<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Seth Levine on Medium]]></title>
        <description><![CDATA[Stories by Seth Levine on Medium]]></description>
        <link>https://medium.com/@levine.seth.p?source=rss-70f6655bd458------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*1qlKbYbJj6_wfku2.</url>
            <title>Stories by Seth Levine on Medium</title>
            <link>https://medium.com/@levine.seth.p?source=rss-70f6655bd458------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 00:36:59 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@levine.seth.p/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Learning from Machine Learning | Sebastian Raschka: Mastering ML and Pushing AI Forward Responsibly]]></title>
            <link>https://medium.com/data-science/learning-from-machine-learning-sebastian-raschka-mastering-ml-and-pushing-ai-forward-responsibly-aac39bd4af83?source=rss-70f6655bd458------2</link>
            <guid isPermaLink="false">https://medium.com/p/aac39bd4af83</guid>
            <category><![CDATA[responsible-ai]]></category>
            <category><![CDATA[interview]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Seth Levine]]></dc:creator>
            <pubDate>Wed, 20 Sep 2023 16:06:58 GMT</pubDate>
            <atom:updated>2025-01-19T21:26:52.505Z</atom:updated>
            <content:encoded><![CDATA[<h4>Sebastian Raschka has helped demystify deep learning for thousands through his books, tutorials and teachings</h4><p>Sebastian Raschka has helped shape how thousands of data scientists and machine learning engineers learn their craft. As a passionate coder and proponent of open-source software, a contributor to scikit-learn and the creator of the mlxtend library, his code runs in production systems worldwide. But his greatest impact is through his teachings — his books <a href="https://amzn.to/3Pk6hXf">Machine Learning with PyTorch and Scikit-Learn</a>, Machine Learning Q and AI and <a href="https://amzn.to/3DK4Gqp">Build a Large Language Model (From Scratch)</a> have become essential guides for practitioners navigating the complex landscape of modern AI.</p><p>Drawing from over a decade of experience building AI systems and teaching at the intersection of academia and industry, Sebastian offers a unique perspective on mastering machine learning fundamentals while staying adaptable in this rapidly evolving field. As Senior Staff Research Engineer at Lighting AI, he continues to bridge the gap between cutting-edge research and practical implementation. In our in-depth discussion in this installment of <a href="https://youtu.be/gOEey-pbM90?si=VkiTSjNb5BJtrKhH&amp;t=21">Learning from Machine Learning</a>, he shared concrete strategies for everything from building reliable production systems to thinking critically about the future of Artificial General Intelligence (AGI).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*XaprQataxU2vdJ_kIJIOpg.png" /><figcaption>What wisdom does one of the world’s top AI educators have for mastering machine learning? (Image by Author)</figcaption></figure><p>Our wide-ranging discussion yielded many insights, which are summarized into 13 key lessons:</p><ol><li><a href="#82d0">Start simple and be patient</a></li><li><a href="#7e61">Learn by doing</a></li><li><a href="#2c47">Always get a baseline</a></li><li><a href="#b4dc">Embrace change</a></li><li><a href="#4c55">Find balance between specialized and general systems</a></li><li><a href="#6deb">Implement from scratch when learning</a></li><li><a href="#d757">Use proven libraries in production</a></li><li><a href="#fd2d">It’s the last mile that counts</a></li><li><a href="#db45">Use the right tool for the job</a></li><li><a href="#04bc">Seek diversity when ensembling models</a></li><li><a href="#c186">Beware of overconfidence</a> (overconfident models :)</li><li><a href="#4fdf">Leverage Large Language Models responsibly</a></li><li><a href="#0775">Have fun!</a></li></ol><h3>Lessons from a Pioneer</h3><h4>1. Start simple and be patient</h4><p>Approach machine learning with patience, taking concepts step-by-step, in order to build a solid foundation. “You should, make sure you understand the bigger picture and intuition.” Grasp the high-level concepts before getting bogged down in implementation details. Sebastian explains, “I would start with a book or a course and just work through that, almost with a blindness on not getting distracted by other resources.”</p><blockquote>“I would start with a book or a course and just work through that, almost with a blindness on not getting distracted by other resources.”</blockquote><p>Borrowing from Andrew Ng, Sebastian shares, “If we don’t understand a certain thing, maybe let’s not worry about it. Just yet.” Getting stuck on unclear details can slow you down. Move forward when needed rather than obsessing over gaps. Sebastian expands, “It happens to me all the time. I get distracted by something else, I look it up and then it’s like a rabbit hole and you feel, ‘wow, there’s so much to learn’ and then you’re frustrated and overwhelmed because the day only has twenty four hours, you can’t possibly learn it all.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/500/0*OCFPM_KDjFyKrGuS.jpg" /><figcaption><a href="https://imgflip.com/i/9h0pu1">Meme</a> by Author</figcaption></figure><p>Remember it’s about “doing one thing at a time, step by step. It’s a marathon, not a sprint.” For early data scientists, he stresses building strong fundamentals before diving into the specifics of advanced techniques.</p><h4>2. Learn by doing</h4><p>“Finding a project you’re interested in is the best way to get involved in machine learning and to learn new skills.” He recalled getting hooked while building a fantasy sports predictor, combining his soccer fandom with honing his data abilities. Sebastian explains, “That’s how I taught myself pandas.” Tackling hands-on projects and solving real problems that you feel passionate about accelerates learning.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/0*CIFrk8YTPRCprB-v" /><figcaption>Combining his soccer fandom with honing his data abilities (<a href="https://hotpot.ai/s/art-generator/8-eRDI5sBGcnV5NlL">link</a>) Created using hotpot.ai/art-generator</figcaption></figure><blockquote>My first project in machine learning… was a fun one… I was working on fantasy sports predictions back then. I was a big soccer fan. Based on that I built machine learning classifiers with scikit-learn, very simple ones, to basically predict [who] the promising players were, and that was very interesting as an exercise because that’s how I taught myself pandas… I tried to automate as much as possible, so I was also trying to do some simple NLP, going through news articles, basically predicting the sentiment and extracting names from players who were injured and these types of things. It was very challenging, but it was a very good exercise to learn data processing and implementing simple things.”</blockquote><h4>3. Always get a baseline</h4><p>When beginning a new ML project you should always find some baseline performance. For example when starting a text classification project, Sebastian says, “Even if you know more sophisticated techniques, even if it makes sense to use a Large Language Model… Start with a simple logistic regression, maybe a bag of words to get a baseline.”</p><p>By building a baseline before trying more advanced techniques you can get a better understanding of the problem and the data. If you run into issues when implementing more advanced techniques, having a baseline model where you already read and processed the data can help debug more complex models. If an advanced model underperforms the baseline, it may be an indicator that there are data issues rather than model limitations.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/0*KSYw5lVjIg3B_cX6" /><figcaption>Getting a baseline in a machine learning model for a complex task (<a href="https://hotpot.ai/s/art-generator/8-Lpd6a5yPQraWp6a">link</a>) Created using hotpot.ai/art-generator</figcaption></figure><blockquote>“I would say always start with [simple techniques] even if you know more sophisticated techniques if we go back to what we talked about with large language models even if it makes more sense for a classification problem to fine-tune a large language model for that, I would start… with a simple logistic regression classifier, maybe bag-of-words model to just get a baseline. Use something where you are confident, it’s very simple and it works, let’s say using scikit-learn before trying the more complicated things. It’s not only because we don’t want to use the complicated things because the simple ones are efficient, it’s more about also even checking our solutions like if our fine-tuned model or let’s say BERT or LLM performs worse than the logistic regression classifier maybe we have a bug in our code, maybe we didn’t process the input correctly, [maybe we didn’t] tokenize it correctly - it’s always a good idea to really start simple and then increasingly get complicated or improve - let’s say improve by adding things instead of starting complicated and then trying to debug the complicated solution to find out where the error is essentially.”</blockquote><h4>4. Embrace change</h4><p>The field is changing quickly. While it’s important to start slow and take things step by step it is equally important to stay flexible and open to adopting new methods and ideas. Techniques and approaches in machine learning have a tendency to come in and out of style.</p><p>Sebastian stresses the importance of adaptability amid relentless change. “Things change completely. We were using [Generative Adversarial Networks] GANs [a few years ago] and now we’re using diffusion models… [be] open to change.” Machine learning rewards the nimble. He emphasizes being open to new experiences both in machine learning and life.</p><h4>5. Find balance between specialized and general systems</h4><p>The pursuit of Artificial General Intelligence (AGI) is a worthy goal but specialized systems often provide better results. Depending on the use case, a specialized system may be more appropriate than a one-size-fits-all approach. Sebastian discusses how systems may be a combination of smaller models where the first model is used to determine which specialized model the task should be directed to.</p><p>Regardless, the pursuit for AGI is an incredible motivator and has led to many breakthroughs. As Sebastian explains, the quest for AGI pushed breakthroughs like DeepMind’s AlphaGo beating the best humans at Go. And while AlphaGo itself may not be directly useful, “it ultimately led to AlphaFold, the first version, for protein structure prediction.”</p><p>The dream of AGI serves as inspiration, but specialized systems focused on narrow domains currently provide the most value. Still, the race towards AGI has led to advances that found practical application.</p><blockquote>“I think no one knows how far we are from AGI… I think there’s a lot more hype around AGI it appears closer than before of course because we have these models. There are people though who say okay this is the totally wrong approach we need something completely different if we want to get AGI no one knows what that approach looks like so it’s really hard to say…</blockquote><blockquote>…The thing though what I always find interesting is do we need AGI more like a philosophical question… <strong>AGI is useful as the motivation</strong>. I think it motivates a lot of people to work on AI to make that progress. I think without AGI we wouldn’t have things like AlphaGo where they had the breakthrough they basically beat the best player at go… how is that useful - I would say maybe go and chess engines are not useful but I think it ultimately led to AlphaFold the first version for protein structure prediction and then AlphaFold 2 which is not based on large language models but uses large language models. So in that case I think without large language models and without the desire maybe to develop AGI <strong>we wouldn’t have all these very useful things in the Natural Sciences</strong> and so my question is <strong>do we need AGI or do we really just need good models for special purposes…</strong>”</blockquote><h4>6. When learning, implement from scratch</h4><p>Coding algorithms without depending on external libraries (e.g., using just Python) helps build a better understanding of the underlying concepts. Sebastian explains, “Implementing algorithms from scratch helps build intuition and peel back the layers to make things more understandable.”</p><blockquote>“Implementing algorithms from scratch helps build intuition and peel back the layers to make things more understandable.”</blockquote><p>Fortunately, Sebastian shares many of these educational implementations through posts and tutorials. We dove into Sebastian’s breakdown of <a href="https://sebastianraschka.com/blog/2023/self-attention-from-scratch.html">Self-Attention of LLMs from Scratch</a> where he breaks down the importance of the “self-attention” mechanism which is a cornerstone of both transformers and stable-diffusion.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DidT8VniVRFn6p6aWOlBVA.jpeg" /><figcaption>Two cavemen reinventing the wheel (generated <a href="https://creator.nightcafe.studio/creation/qDnDBBHRcNg29llDdOoD/two-cavemen-computer-scientists-reinventing-the-wheel?ru=Splevine">here</a>)</figcaption></figure><h4>7. In production, don’t reinvent the wheel!</h4><p>In real world applications, you don’t have to reinvent the wheel. Sebastian expands for things that already exist, “I think that is a lot of work and also risky.” While building from scratch is enlightening, production-ready applications rely on proven, battle-tested libraries.</p><blockquote>“what I did was for education… let’s implement a principal component analysis from scratch or let’s implement a self-attention mechanism from scratch and writing the code but not necessarily as a library because I think <strong>there are already a lot of efficient implementations out there so it doesn’t really make sense to reinvent the wheel</strong> but it’s more about let’s peel back a few layers make a very simple implementation of that so that people can read them because that’s one thing — deep learning libraries are becoming more powerful. If we look at PyTorch for example but they are also becoming much much harder to read — so if I would ask you to take a look at the convolution operation in PyTorch I wouldn’t even understand… I wouldn’t even know where to look… to start with it… I mean for good reason because they implemented it very efficiently and then there’s cuda on top of that… but as a user if I want to customize or even understand things it’s very hard to look at the code so in that case I think there’s value in peeling back the layers making a simple implementation for educational purposes to understand how things work.</blockquote><h4>8. It’s the last mile that counts</h4><p>Getting a model to relatively high performance is much easier than squeezing out the last few percentage points to reach extremely high performance. But that final push is vital — it’s the difference between an impressive prototype and a production-ready system. Even if rapid progress was made initially, the final seemingly marginal gains to reach “perfection” are very challenging.</p><blockquote>Even if rapid progress was made initially, the final seemingly marginal gains to reach “perfection” are very challenging.</blockquote><p>Sebastian uses self-driving cars to drive this point across. “Five years ago, they already had pretty impressive demos… but I do think it’s the last few percent that are crucial.” He continues, “Five years ago, it was almost let’s say 95% there, almost ready. Now five years later, we are maybe 97–98%, but can we get the last remaining percent points to really nail it and have them on the road reliably.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DqYdNPnIggvoN9ajmUMB8A.png" /><figcaption>This final push, though it may seem marginal in terms of numerical improvement, can be the most challenging and crucial step in the development process. (Image by Author)</figcaption></figure><p>Sebastian draws a comparison between ChatGPT and Self-Driving cars. While astounding demos of both technologies exist, getting those last few percentage points of performance to reach full reliability has proven difficult and vital.</p><h4>9. Use the right tool for the job</h4><p>Sebastian cautions against forcing ML everywhere, stating “If you have a hammer, everything looks like a nail… the question becomes when to use AI and when not to use AI.” The trick is often knowing when to use rules, ML, or other tools. Sebastian shares, “Right now, we are using AI for a lot of things because it is exciting, and we want to see how far we can push it until it breaks or doesn’t work… sometimes we have nonsensical applications of AI because of that.”</p><p>Automation has limits. Sometimes rules and human expertise outperform AI. It’s important to pick the best approach for each task. Just because we can use AI/ML as a solution doesn’t mean we should for every problem.</p><blockquote>“[There’s] a saying <strong>if you have a hammer everything looks like a nail</strong>, and I think this is right now a little bit true with ChatGPT because we just have fun with it… let me see if it can do this and that but it doesn’t mean we should be using it for everything… now the question is basically the next level… <strong>when to use AI and when not to use AI… </strong>because right now we are using AI for a lot of things because it’s exciting and we want to see how far we can push it until it let’s say breaks so it doesn’t work but sometimes we have nonsensical applications of AI because of that. …like training a neural network that can do calculation… but we wouldn’t let it do the math matrix multiplication itself because you know it’s non-deterministic in a sense so you don’t know if it’s going to be correct or not depending on your inputs and there are definite rules that we can use so why approximate when we can have it accurate”</blockquote><h4>10. Seek Diversity in Model Ensembles</h4><p>Ensemble methods like model stacking can improve prediction robustness, but diversity is key — combining correlated models that make similar types of errors won’t provide much upside.</p><p>As Sebastian explains, “Building an ensemble of different methods is usually something to make [models] more robust and [produce] accurate predictions. And ensemble methods usually work best if you have an ensemble of different methods. If there’s no correlation in terms of how they work. So they are not redundant, basically.”</p><p>The goal is to have a diverse set of complementary models. For example, you might ensemble a random forest model with a neural network, or a gradient boosting machine with a k-nearest neighbors model. Stacking models that have high diversity improves the ensemble’s ability to correct errors made by individual models.</p><p>So when building ensembles, seek diversity — use different algorithms, different feature representations, different hyperparameters, etc. Correlation analysis of predictions can help identify which models provide unique signal vs redundancy. The key is having a complementary set of models in the ensemble, not just combining slight variations of the same approach.</p><blockquote>“…<strong>building an ensemble of different methods is usually something to improve how you can make more robust and accurate predictions and ensemble methods usually work best if you have an ensemble of different methods — if there’s no correlation in terms of how they work. So they are not redundant</strong> basically. That is also one argument why it makes sense to maybe approach the problem from different angles to produce totally different systems that we can then combine.”</blockquote><p>Models with varying strengths and weaknesses can effectively counterbalance each other’s shortcomings, leading to more reliable overall performance.</p><h4>11. Beware of overconfidence</h4><p>“There’s a whole branch of research on [how] neural networks are typically overconfident on out of distribution data.” ML predictions can be misleadingly overconfident on unusual data. Sebastian describes, “So what happens is if you have data that is slightly different from your training data or let’s say out of the distribution, the network will if you program it to give a confidence score as part of the output, this score for the data where it’s especially wrong is usually over confident… which makes it even more dangerous.” Validate reliability before deployment, rather than blindly trusting in confidence scores. Confidence scores can often be high for wrong predictions making them misleading for unfamiliar data.</p><blockquote>Validate reliability before deployment, rather than blindly trusting in confidence scores.</blockquote><p>To combat overconfidence in practice, start by establishing multiple validation sets that include both edge cases and known out-of-distribution examples, keeping a separate test set for final verification. A robust monitoring system is equally crucial — track confidence scores over time, monitor the rate of high-confidence errors, set up alerts for unusual confidence patterns, and maintain comprehensive logs of all predictions and their associated confidence scores.</p><p>For production systems, implement fallback mechanisms including simpler backup models, clear business rules for low-confidence cases, and human review processes for highly uncertain predictions. Regular maintenance is essential: as new data becomes available it may be worthwhile to retrain models, adjust confidence thresholds based on real-world performance, fine-tune out-of-distribution detection parameters, and continuously validate model calibration. These practices help ensure your models remain reliable and self-aware of their limitations, rather than falling into the trap of overconfidence.</p><h4>12. Leverage Large Language Models responsibly</h4><p>ChatGPT (and other generative models) are good brainstorming partners and can be used for ideation when “it doesn’t need to be 100% correct.” Sebastian warns that the model’s output should not be used as the final output. Large language models can generate text to accelerate drafting but require human refinement. It’s important to be fully aware of the limitations of the LLMs.</p><h4>13. Remember to have fun!</h4><p>“Make sure you have fun. Try not to do all at once.” Learning is most effective and sustainable when it’s enjoyable. Passion for the process itself, not just outcomes, leads to mastery. Sebastian emphasizes to remember to recharge and connect with others who inspire you. Sebastian shares, “Whatever you do, have fun, enjoy, share the joy… things are sometimes complicated and work can be intense. We want to get things done, but don’t forget… to stop and enjoy sometimes.”</p><h3>Conclusion</h3><p>While the field’s rapid advancement and complexity can be overwhelming, Sebastian offers a clear path forward: build rock-solid fundamentals, always start with baseline models, and maintain systematic approaches to combat common pitfalls. He advocates for implementing algorithms from scratch before using high-level, optimized libraries to ensure deep understanding. He offers practical strategies — such as including diversity in ensemble models, critically assessing model confidence, and recognizing the difficulty of the “last mile” — for developing reliable and dependable production-quality AI systems.</p><p>Sebastian stresses that <strong>mastering machine learning isn’t about chasing every new development</strong>. Instead, it’s about building a strong foundation that enables you to evaluate and adapt to meaningful advances. By focusing on core principles while remaining open to new techniques, we can build the confidence to face increasingly complex challenges. Whether you’re implementing your first machine learning project or architecting enterprise-scale AI systems, the key is to embrace the learning process: start simple, evaluate thoroughly, and never stop questioning your assumptions. In a field that seems to reinvent itself almost daily, these timeless principles are our most reliable guides.</p><h3>Check out the Full Interview</h3><p>Enjoy these lessons and check out the full <a href="https://rss.com/podcasts/learning-from-machine-learning/">Learning from Machine Learning</a> interview here:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FgOEey-pbM90%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DgOEey-pbM90&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FgOEey-pbM90%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/2a92d558576fd43d0e05b8a625c3c383/href">https://medium.com/media/2a92d558576fd43d0e05b8a625c3c383/href</a></iframe><p>Listen on your favorite podcast platform:</p><p><a href="https://rss.com/podcasts/learning-from-machine-learning/881901/">Learning from Machine Learning - Sebastian Raschka: Learning ML, Responsible AI, AGI | Learning from Machine Learning #4 | RSS.com</a></p><p>Resources to learn more about Sebastian Raschka and his work:</p><ul><li><a href="https://sebastianraschka.com/">Sebastian Raschka</a></li><li><a href="https://lightning.ai/">Lightning.ai</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=aac39bd4af83" width="1" height="1" alt=""><hr><p><a href="https://medium.com/data-science/learning-from-machine-learning-sebastian-raschka-mastering-ml-and-pushing-ai-forward-responsibly-aac39bd4af83">Learning from Machine Learning | Sebastian Raschka: Mastering ML and Pushing AI Forward Responsibly</a> was originally published in <a href="https://medium.com/data-science">TDS Archive</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Learning from Machine Learning | Vincent Warmerdam: Calmcode, Explosion, Data Science]]></title>
            <link>https://medium.com/data-science/learning-from-machine-learning-vincent-warmerdam-calmcode-explosion-data-science-e07951a07cf4?source=rss-70f6655bd458------2</link>
            <guid isPermaLink="false">https://medium.com/p/e07951a07cf4</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[career-advice]]></category>
            <category><![CDATA[deep-dives]]></category>
            <category><![CDATA[interview]]></category>
            <dc:creator><![CDATA[Seth Levine]]></dc:creator>
            <pubDate>Wed, 14 Jun 2023 17:22:50 GMT</pubDate>
            <atom:updated>2023-08-23T22:26:02.738Z</atom:updated>
            <content:encoded><![CDATA[<h4>The Past, Present, and Future of Data Science: An Interview with Vincent Warmerdam, Explosion ML Engineer and creator of calmcode</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FP9LdeC0RkF6Lcj7uIgYDA.jpeg" /><figcaption>Image by Author</figcaption></figure><p>Welcome back to “<a href="https://www.youtube.com/@learningfrommachinelearning">Learning from Machine Learning</a>,” a series of <a href="https://rss.com/podcasts/learning-from-machine-learning/">interviews</a> that delves into the fascinating world of machine learning. As AI revolutionizes our world, each episode offers technical insights along with career advice and life lessons from leading practitioners in the field.</p><p>In this <a href="https://youtu.be/yvgxRzqx1Jg">episode</a>, Vincent Warmerdam, creator of calmcode and <a href="https://koaning.io/">koaning.io</a>, ML Engineer at Explosion, and former Research Advocate at Rasa, shares his experiences and advice. He has held numerous roles in the data science world including consultant, advocate, trainer, lead, educator and recruiter. His work has focused on simplifying processes, utilizing models effectively and educating others.</p><p>His site <a href="http://calmcode.io/">calmcode</a> provides high-quality, easily consumable educational data science content and has gained popularity and garnered positive feedback. It currently attracts thousands of users monthly and has become a valuable resource for data scientists. Vincent helped organize PyData Amsterdam and he has consistently been a speaker at PyData and NormConf. His talks are enjoyable and engaging (check them out on YouTube).</p><p>Vincent has created dozens of tools to help in machine learning development, data processing, testing, and natural language processing.</p><p><a href="https://github.com/koaning">koaning - Overview</a></p><p>Vincent’s passion for building practical tools, sharing his work openly and addressing real-world problems with simplicity and calmness has influenced the way that I do my work.</p><h4>Summary</h4><p>Vincent has had an unconventional career journey and entered data science while it was still emerging. He attributes his success to a combination of luck, the popularity of his blog, organizing meetups, and his open-source contributions, which resulted in job offers directly from CEOs and CTOs instead of going through the traditional recruitment processes. His academic background includes studying operations research and design which gave him a unique perspective on algorithms and constraints. He also mentioned that his gig as a bartender at a Comedy Club helped sharpen his presentation skills.</p><p>Inspired by a family member’s wrong medical diagnosis, Vincent was attracted to machine learning to use algorithms to make better decisions and predictions. During his Masters thesis he took the opportunity to focus on machine learning.</p><p>His open source work has been driven by his desire to solve specific problems and sometimes to scratch his own itches. He started with his first PyPI project, evol, and developed libraries like scikit-lego, human-learn, whatlies, and doubtlab based on his needs and experiences at different companies.</p><p>The bulk library emerged as a result of his work on human-learn and whatlies. It leverages embeddings, dimensionality reduction and clustering to facilitate bulk labeling and data exploration. While the tool does not provide perfect labels, it helps lower the barrier to entry and build intuition when working on new datasets.</p><p>He started calmcode as a response to the lack of quality educational content in data science, aiming to provide concise and opinionated tutorials that focus on getting the idea across effectively. His lessons consist of a series of short, easily digestible videos. He focuses on teaching tools and approaches to improve day-to-day data science work.</p><p>Vincent’s experience in the field provided great perspective on the current state of machine learning. We were able to discuss the limitations of generative models, the role of rule-based systems to complement ML models and the complexities of ground truth labeling.</p><p>Vincent’s advice was real and poignant. He cautions against placing too much emphasis on monetary gains early in your career and to prioritize discovery of your passion and true motivations. He grapples with the contrasting feelings of the usefulness of the ML field and the need for skilled professionals in it while also recognizing the abundance of hype. Most importantly, Vincent emphasizes gaining a clear understanding of what you can and cannot influence and to focus your energy on areas where you can make a difference.</p><h4>Advice &amp; Takeaways</h4><ol><li>Understanding the problem correctly is most important — a better algorithm may not always yield better outcomes if applied to the wrong problem.</li><li>Rephrasing problems can be beneficial — think holistically about system optimization rather than focusing solely on improving individual components.</li><li>“People often forget that the algorithm is usually just a cog in the system. And we’re interested in building a better system, not a better cog. So, if you’re building a better cog but it doesn’t fit the rest, it’s not a better cog because you don’t get a better system.”</li><li>Consider starting a blog where you share what you learn. Writing short “today I learned” snippets can be an easy way to build an online presence and demonstrate your continuous learning.</li><li>Don’t feel the pressure to know everything when starting out. Consider related roles like being an analyst, which can provide valuable skills and knowledge that will benefit your data science journey.</li><li>“You can make python packages more often than you might think, so just build one.” More people should create their own Python packages, even for internal tools or small helper functions. Not enough people take advantage of the benefits of packaging and reusing their own code.</li><li>Stepping outside the machine learning bubble can be refreshing and provide inspiration. Interacting with people who are not in the field and understanding their perspectives can help you create better applications and focus on the human aspect of your work.</li><li>Be careful of getting too focused on money early in your career. It’s important to figure out what you enjoy and what makes you tick rather than just chasing a higher salary.</li><li>“I’m kind of a mixed bag when it comes to the whole machine learning thing. Part of my opinion is it’s a super useful tool and we need more good people doing machine learning. But at the same time, it’s like a gross bucket of hype that we really want to have less of. And my day to day is to sort of deal with both of these feelings.”</li><li>“Calling <em>.fit</em> and <em>.predict</em> are the easy bits. It’s all the stuff around that. It’s way trickier. Especially when you consider themes of fairness, all the things that can go wrong, can we really know that upfront? I don’t know if you always can.”</li><li>“There’s some stuff that you can control, some stuff that you can’t. Just make sure that you understand what you can and can not control and then move on from there.”</li></ol><h3>Table of Contents</h3><blockquote><a href="#26e3">Introduction</a><br><a href="#89a2">Summary</a><br><a href="#a5c2">Advice &amp; Takeaways<br></a><a href="#c58f">Full Interview</a><br> — <a href="#78d6">Welcome</a><br> — <a href="#eccb">Background</a><br> — <a href="#fb97">Academic Background</a><br> —<a href="#8fa8"> Machine Learning Attraction</a><br> — <a href="#e684">Calmcode</a><br> — <a href="#b85e">Open Source</a><br> — <a href="#7396">Bulk</a><br> — <a href="#4198">Understanding the Problem</a><br> —<a href="#ef45"> Unanswered Questions in Machine Learning</a><br> — <a href="#5bf3">Generative and Predictive Machine Learning</a><br> — <a href="#7a2a">Influences</a><br> — <a href="#f1fc">Career Advice</a><br> — <a href="#ad30">Advice for New Data Scientists</a><br> — <a href="#5b18">Learning from Machine Learning</a><br><a href="#3b15">Video Interview</a><br><a href="#05de">Spotify Audio</a><br><a href="#9210">Previous Episode featuring Maarten Grootendorst</a><br><a href="#a6b6">References</a></blockquote><h3>Full Interview</h3><p>(Note: This interview was lightly edited for clarity)</p><h4>Welcome</h4><p><strong>Seth</strong>: Welcome to Learning from Machine Learning. On this episode, I have the pleasure of having Vincent Warmerdam. He’s currently a machine learning engineer at Explosion, the company behind SpaCy and Prodigy. Vincent is an educator, a blogger, consistent PyData speaker.</p><p>He’s created many valuable open source tools. He’s endorsed for awesomeness on LinkedIn over a hundred times. And truly an inspiring force in the data science community. Welcome to the podcast.</p><p><strong>Vincent</strong>: Hi. That comment. It’s making me check. Do I have a hundred endorsements? I didn’t know.</p><p><strong>Seth</strong>: Over.</p><p><strong>Vincent</strong>: Oh, okay, cool.</p><p>It’s an inside joke I have with a former colleague whoever can get the most. I’ve got more than ninety nine endorsements for awesomeness on LinkedIn now, Nice!</p><h4>Background</h4><p><strong>Seth</strong>: Why don’t you give us some background on your career journey? How did you get to where you are today?</p><p><strong>Vincent</strong>: It’s a little bit hard to give proper career advice because I want to just recognize I’m a little bit privileged and I got lucky a whole bunch because when I started this whole data science thing, this was the era when random forests were kind of new. And if you could just use a random forest, you’re already way better than all the economic traditions with linear vols. So can you run fit and predict — bang. You’ve got a job. And, I was kind of in that era at the right time.</p><p>After that though, I started blogging. I started helping out by arranging some meetups. There’s a machine learning meetup in Amsterdam — I helped organize PyData Amsterdam.</p><p>And I had a pretty popular blog as well, so people started recognizing me for that. And, at some point, that recognition gets you places, you get invited to speak, and then people sort of see you as an authority figure. I’m not, I like to think I have sensible ideas, but I try to be modest about that. But that has kind of been the story of my career though because people knew my blog, which would usually include a CTO of a company. And then a CTO would say, “Hey, I like your blog. Can we just have a beer?”</p><p>And then usually, that led to the job offer. I have yet to talk to a recruiter. Honestly. I’ve never been hired through the recruitment pipeline so far in my career.</p><p>It’s always been via the CEO or CTO because they knew of my work beforehand. And this is very weird, this is a very bad story because this is very hard to replicate for others because I just got very lucky when it comes to this. I do think having a blog and being able to get recognized is very useful, just very hard to replicate. But I also like to think that some of the side projects that I did for open source definitely helped out. Calmcode is something that people know me of these days. There’s a saying, plant a thousand flowers and one of them will be a lotus.</p><blockquote>…plant a thousand flowers and one of them will be a lotus</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*4K3OiUwl-RdPUGR3BJr--w.jpeg" /><figcaption>Image generated using <a href="https://huggingface.co/CompVis/stable-diffusion-v1-4">https://huggingface.co/CompVis/stable-diffusion-v1-4</a> with prompt “Plant a thousand flowers and one of them will be a lotus”</figcaption></figure><p>I subscribe to that idea, but a lot of it comes down to luck and a bit of privilege as well. I just want to be honest about that. Being able to be recognizable has proven to be useful to me.</p><p><strong>Seth</strong>: Yeah. Definitely. I mean, everyone’s journey is unique. What are the roles that you’ve played at different companies?</p><p>I know you’ve been a consultant at one point, an advocator. You’ve had some interesting titles. Right?</p><p><strong>Vincent</strong>: There was a phase at a previous company where they would really let you pick any name you could and me and a couple of colleagues would think it’s funny to really see how far we could stretch it. So I called myself a Pokemon master because I thought that it was kind of funny. My favorite title was senior person because at some point, I’m just one of the old guys at the company. So I just called myself that.</p><p>As far as roles goes, I’ve been doing lots of training, so at some point you call yourself a trainer. Because you’re doing consultancy, you have different roles within different teams. So I was a lead at some point as well.</p><p>I was also helping a recruiting company recruit people at companies for specific data team. Sometimes my role would be super temporary. Sometimes it’d be for two years. But usually, it’s been I was a person on a data team trying to get the team to become productive, doing whatever. Usually, the things I like to think I’m pretty good at is keeping things very simple.</p><p>I can get a lot of mileage from linear models, which tends to work quite nicely. And before that, I also had lots of different jobs in college, and I do like to think that that kind of helped as well. My background is in operations research, but before that, I actually studied design. And my evening job, I was a bartender at a comedy theater in the Netherlands, which I do like to think that might have helped my presentation skills in a way. Having studied design for a while also makes me think differently about algorithms.</p><p>And how operations research makes me think about constraints a lot when I’m doing data science things. I like to think that I have a somewhat diverse background, and it’s that somewhat diverse background that makes it easy for me to do stuff that I’m doing now. That’s a summary.</p><h4>Academic Background</h4><p><strong>Seth</strong>: That makes sense. So diving into your academic background, what was it? So, it was some operations research and some design? That sounds pretty unique there.</p><p><strong>Vincent</strong>: Yeah, I studied industrial design engineering for a year, and then I found out it wasn’t for me. So I had to switch majors. And then the bachelor’s was econometrics and operations research, and the masters was operations research. And I thought that it was the most interesting of the two. But also because that had a little gateway to computer science.</p><p>So if I wanted to do the computer science courses, that’d be kind of an easy way for me to get courses I wanted to take. But, I was also rowing when I was in college and did a bit of partying, so I wasn’t allowed to start my masters just yet. So I had this one year where I just took whatever course interested me. I did a couple of courses in neuroscience, which was pretty interesting, some psychology and biology and what not.</p><p>Again, I like to think that the diversification of knowledge has proven to be quite useful, but the official title for my academic background, Operations Research. So that’s the mathematics behind optimizing systems. That’s kind of the thing that you’re taught there.</p><p><strong>Seth</strong>: Very cool. What’s the quintessential or canonical problem in operations research that that people try to solve?</p><p><strong>Vincent</strong>: So, a very textbook example in machine learning, you usually try to optimize towards something. You want to get the loss as low as possible or the accuracy as high as possible. And you’ve got algorithms for that. You typically take your data and the label you want to predict and you come up with some sort of loss function, you try to get those as small as possible. And in operation research, you kind of do a very similar thing.</p><p>It’s just that in operation research, you typically aren’t dealing with a machine learning algorithm, but you’re dealing with, let’s say, “Hey, we have stocks that we would like to invest in and Oh, by the way, that also introduces constraints. Because, yes, we want to get the highest return, but of course, we don’t want to overshoot the budget, and we also have a risk preference.” And those are defined as hard mathematical constraints. Then if you want to optimize, it’s kind of a different ball game because if your algorithm ever exceeds the bounds of the constraint, then you’re in sort of bad territory. So I would argue that’s he main thing.</p><p>If you’re doing operations research then you’re taught constraints really matter and you want to deal with those in mathematically proper ways. And that’s a different ballgame. But that’s the main thing that they’re dealing with, this constrained optimization. That’s what they do.</p><p><strong>Seth</strong>: Very cool. Yeah. Sounds like you need a very strong math background. Some linear algebra in there too?</p><p><strong>Vincent</strong>: Lots of calculus and linear algebra. Although, I will say it depends a bit on what you do. When you do the master’s degree, of course, you do theoretical courses and you got to do the proofs. But the moment that you start doing your thesis [things change]. I had a professor who basically said, “I know nothing about machine learning Vincent, but you seem so eager. You do machine learning and you just teach me how it works because I’ve got no idea.”</p><p>And that was great. The professors just let me do what I wanted to do, and I was also able to teach myself that way. But if you really want to do the proper operations research and especially if you want to do a PhD, It is super math heavy. That’s true.</p><p>A little bit too heavy for my comfort to be honest. I’m a little bit more on the applied side of things, but still know people who end up doing a PhD, and they are definitely the math proof kind of people. They’re the kind of this cookbook of linear algebra kind of a person. That’s definitely in the field.</p><h4>Machine Learning Attraction</h4><p><strong>Seth</strong>: What was it that attracted you to machine learning? What initially got you interested in it?</p><p><strong>Vincent</strong>: There’s always something cool about making predictions. Right? So there’s something about that I thought was pretty interesting. I think the longer story, though, I do remember a very close family member of mine got a wrong medical diagnosis.</p><p>The wrong diagnosis was they told the person you have a very bad disease and the person didn’t. And we found out just in time, thank God. But, they might have made some really weird life decisions like sell the house immediately because of that decision. So that made me think, okay, there’s a definite consequence to making the wrong decisions. Anything that we can do to make better decisions is interesting.</p><p>And maybe there’s something about this machine learning. The whole idea that you try to learn more from data by using a machine, there’s something plausible about that. It seems very interesting. So that was kind of around the time that I did think, Hey, yeah, Let’s see if these algorithms might be able to do something.</p><p>And then the career prospects turned out to be amazing. So that’s another motivation to go into that realm. But the initial spark was a wrong decision got made. That’s how I started thinking, hey, maybe there are systems that we can improve here.</p><h4>Calmcode</h4><p><strong>Seth</strong>: Very interesting, and we’re going to dive into machine learning in a bit. But having the creator and maintainer of calmcode which is such an unbelievable resource.</p><p>[Vincent Laughs]</p><p><strong>Seth:</strong> No, nothing nothing to laugh about. Calmcode is incredible.</p><p>It’s in the top two or three things that I recommend to every new data scientist. The way that you break down really complex things into a nice calm and logical and rational way is extremely valuable.</p><p><strong>Vincent:</strong> Glad to hear.</p><p><strong>Seth:</strong> So can you talk a little bit about calmcode and why’d you start it? And well, what is it? Give everyone a breakdown</p><p><strong>Vincent</strong>: So first of all, happy to hear it — that you like calmcode. I’m happy to hear that it helps. So, basically, the story behind calmcode was, at some point, I was looking at educational content around data science. And, as an educator, I just started noticing there’s just so much gunk.</p><p>So to give an example, the number one tutorial, maybe four or five years ago, on how scikit-learn works, what happened with this data set called Load Boston, which is about Boston House prices. There’s so many tutorials that use that data set from all your O’Reily books to a lot of open source packages. But then you look at the data and it turns out that one of the variables that they’re using to predict the house price is skin color.</p><p>I forgot the exact name, but it was something like percentage of blacks in the town. You don’t want to put that in the predictive model. It’s a really, really bad idea. Also, why is this dataset on scikit-learn? Why are so many people using it?</p><p>So that led to a lot of frustration on my end. And then I also noticed that there are these enterprise courses that use Load Boston that charge a thousand dollars a day. And you look at it, you kind of go, this is a mess. And then I figured if I’m this frustrated, maybe I can get energy out by putting this stuff out there for free.</p><p>I knew that I was knowledgeable enough to be able to teach these kinds of topics because I’ve taught them before. But I’ve also noticed that a lot of this educational content seems to focus more on the creator and less about just getting the idea across. So I figured it would be kind of like a fun little experiment. If I were to make a learning platform, how would I do it? And that’s how CalmCode got created.</p><p>The idea is you just have a maximum of five one minute videos to explain a single topic and the sequence of those can be a little course on pandas or can be a little course on whatever.</p><blockquote>…for every single topic I can say, Is this a calm tool? Is it something that makes your day to day nicer? And if the answer is maybe not, then I just don’t teach it.</blockquote><p>What I like about doing the calmcode thing is for every single topic I can say, Is this a calm tool? Is it something that makes your day to day nicer? And if the answer is maybe not, then I just don’t teach it, which is also one of the reasons why I don’t teach Spark, to be honest. Because installing it is just such a pain. And sometimes there’s easier ways of analyzing the data than resorting to a very big data tool.</p><p>So it’s just a very opinionated learning environment that people seem to have really liked. I’ve gotten lots of very nice responses. Ever since the baby showed up, I have been doing way less.</p><p>But it has been very cool to see just this little hobby project of mine without distractions that’s very calm. Seems to be getting between ten and twenty thousand people a month. And I get lots of people buying me beers at conferences out of the blue. That kind of stuff is pretty cool.</p><p><strong>Seth</strong>: Yeah. It’s a great resource. Definitely paying it forward, creating a place for data scientists to go to learn anything. Have you ever found yourself going back to an old calmcode to refresh your self on some of these skills?</p><p><strong>Vincent</strong>: Yeah. That’s another reason. So one thing that calmcode has proven to be quite nicely for me is it’s kind of like a snippets library. Because I knew the course that I made and I knew that I definitely mentioned this in that course and I kind of need a config file, where is it?</p><p>Just today, I was looking at my typer course because I needed “oh, how do options work again? Copy paste.” So it’s also almost a snippet tool from myself at this point as well. Not the original intent, but it is something that seems to be happening.</p><p>And also, I’m building search for calmcode now as well. It’s kind of as a hobby project. I’m contemplating, hey, maybe the main thing the search feature should do is just find the right snippets, which is kind of like an interesting search problem on its own.</p><p>But, yeah, totally, I need a reminder too. There’s many courses. I don’t have all of them in my head, all of the time. I still watch my own stuff in that sense. Yeah.</p><p><strong>Seth</strong>: Yeah. It’s a good resource for you that you can consume which also became something so many other people can use. I’m trying to think of my first usage of it. I think it was, like, args kwargs, which is one of the first ones on there. Yeah. I revisit it every so often.</p><p><strong>Vincent</strong>: Nice. Yeah. Nice.</p><p><strong>Seth:</strong> Thank you.</p><p><strong>Vincent:</strong> Well, so then I would love to do more.</p><p>But the simple fact of the matter is my life is a little bit different now because of the baby. So, there’s so many ideas I have that I could do with calmcode. The one thing I also kind of like about the project is I cannot spend any effort on it and the site will just still run. Right. So that’s also kind of the calm design of it.</p><p>I really like having a hobby project where it’s impossible to break. And if any of it breaks, it’s super easy to fix because it’s just a static website. So that makes it super easy.</p><p><strong>Seth</strong>: If there were no time constraints or any resource constraints, what would you do to improve calmcode?</p><p><strong>Vincent</strong>: There’s a couple of courses particularly that I would love to do. One of them is just embeddings I think that there seems to be a bit of hype around it, but also just you can make embeddings do different things and there’s reasons why they work. But they don’t solve every problem and I can do a fun course where you start with letter embeddings and you move on to other embeddings and images, and then you also show how they can fail. I think that that could be super cool.</p><p>Bayesian MCMC [Markov Chain Monte Carlo] stuff, would be nice to have as well because you can make very articulate models, which is a trick not enough people are appreciative enough of.</p><p>And then I would love to have a new section on the site, which is all about demos and benchmarks. And that’s because I think it’s very hard to do a benchmark wrong. All benchmarks are wrong, but some of them can be very insightful. And I think just celebrating that a bit more would also just be fun. I’ve got a couple of examples lined up, but I have no time to actually produce it.</p><p>But things like, hey, what can you do to actually make numeric algorithms converge a bit quicker. Does standardization really help or not? And, just exploring that a little bit could be just super fun.</p><p>Stuff like that’s in my mind. There’s always stuff to make. And, like, another thing I’m playing with is, like, would it be fun to collaborate on that project? Maybe. I don’t know. But of course, There’s no rush. So it’s also kind of fine if I don’t spend time on it right now. That’s also cool.</p><h4>Open Source</h4><p><strong>Seth</strong>: Yeah. Awesome. Switching gears a little bit into some of your open source work. I think the first library of yours that I was exposed to was bulk. Maybe something else before that but that was the first one I was really using. And then you also have embetter, human-learn, whatlies, doubtlab, cluestar. Those are the ones that I’m most familiar with. I know there’s about another two or three dozen</p><p><strong>Vincent</strong>: Yeah. Small dozen at this point.</p><p><strong>Seth</strong>: When do you decide like — this project deserves an open source library? When you think it’s a tool?</p><p><strong>Vincent</strong>: It helps a bit to maybe explain how the open source thing kind of got started. So my first open source project that I put on PyPI was called evol, which is basically a DSL for evolutionary programming. I made it with a colleague of mine. It was a very cute idea. And I wanted to have my own little library.</p><p>So I was looking for a problem. And then I just find out that, if I have a population object and evolution object, and those two can interact in nice ways and super easy to make genetic algorithms. Alright. Cool library, I did a bunch of talks on that.</p><p>But then at some point, I taught myself how to make Python packages. And then I was a consultant, and I started noticing that at different clients that I would be writing the same, scikit-learn components. So I figured, I have to have a library with these components that I keep reusing. And that’s how scikit-lego came to be, and that’s how I familiarized myself with the scikit-learn ecosystem.</p><p>And then, I started working at Rasa. And there, we do lots of benchmarks on sentence classification because Rasa builds chatbots. And when you’re building chatbots, sentence comes in and we need to figure out the intent. Okay. So I wrote a bunch of benchmarking tools because that’s what I needed and some of those can be open sourced.</p><p>Whatlies was an example of that because I wanted to have a library where very quickly I could have many non-English embeddings and see if they were better. And then it turned out that there’s a whole non-English community around Rasa who was super interested in that.</p><p>So I was able to build some Rasa plugins to support all these non-English tools. And, then at some point, I started maintaining my own libraries, and then I noticed that I need some unit tests for my docs because I don’t want my docs to break. So I made a couple of tools to help me do that. <a href="https://github.com/koaning/mktestdocs">Mktestdocs</a> — That’s one of these tools.</p><p>I noticed the tests at Rasa were running super slow so I made <a href="https://github.com/koaning/pytest-duration-insights">pytest duration insights</a> so I could figure out which tests were slowest. And you can see how all these things accumulate, but it’s always because I’m scratching another itch. And my preferred way of operating is to do that in public.</p><p>And of course, there are tools that I can’t do in public. I work at a company. Some tools are private. That’s fine. But most of the time, I’ve encountered a problem, and I just want to be able to solve it again later with very low effort. And because I’ve made packages before, it’s just super easy to repeat.</p><p>And that’s also how <a href="https://koaning.github.io/doubtlab/">doubtlab</a> happened, and it’s also how <a href="https://github.com/koaning/embetter">embetter</a> happened, and honestly, also how bulk happened. It’s just that at some point, I figured I need this for my work. It’s nice to have around, so let’s just package it and go build in public, and that works very well for me. That’s the main story there.</p><p><strong>Seth</strong>: Yeah. Very cool. And that’s a great story. It seems like building one tool, you build up certain skills, and then one thing kind of leads to another, and then it’s not such a big deal.</p><p>Once, I guess, you have around three dozen amazing tools to add that thirty seventh tool.</p><p><strong>Vincent</strong>: So, yes, but I do want to make one comment because I do think that in general, if I look at the companies that I visited, with my background as a consultant, I do think not enough people make their own Python packages.</p><p>For example, imagine that you have a pandas query that has to deal with time series or something that is working on this very specific database. Okay. Then the function that reads out the data from the database probably can be a function that needs to be reused. And maybe you have to add sessions or maybe you have a very specific machine learning model that you want to reuse.</p><p>And for all of these utilities, you don’t want them to live in a notebook. You want them to live in a Python package. And I have seen that not enough people make their own internal tools which I do think is a shame. I was around a couple of mature colleagues at the time and we would write our own Python tools internally.</p><blockquote>…you can make Python packages more often than you might think. So, just build one even if it’s for your own little helper functions in pandas that you like to use.</blockquote><p>And because we had that habit, it was also quite easy for me to make one that was just public. So, this is advice I might have for a more general crowd, you can make Python packages more often than you might think. So, just build one even if it’s for your own little helper functions in pandas that you like to use. That’s a totally legitimate use case.</p><h4>Bulk</h4><p><strong>Seth</strong>: Yeah. To dive into the one that I’ve used the most, bulk. Can you talk about bulk? What is the pipeline and the requirements for it? What are the mechanisms at play?</p><p><strong>Vincent</strong>: Yeah. So it might be fun to also explain how that library accidentally happened. So I had a library called human-learn. There’s a couple of really cool features, but the whole thing with human-learn is that as a human, you can now make scikit-learn models without knowing anything about machine learning. One thing you can do is turn a Python function into a scikit-learn compatible component which is useful. So you can grid search over the kwargs and all that.</p><p>However, one thing I thought was kind of cool too is, usually you see a plot, of some blue dots there, some yellow dots there, some red ones there. And people say, this is what we need machine learning for and then an algorithm dissects them. But then I figured, you know, you can just draw a circle around the green dots and a circle around the blue ones and just translate that circle into a scikit-learn model. So that’s a feature of human-learn. In human-learn, we have bokeh components that can do that from a notebook.</p><p>And while I was working on that, I was also working on <a href="https://koaning.github.io/whatlies/">whatlies</a> over at Rasa for all of these word embeddings. Then at some point, it started dawning on me that when you take these word embeddings and when you pass them through <a href="https://umap-learn.readthedocs.io/en/latest/">UMAP</a>, you kind of get these clusters. And then I figured, oh, I just want to select them. Oh, hang on. I’ve got this tool called human-learn that just does that.</p><p>And within, like an hour, I had that working in a notebook. Then, I showed it to a bunch of colleagues and they all kind of went, “This is super useful, Vinny. Well done.” So that was a notebook that got shared around a lot.</p><p>And now, I no longer work at Rasa, and I started working at this company called Explosion. We have an annotation tool. And I felt like doing the bulk trick again, but I didn’t feel like using that in a notebook. So I just turned that into a little web app that you can run locally and it’s one of the pre-processing steps I like to use as the thing you do before you start annotating in Prodigy. You just take your data, you embed it into a 2-D plot using UMAP and then you typically see clusters and you try to explore that space, make a selection, and that’s it.</p><p>It’s a very nice way to do bulk labeling because clusters tend to appear from these embeddings. And that’s basically the whole trick. These bulk labeling techniques, they kind of work, but they’re not perfect. They seem pragmatic enough for me to go ahead and get started within an hour. And that’s kind of the power of it. Stuff that used to take me six hours and now it takes me only one hour.</p><p>And it’s a trick that only works for getting started, but I get started a lot on a lot of new data sets. So for me, it totally solves a problem. Bulk is also one of these projects where I would love to have more time to fix some of the rough edges, but it is a little hack that totally works and I love using it. And there seems to be a little crowd of people who seem very appreciative of that tool as well, especially because it does text but also images. Out of the box, it just does that.</p><p><strong>Seth</strong>: Very cool. Yeah. I used bulk when it was in a notebook. I know I reached out to you. You were very generous with your time trying to help me getting running in different environments.</p><p><strong>Vincent</strong>: Yeah, the first notebook was definitely buggy. That’s definitely true. Yeah. I definitely remember that.</p><p><strong>Seth</strong>: Still did the trick.</p><p><strong>Vincent</strong>: Yeah. Well, the thing also back at Rasa, I would make a habit of making these videos. So, bulk that had a YouTube video attached as well, which is how a lot of people found out about it. And I think there’s this one repository that happens to have that notebook, which is still getting stars these days.</p><p>But I recommend people just use the command line thing now because less distraction and a bit more stable.</p><p><strong>Seth</strong>: Yeah. And then interestingly for me, as I was moving a lot of my work outside of notebooks and into scripts. I came across bulk again and now I’m using more of the web app. I love both of them. They’re great tools and you make a good point.</p><p>Sometimes lowering the barrier to get started on a problem is just so important because then you start to get the ball rolling, you start to get some thoughts going, and you can make some meaningful progress. What I like about it is you start building some intuition, by exploring the data and you start to think, “Oh, okay. These could be some potential categories.”</p><p><strong>Vincent</strong>: There’s definitely a human in the loop who is learning aspect of it that I also think is really useful. Especially when they dump a new dataset on you. Yeah, you can start throwing it into an algorithm and that’s [fine]. But genuinely, understanding what’s in the dataset typically is the thing that takes the most time. And it’s nice that as a side effect of bulk, you are at least exposing yourself to these clusters. And that on its own seems quite useful.</p><p>Right now, you can do bulk labeling on sentences and images. One of the things I am working on is doing that for phrases as well, for substrings in text. So right now I can embed the entire sentence, but what I want to move towards is that I’m also able to say, take every noun phrase in that sentence and make a small little point for that. Because that way, if you’re interested in doing named entity recognition or something like that, we can also do bulk labeling for you.</p><p>And, especially things like video games that might be abbreviations — Star Wars are two tokens. It’d be nice if we can turn it into a single phrase. And over at our company Explosion, we have lots of tricks that totally solve all of this. It’s just that I need to have an afternoon to make that work inside of bulk.</p><p>But it’s stuff that is on the road map that I am definitely interested in solving some of those problems as well.</p><h4>Understanding the Problem</h4><p><strong>Seth</strong>: Yeah. So I’ve noticed going through some of your work. A lot of it is focused on creating high quality datasets. But something before that is actually understanding the problem. And I watched one of your PyData talks basically about rephrasing the problem.</p><p>And you gave an incredible example about a problem where someone is looking for beans, beef, and bread.</p><p><strong>Vincent</strong>: Oh, yeah.</p><p><strong>Seth</strong>: Can you can you talk about that one?</p><p><strong>Vincent</strong>: So this was not my tale. I actually met the person who works at the World Food Program doing Operations Research. And one of the problems that they had was [dealing with] hunger in the world. And, sometimes a village with hunger says, we need more beans or we need more chicken or there’s demand for certain products. And then part of what the World Food Organization tries to do is to source those foodstuffs cheaply.</p><p>And then part of the cost picture here is the logistics of it. So, can we get the food on the truck? And how expensive is it to get the truck? And all the logistics. And as this person was saying, they defined the problem the wrong way originally because when a person says, I need beans, yes, they can say that, but it’s not beans that they need, it’s nutrients. And beans, they’re high in fiber and high in protein.</p><p>Okay. There’s other food like lentils that is also high in fiber and high in protein. And if we’re fighting hunger, then we’re not going to be very picky about whether or not we get beans or lentils. And maybe if we do that, we can get the foodstuff without needing a shipyard. We can just send the truck.</p><p>And just by redefining that problem, I believe they got like a five percent cost reduction, which is a crazy high number for an operation — for a problem that people have already spent years on trying to optimize. Getting a five percent cost reduction is almost unheard of, but it was basically because they were solving the wrong problem. And my theory is at least that, like, this is an anecdote of a thing that happened to this one person for the World Food program. Quite typically, this whole act of rephrasing is a very useful exercise and maybe not enough of us do.</p><p>An example in NLP, one of the problems we’ll sometimes see on our support forum is, let’s say, they have a resume that they want to parse. And then they say, well, I want to have the start date and the end date per job. So I want to have an algorithm that can detect the start date. And, you know, you can build an algorithm that can detect the start date, that’s fine. But if you rephrase the problem into, let’s first find all the dates, and then afterwards figure out which one’s a start date and the end date, then the second problem becomes, well, the start date is probably first and the end date is probably after that. Oh, the whole problem just becomes a whole lot simpler if you just rephrase the problem into a two step approach instead of considering it end to end.</p><p>And there’s lots of these opportunities that people forget about. And I, again, to come back to calmcode, I fear that partially some of the machine learning textbooks are to blame because very few machine learning books actually tell you that you can choose to ignore half the data if it makes more sense. You can choose to just solve a different problem if that’s easier to solve. But that’s not the mode of thinking I seem to see, especially with new graduates. Which is bit of a shame.</p><p>But with that World Food Program story, I have to trust the person on stage who told it to me, but that definitely happens. Like, the World Food Program found a way to reduce the cost of transportation by five percent just by rephrasing a mathematical problem. And definitely something that happens in real life.</p><blockquote>“It wasn’t the algorithm that saved the day, rather the understanding of the world. A better algorithm would yield a worse outcome if it is used on the wrong problem.”</blockquote><p><strong>Seth</strong>: Right, yeah. And doing something at that scale any sort of reduction, a five percent reduction is massive. My favorite quote from that presentation you said, “It wasn’t the algorithm that saved the day, rather an understanding of the world. A better algorithm would yield a worse outcome if it is used on the wrong problem.” I really liked that one.</p><p><strong>Vincent</strong>: Oh, happy to hear it. So, there’s more anecdotes in that story. But if people are interested in this, there is an operations researcher, [Russell] Akhoff.</p><p>And he wrote this one paper about, the title was <a href="https://www.tandfonline.com/doi/abs/10.1057/jors.1979.22">The Future of Operations Research is Past,</a> which he wrote in like the eighties. It basically outlines why operations research algorithms can fail. And it’s reasons related to this anecdote. The reason why I want to bring this up is because some of those arguments work for data science too. It’s an article from the eighties, but everyone should read it: The Future of Operations Research is Past.</p><p>And I wrote a similar article called <a href="https://koaning.io/posts/the-future-is-past/">The Future of Data Sciences is Past</a> just by repeating a couple of those arguments. But people often forget that the algorithm — It’s usually just a cog in the system, and we’re interested in building a better system, not a better cog. So if you’re building a better cog but doesn’t fit the rest. It’s not a better cog because you don’t get a better system.</p><blockquote>But people often forget that the algorithm — It’s usually just a cog in the system, and we’re interested in building a better system, not a better cog. So if you’re building a better cog but doesn’t fit the rest. It’s not a better cog because you don’t get a better system.</blockquote><p>Another thing that Akhoff does very well in his books, he basically explains a lot of these systems theories. And one one quote there that I can recommend people think more about is maybe instead of making, let’s say, a better cog. Instead of thinking, “Hey, maybe there’s like one part of the system that we can optimize.” Maybe instead try to see if you can make communication between two parts better. Because if you think about it from a systems perspective, by doing that, you’re optimizing two things.</p><p>And also, you’re gaining clarity, so that’s always good. And it’s definitely this sort of let’s think about a problem by reducing it down to a single number and not consider anything else. That’s usually like a rabbit hole where people lose themselves in as well in data science, I think.</p><p><strong>Seth</strong>: Yeah. It’s super interesting because I think that there are are a lot of times when people approach problems sometimes they focus on sort of the different modules and they have this modular way of thinking about things and they go, oh, if I make this one thing the best that it could be, then the whole system will be better. And in some cases, it will make a great improvement. But other times, it’s very important to understand the supporting system and how it integrates. Reminds me that you have to have good integration tests and you need to make sure that everything fits into the system properly.</p><p><strong>Vincent</strong>: To give an anecdote here, the former CEO of bol.com wrote this in his autobiography. So, bol.com is like the Dutch Amazon. Amazon’s not that big here. Bol.com is basically Amazon, but blue, and Dutch — It’s kind of a thing we have here.</p><p>But they hired their first data scientist at some point. And this book has a chapter on that — What happened when we got our first data scientist? And in the book, the first data scientist is portrayed as kind of an arrogant kind of a person. Who’s always complaining that all these humans are not as good as my algorithm.</p><p>And then one of the things that he does is he figures out that there’s an optimal time to tweet about new video games that come out on their social channels and etcetera. So that’s like a thing he did. In Holland, we have this thing called Remembrance Day. And I believe it’s seven o’clock could be eight, but during Remembrance Day, we remember the Second World War. And basically, the entire country goes for two minutes of silence.</p><p>You might have seen some of the photos where people on their bikes delivering pizzas would step off the bike, just stand still for two minutes. It’s a thing that people take quite serious. So seven o’clock on Remembrance Day will be a very bad time to tweet about the new Call of Duty shooting game where you can shoot a bunch of people. And it will be especially bad if you would tweet that you’re super excited about the prospect of shooting people during Remembrance Day. But that’s exactly what happened because his algorithm determined that seven o’clock was the optimal time to start tweeting about this sort of thing.</p><p>And there’s so many of these stories. Right? And, on their own, on paper, I cannot necessarily blame the data scientist for doing his or her work. But this is the systems thing. Group one has concern that something might go wrong, group two does not.</p><p>If you just get them talking to each other, then usually the world’s a better place. That’s the theme, I would say.</p><p><strong>Seth: </strong>When you get the answer to your problem and you have to ask yourself, does this make sense? That’s sometimes a little step that a lot of people skip over, and it’s extremely important.</p><p><strong>Vincent</strong>: I do want to acknowledge that it’s also hard, right? I think calling .fit and .predict are the easy bits.</p><p>It’s all the stuff around that. It’s way trickier. Especially when you consider themes of fairness, all the things that can go wrong, can we really know that upfront? I don’t know if you always can.</p><p>To give one shout out though? There is this project. It’s called <a href="https://deon.drivendata.org/">deon</a>, the <a href="https://deon.drivendata.org/">Deon checklist</a>. There is a <a href="https://calmcode.io/deon/introduction.html">calmcode course</a>.</p><p>Deon is a data science checklist. So just a bunch of stuff that has gone wrong at different companies where there’s newspaper articles, like explaining how bad the situation became. They just have a checklist of stuff that’s like “hey, check for this before you push live because stuff might go wrong.” And for every item on that checklist, they also have two newspaper articles of stuff that happened in the past. So you as the data scientists can go up to your boss and say, “I want to [minimize] risk because this actually went wrong.”</p><p>And it’s a really cool project just because they actually did the proper collecting of anecdotes, which is a powerful act in this day and age.</p><h4>Unanswered Questions in Machine Learning</h4><p><strong>Seth</strong>: Yeah. A hundred percent. Having a story connected to anything in data science is always valuable.</p><p>To zoom out and talk about machine learning in general, what’s an important question that you believe remains unanswered in machine learning?</p><p><strong>Vincent</strong>: Okay. So I was drinking at a PyData Afterparty. And a couple of people came up to me, and these were people I would consider relatively senior. They knew their stuff and they asked me to predict the future of machine learning.</p><p>And I kind of felt like making a joke because, you know, you’re at the bar. I wasn’t really inclined to go super into that. As a joke, I figured I would say, “you know what I think of the future of data science, people are going to really realize just the sheer amount of nonsense that’s in our field. And we should maybe just stop altogether.”</p><p>But I decided to think about it more and I will say, there is some truth in that actually. I do kind of worry that maybe a lot of the stuff that we’re doing is more the hype thing instead of, are we sure that we understand the problem?</p><p>So what’s missing in machine learning? Well, maybe we’re doing too much of it. This is kind of a feeling that I have.</p><p>And of course, there’s a place in machine learning in the future. It’s definitely going to happen, but it doesn’t have to be everything. That’s kind of more the thing that I’m afraid of.</p><p>There’s an author who writes a book about artificial weirdness. Just like all the weird gunk that artificial intelligence can produce.</p><p>And the book is called, <a href="https://amzn.to/3pd3t3w">You Look Like a Thing and I Love You</a> by Janelle Shane. Have a read. The book starts by saying I have all of these Tinder texts, and I want to have an algorithm figure out the best Tinder text to send. And the algorithm came with, “you look like a thing and I love you.” Which is kind of a hilariously brilliant thing, but it’s not the thing you should send, I think, on Tinder.</p><p>But the book is full of these examples where you kind of have to be careful that artificial stupidity is not happening at the same time as well. Right? There’s plenty of examples where that happens. Like, the Call of Duty thing is just one example.</p><p>I find myself to be kind of the grumpy old guy who sort of yells at clouds. Kind of a, “sure, machine learning has a place, but can we do without it first?” First try the simple thing because that’s something that people seem to forget to do. And that’s a more pressing concern, personally.</p><p><strong>Seth</strong>: In a similar vein with everything that’s going on in natural language processing right now with generative models and ChatGPT. How do you view the gap between the hype and the reality? I’m excited to get the old grumpy guys perspective on this.</p><p><strong>Vincent</strong>: So I am actually professionally toying with this stuff. If you have a look, the Explosion repository now <a href="https://github.com/explosion/prodigy-openai-recipes">openai/prodigy recipes</a>. That’s the name of the repository. So we are experimenting a little bit with, like, hey, can ChatGPT just say, here’s a sentence, detect all the dates.</p><p>Just so we can pre-highlight that in our prodigy interface. It’s something we are exploring right now. And it turns out it’s actually really good at some of these examples. And it’s really bad at others. We don’t fully understand why yet.</p><p>But I will acknowledge that can be quite useful. If that’s something that you can use to get better training data quicker because the annotation is just a lot easier just saying yes or no is quicker than sort of highlighting every single item in the user interface. That seems totally fine.</p><p>What I think is a bit more of a concern though, is that people sort of say, “oh, it’s magic. That’s how this works. It’s magic.” It’s not magic. This is to some extent kind of like the <a href="https://en.wikipedia.org/wiki/Markov_chain#:~:text=A%20Markov%20chain%20or%20Markov,the%20state%20of%20affairs%20now.%22">Markov Chain</a> thing where it just predicts the next word. And you can imagine that if you just give that enough text and enough compute power, you might be able to have it generate very plausible text that you might find on the Internet. Then you can ask questions, like, is it generalizing? Or is it just remembering?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*vMUhM9jrGvKJkmt52cC5Kw.jpeg" /><figcaption>Magic Floating through the air generated by <a href="https://huggingface.co/CompVis/stable-diffusion-v1-4?text=magic+floating+through+the+air">Stable Diffusion v1–4</a></figcaption></figure><p>And, those are all fair questions. But it’s not intelligence just yet. It’s not real reasoning. And I have plenty of silly examples that demonstrate that it’s not actual intelligence that is happening under the hood.</p><p>That said again, as long as there’s a human in the loop and it proves to be useful and productive, then I think it’s fine. But again, that’s when I’m wearing the lens of, hey, there are professional interests. There are, of course, harmful factors that I do think need to be taken into consideration as well. You can definitely send more mass emails in bulk and maybe have more Twitter bots and all those things that I’m not particularly fond of.</p><p>So anyway, that’s one aspect. Another thing I do also maybe want to highlight because I also tried the Midjourney thing. I’ve tried to generate Magic: The Gathering cards.</p><p><strong>Seth</strong>: Okay. I’ve seen them and they’re pretty funny.</p><p><strong>Vincent</strong>: I thought at some point it would be kind of funny to say, hey, let’s make Magic: The Gathering cards of orcs in the office. You would have an orc warlord product manager and an orc venture capitalist and an orc TED keynote speaker. And immediately, this idea is pretty funny because if you think about the office, you kind of think of like a dull gray suit. And if you think of an orc, you think of World of Warcraft and like a warmonger, etcetera. So, that was pretty funny.</p><p>But then the next question is, can we actually generate the really funny pictures? And that turned out to be somewhat hard. So I have this one picture of an orc paladin, like, totally covered in iron basically, mesh like, behind the computer. And you kind of go, okay, data engineer, kind of okay. That’s kind of funny already.</p><p>But I wanted this orc to be a data analytics engineer because they are talking about data lakes. And then I thought the funny thing would be heavy ironclad orc but with a little yellow snorkel comes out of the helmet. That would just be the funniest thing. And I for the life of me, I could not get it to generate a yellow snorkel.</p><p>And you start thinking about, why might that be? And then you also think, like, well, Vincent, you’re already kind of stretching it to have these World of Warcraft Dungeon and Dragon styles in an office. The fact that those two styles are even compatible is already kind of a stretch. Let alone that you also generate some sort of weird snorkel from it. Right?</p><p>So if people consider these tools like magic, the best advice that I do have is try to come up with a kind of an awkward weird task that tries to touch where the edges are of where such algorithms are comfortable, and that’s usually going to give you examples that can maybe help you consider that it’s not really magic what’s happening. There’s just this; It’s trying to remember. It’s trying to sort of generate stuff that it’s seen before. And there’s plenty of edge case examples where this sort of stuff is just “<a href="https://amzn.to/3WT5wGC">You look like a thing and I love you</a>.” Read that book. It has really compelling examples and the style of the book is lovely too. I highly recommend it.</p><h4>Generative and Predictive Machine Learning</h4><p><strong>Seth</strong>: Thank you. Yeah. I will check it out. I think generative models are super interesting because unlike predictive models where for example you’re doing text categorization, where you can sort of know if it’s correct or not? There usually is a ground truth. With generative models where you’re doing something like you want to create an orc that’s wearing a snorkel, you know, how do you know that it’s correct?</p><p>It’s not so clear cut.</p><p><strong>Vincent</strong>: How many labels of unrelated photos do you need to actually generate that? Right? Oh, but also here is why also part of the solution here is obviously, user interface as well. There are amazing things you can do having text as an input. But in this case, you are also okay. We’re almost there. I just want to select the region around the helmet where a yellow snorkel needs to appear.</p><p>Something like that is going to happen at some point, and that’s going to make these systems better. And then I can move on to enterprise elves and figure out some other edge case. Right? And that will kind of be a continuous thing.</p><p>But, yeah, in general, because you mentioned ground truth — ground truth is tricky too. And this is also where a lot of artificial stupidity kind of comes from. And my personal gripe with that — so consider, image classification, the famous cat dog thing: Is this a picture of a dog or is this a picture of a cat?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/262/1*GFn7CoQ5iZbejCDTTicCyQ.jpeg" /><figcaption>Source: Catdog Logo <a href="https://en.wikipedia.org/wiki/File:CatDog.jpeg">Wikipedia</a></figcaption></figure><p>Standard classification would say, okay, this is a binary task. But then you kind of go, well, we can have photos of no cats or dogs. So, we need three classes? Okay. What do you do with photos that have both a cat and a dog? Oh, yeah. Okay. That can happen too, right.</p><p>Okay. Reality is more complex. And what do we do then?</p><p>Well, maybe we have to say, is there a dog in the photo? Yes or no. And is there a cat in the photo? Yes or no. Maybe those should just be two binary classifiers. Maybe that’d be more sensible. Okay. What do you do when there’s four dogs in the photo?</p><p>Again, the more and more you start thinking about it, you also kind of realize, even the well defined text classification doesn’t always mix well with reality either. And even if you have ground truth labels, you kind of have to wonder, well, the ground truth labels maybe don’t mix with reality either it it’s defined as a classification task because a sentence can be about more than one topic and the photo can be about more than one thing as well.</p><p>So taking a step back and just really wondering, well, some of these things can be details as long as we really understand the problem, but maybe we should focus on that then. Maybe we should skip hyperparameter tuning and only worry about — Do we really understand the problem?</p><p><strong>Seth</strong>: Yeah. That’s a really good point. I think that when you’re approaching a problem, people tend to jump to a solution. If you’re doing something like text classification — oh, okay. I’m going to create a multi-class text classifier. Well, it turns out that it is never really quite that simple. Right?</p><p>It’s really multi-label. Should I use a hierarchy? Should I do this? Should I do that? And, you know, getting a better understanding of the problem always helps you figure out more. It’s so much more valuable than doing hyper parameter tuning on that original multi-class text classifier.</p><blockquote>“The model can do one step, but your system can do two or three if need be. So definitely feel free to consider the two-step system where we have a couple of classifiers that detect a couple of properties, and then we have a rule-based system after that’s going to say, ‘Okay, this combination of things that seems interesting. Let’s go for that.’ People forget about the rule-based system that can be built on top of. And that’s, you know, a bit of a miss. But it’s also, like 80% of the time, that’s also the fix.”</blockquote><p><strong>Vincent</strong>: Well, so the main thing, I do have a little bit of advice in general. I’m on the Prodigy forum and I help some SpaCy users with their problems. The most general advice that I do give people in this domain is to consider that maybe the model can do one step, but your system can do two or three if need be. So definitely feel free to consider the two step system where you have a couple of classifiers that detect a couple of properties, and then we have a rule based system after that’s going to say, “oh, okay. This combination of things seems interesting. Let’s go for that.”</p><p>People forget about the rule-based system that can be built on top of. And that’s a bit of a miss. But it’s also, like eighty percent of the time, that’s also the fix. So do with this information, what you will, dear audience, but I do think that there’s a two step approach that definitely does work in general.</p><p><strong>Seth</strong>: I think that’s really good advice especially right now with all of the hype with deep learning. I think we’re still in a world where finding the right combination between machine learning models and heuristics, sometimes pretty basic heuristics, often yields the best results.</p><h4>Influences</h4><p><strong>Seth:</strong> To move into the learning from machine learning portion of our talk. We’ll start with this. Who are some people in the machine learning fields that influence you?</p><p><strong>Vincent</strong>: I’ve had some really lovely direct colleagues that I still hang out with. So, those obviously. Back when I started, I was learning R, so <a href="https://hadley.nz/">Hadley Wickham</a> was a person that I definitely looked up to a lot. And I also met him on a couple of occasions, which is super cool. He did an advanced course, like, five years ago, and I was a TA. Great great experience, I got to meet the guy.</p><p><a href="https://www.linkedin.com/in/katharinejarmul/?originalSubdomain=de">Katharine Jarmul</a> is a person who also comes to mind. She was one of the kickstarters behind <a href="https://pyladies.com/">PyLadies</a>, but she also has been a great advocate for privacy and fairness in machine learning. And she has reviewed my slides in the past for a couple of talks and she’s just great. She comes to mind.</p><p><a href="https://vickiboykis.com/">Vicki Boykis</a>, I think, is one of the funniest people — she deserves way more credit for shit posting, she’s great. The NormConf thing was also an amazing thing that she helped kick start there, it was great.</p><p>And then <a href="http://worrydream.com/">Bret Victor</a>, I think, has the best talk I’ve ever seen, that I will ever see. <a href="https://www.youtube.com/watch?v=8pTEmbeENF4&amp;ab_channel=JoeyReid">The future of programming</a> by Brett Victor. That’s a thing I watch every year, basically. That’s the most gobsmack, most inspirational thing I’ve ever seen. I won’t tell you what the thing is about. Just watch it.</p><p><strong>Seth</strong>: I’m looking forward to it.</p><p><strong>Vincent</strong>: And then, I guess, [Russel] Ackoff, but the main thing with Ackoff was I did this whole master’s degree in Operations research and then a professor was going to retire and I was one of the speakers at his party. And then at some point, he said the reason I wanted you here is because you really remind me of Ackoff. I was like “who is he?”</p><p>“He’s this amazing guy. Just buy his book.” And then you read this stuff. He’s like me in the 80s. So that was definitely also a good source of inspiration.</p><blockquote>…the average Joe is pretty inspirational, but the average Joe doesn’t think that he or she should be on stage.</blockquote><p>One thing I do want to mention about this is back when I was organizing a PyData. You kind of think, “okay, who are good keynote speakers and who are good, invited speakers, etcetera.” And my impression is that the average Joe is pretty inspirational, but the average Joe doesn’t think that he or she should be on stage.</p><p>And the best example of this is, at PyData London, there was a normal talk by a guy who was building drones to find endangered species of orangutan in the rainforest of Borneo.</p><p><strong>Seth</strong>: Wow!</p><p><strong>Vincent</strong>: And he had the small room, but his talk was amazing. So I figured, screw this. You’re the keynote at Amsterdam. This is the most amazing thing I’ve ever heard. This is your hobby.</p><p>So he was the keynote speaker the next year. And he was grateful and very good fun. But, he didn’t realize that that was definite keynote material. And similarly, I’ve read this blog post once where this guy was trying to figure out which words are the most metal.</p><p>And the way he did that was by training a huge Markov Chain on metal lyrics and non-metal lyrics. And the conclusion of the blog post was that the least metal word is cooperation because it only appears in the corpus once. And you read this, this is amazing. Because you’re basically applying the theory correctly on a pretty humorously silly problem, maybe.</p><p>But there’s passion here. And the guy, when I did approach him, [I said] you really need to apply for PyData. I don’t have to review your thing, I think you’re going to be in.</p><p>And it just hadn’t occurred to him that this was something he could do. And I like to think that there are so many more people who suffer from this, that they might have a really grand amazing inspirational moment, but don’t consider that they’re able to share that. And of course, some people are, properly introverts, which is also just fine. But one lesson I have learned at PyData is that the inspiration can really come from surprising angles that you don’t expect. So don’t focus too much on the big names.</p><p>That’s also the thing.</p><p><strong>Seth</strong>: Yeah. One of the best types of people are very humble and they do such a good job with their work and you can tell how much they care about what they do and how much, I don’t know if pride’s the right word, but they take their work very seriously. They care..</p><p><strong>Vincent</strong>: They care, Yeah. You can be the smartest person, but if you don’t care about your topic it’s not going to be a great talk.</p><p>And let’s say that maybe you’ve cut a few corners, but you calculated the optimal Pokemon. I don’t know, something like that. It can still be a great talk.</p><p>And again, more people should do it. If people are interested in doing more blogs and talks by the way, consider Lightning Talks and very short blog posts are called “Today I Learned”. The world definitely needs more of that. And I’m happy to see PyAmsterdam, the meetup, once a year, they do the lightning talk meetup, where ten people give five minute presentations.</p><p>Those meetups tend to be amazing. Any PyData organizers listening, feel free to steal this idea. Those meetups are always fun.</p><h4>Career Advice</h4><p><strong>Seth</strong>: Very cool. So you’ve you’ve given a lot of advice so far, but I want to ask, what’s one piece of advice or something that’s stuck out that you’ve received that’s helped you in your machine learning or career journey.</p><p><strong>Vincent</strong>: I got this very early on in my career. My former CTO that I still hang out with gave me pretty good career advice when I was twenty three. And he said, “Be careful of getting a raise. Because if your job starts earning a lot of money, but it’s kind of getting boring, then the money might be a reason that you’re going to stick around.”</p><p><a href="https://youtube.com/shorts/i1QEuc8ioDs?si=BWrpO4FVfYUI_r6q">Be careful of getting a raise early in your career</a></p><p>And that’s a dangerous thing early in your career because maybe you have to figure out what you like in life. And maybe you have to figure out what makes you tick. And if you’re going to hyper focus on the money, it’s kind of like hyper focusing on the metric. You’re going to over optimize for something that might not matter as much. So that was pretty cool advice, kind of on the meta side. But I do think in general, I have been able to apply that quite well.</p><p>Again, privileged speaking here. Right? But, I have been able to apply that. So that’s been cool.</p><p>Kind of a weird anecdote, but it’s surprisingly inspirational as well. So I have a lot of friends who do nothing in data science. And I love that. I’m nerdy Vincent, and when I drink beer with them, they say, stop being nerdy Vincent, you’re amongst normal people, you can just talk about life now.</p><p>And I live in a neighborhood where you know all of your neighbors, basically. And it’s sort of still kind of a middle class neighborhood — it’s changing because of gentrification, but, we all know each other. So there’s a guy on my street. And he’s a painter. And when it’s nice weather, he puts a crate of beer on his bench outside of his house.</p><p>And the whole street just goes for a pint, basically. It’s the cutest thing. Cutest neighborhood ever. But the thing with him is he recently became an independent contractor as a painter, which also meant that he bought his first laptop ever. And he’s forty two.</p><p>And, he needs help, not just with his website, but, like, getting word started. His entire life, the main computer that he had was his phone, and he has been fine, but he finds a computer terribly, terribly complex. And to be honest, I find that just such a refreshing thing. And to also be reminded of the fact that the way that I experience computers doesn’t necessarily have to be normal. That’s a very useful reminder. So, it’s the best inspiration in a sense.</p><p>Maybe don’t be in machine learning all the time. It is my advice. Especially if you’re making machine learning for apps that the average person uses. It really helps to remember that they really don’t care about your algorithm. They just don’t. They really, really don’t.</p><p>I have found myself to be stuck in a machine learning bubble at times. And I just find it very refreshing to [step outside].</p><p>I used to do this at consultancy gigs as well. I was making an app that the truckers would have to use for logistics and stuff. And at some point, I would just hang out at the smoker’s corner where all the truckers would hang to sort of understand what kind of people they were. And also just to understand what they found frustrating about the app and doing more of that really. Being more of a human in the loop.</p><p>Focus on the human thing is what I’m trying to do more of and what I find very inspirational.</p><h4>Advice for New Data Scientists</h4><p><strong>Seth</strong>: Yeah. I really like that. For somebody who is just starting out in the field, let’s say, that they just got hired as a junior data scientist, or they’re thinking about starting in data science, what would your advice be to them?</p><p><strong>Vincent</strong>: Okay. So step one, I gave a talk on this topic at NormConf. So there’s a talk titled <a href="https://www.youtube.com/watch?v=S7vhi6RjBZA&amp;ab_channel=NormConf">Group-by statements that save the day</a>. This talk is precisely designed for you.</p><p>Having said that I’m a really bad person to give career advice because I looking back, I just thought that a large chunk of where I am today is due to luck and that’s something that’s kind of hard to optimize for.</p><p>What I do think is useful, in general is maybe to have your own blog where you just share things that you learned today. So just like calmcode is still kind of my snippets library in a way. Your blog can be the same for you. And in general, I have found that for these “today I learned” snippets — If you’re able to write two a month, and it take maybe half an hour per post instead of a big blog post that takes hours, this thing shouldn’t take more than like tens of minutes, let’s say. But you did it for a year and you’ve got a blog with 24 posts.</p><p>If you’re learning and you’re able to share knowledge then people are going to acknowledge that you do have a bit of a resume there that demonstrates that you’re learning stuff. So that seems like a pretty easy thing to do if you want to get something of an online presence with low effort. That’s something I recommend.</p><p>I do think if you are a super junior just getting started, I do want to acknowledge it’s kind of hard. It’s a bit of a shame now the [state of the] hiring market and all that. But one thing that you can do to make it maybe slightly easier for yourself is to consider that you don’t have to know everything in order to get the job.</p><p>You might also be able to get a related job. Some advice that I have given to friends of mine who wanted to get into this data science field is it’s a little bit easier maybe to learn R than it is to learn Python, and it’s maybe a little bit easier to just be an analyst for a year or two.</p><p>And all the skills you learn while being an analyst are going to be super useful if you want to become a data science person later. So if it’s easier and you get paid to learn, <strong>don’t optimize for a title</strong>. Just optimize for the stuff that you learn while on the job. That seems easier. And there’s nothing wrong with being a good analyst.</p><p>Maybe we need more good analysts than we do good data scientists as well. Right? Maybe we need more group by statements that saved the day. Hint, watch the talk.</p><p>But I do think there’s a little bit of snobbery when it comes to job titles. “Like, oh, I’m the super senior staff, mega engineer. Like, sure.” But if you’re just a really decent analyst, we need more very decent analysts. That’s also fine. Go for that.</p><h4>Learning from Machine Learning</h4><p><strong>Seth</strong>: Yeah. That’s definitely good advice. And now the question that we’ve all been waiting for, what has a career in machine learning taught you about life?</p><p><strong>Vincent</strong>: Some problems solve themselves when you ignore them. Seriously, I’ve been in so many of these situations where the problem got solved by just ignoring the machine learning bit, that you kind of start to wonder, well, maybe some problems do solve themselves if you ignore them.</p><p>And I have noticed in a few instances, this is just kind of the case. Especially when you have a child, you do kind of learn that there’s some stuff that you can over optimize for as well. And, like, oh, the baby’s not sleeping well. Well, that problem will sort itself out at some point. It’s not like influence from my end is going to make a very significant impact there.</p><p>And I guess the same thing with machine learning. <strong>There’s some stuff that you can control, some stuff that you can’t. Just make sure that you understand what you can and can not control and then move on from there.</strong></p><p>And again, I’m kind of a mixed bag when it comes to the whole machine learning thing. Part of my opinion is it’s a super useful tool and we need more good people doing machine learning. But at the same time, it’s like a gross bucket of hype that we really want to have less of. And my day to day is to sort of deal with both of these feelings.</p><p>I hope this answered the question in some way, but that that’s kind of where I’m at. Try to do it calmly, that’s my final pun. That’s also something I might recommend.</p><p><strong>Seth</strong>: There you go. Yeah. I think that some problems over time do resolve themselves. And also I like the first rule of machine learning is do you really need to use machine learning?</p><p><strong>Vincent</strong>: Yeah. I agree. And one thing I really do maybe to brag about the employer a bit.</p><p>One thing I really like about SpaCy is you don’t have to use the machine learning bits. You can also just use the non-machine learning bits in SpaCy. And they are also performant, fast and super useful. There are also machine learning packages that allow you to do some rule-based stuff. And if you’re doing NLP, this is really why I love using SpaCy.</p><p>You don’t have to use statistical stuff all the time. The rule based engines are great too. End of pitch.</p><p><strong>Seth</strong>: I’ve been a huge fan of SpaCy for a while now — at least four years, probably more. It’s helped me solve lots of problems from named entity recognition, text classification, cool ways of doing matching, all of that.</p><p><strong>Vincent</strong>: Well, so if I can give one final pitch. So there’s a lot of talk about data-centric AI these days. But the reason why I started getting interested in what these Explosion people were doing back in the day is there’s a blog post from 2017 called <a href="https://explosion.ai/blog/supervised-learning-data-collection">Supervised Learning is great — It’s data collection that’s broken</a>. They were doing data-centric stuff in 2017, but that’s one of the best blog posts I’ve ever read.</p><p>So they talk about data quality and one of the best quotes ever is don’t expect great data if you’re boring the shit out of underpaid people because mechanical turk is still like the way people go sometimes, read that blog post. I will give you a link for the show notes. That’s also like a highly inspirational thing people should read.</p><p><strong>Seth</strong>: Awesome. Thank you so much. It has been such a pleasure to talk with you. You’ve given me tons of great resources. Putting together the show notes for this one is definitely going to be a good time. If there are some places that you would want listeners to learn more about you, what would those places be?</p><p><strong>Vincent</strong>: So I’m on <a href="https://twitter.com/fishnets88?s=20">Twitter</a> and Fosstodon these days. But the main thing is I can’t announce anything just yet. It’s that I work at Explosion and I can see the stuff that’s in the pipeline. So I’m working on very cool stuff and there’s definitely going to be announcements of super cool stuff all my other colleagues are working on — just follow Explosion.</p><p>There’s a bunch of really cool stuff in the pipeline. And if you do that, then you also at some point, will hear about some of the stuff that I’m working on.</p><p><strong>Seth</strong>: Awesome. Thank you so much, Vincent. It has truly been a pleasure.</p><p><strong>Vincent:</strong> Likewise.</p><h4>Listen Now</h4><p>The <a href="https://youtu.be/yvgxRzqx1Jg">video</a> of the full interview can be seen here:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FyvgxRzqx1Jg%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DyvgxRzqx1Jg&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FyvgxRzqx1Jg%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/4c5fd936cb28a79479a9b77719b04b04/href">https://medium.com/media/4c5fd936cb28a79479a9b77719b04b04/href</a></iframe><p>The podcast is now available on all <a href="https://rss.com/podcasts/learning-from-machine-learning/804379/">podcast platforms</a>:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fopen.spotify.com%2Fembed%2Fepisode%2F32UlPjayBKuYIuTO6RNrvX%3Futm_source%3Doembed&amp;display_name=Spotify&amp;url=https%3A%2F%2Fopen.spotify.com%2Fepisode%2F32UlPjayBKuYIuTO6RNrvX&amp;image=https%3A%2F%2Fi.scdn.co%2Fimage%2Fab67656300005f1f87a1155f89111000b7c75a33&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=spotify" width="456" height="152" frameborder="0" scrolling="no"><a href="https://medium.com/media/004ea620c71cbd0c6d260914228f2d1f/href">https://medium.com/media/004ea620c71cbd0c6d260914228f2d1f/href</a></iframe><p>Our previous episode featured <a href="https://www.maartengrootendorst.com/">Maarten Grootendorst</a>, the creator of <a href="https://maartengr.github.io/BERTopic/">BERTopic</a> and <a href="https://maartengr.github.io/KeyBERT/">KeyBERT</a>, and a <a href="https://medium.com/@maartengrootendorst">prolific author here on Towards Data Science</a>. He discussed Open Source projects, the intersection of psychology with machine learning and software development, and the future of Natural Language Processing.</p><p><a href="https://towardsdatascience.com/learning-from-machine-learning-maarten-grootendorst-bertopic-data-science-psychology-9ed9b9b2921">Learning from Machine Learning | Maarten Grootendorst: BERTopic, Data Science, Psychology</a></p><h4>References</h4><p>Resources to learn more about Vincent Warmerdam:</p><ul><li><a href="https://calmcode.io/">calmcode</a></li><li><a href="https://koaning.io/">https://koaning.io/</a></li><li><a href="https://www.youtube.com/watch?v=kYMfE9u-lMo&amp;ab_channel=PyData">Vincent Warmerdam: The profession of solving (the wrong problem) | PyData Amsterdam 2019</a></li><li><a href="https://www.youtube.com/watch?v=S7vhi6RjBZA&amp;t=0s&amp;ab_channel=NormConf">Group-by statements that save the day — Vincent D Warmerdam</a></li><li><a href="https://www.youtube.com/redirect?event=video_description&amp;redir_token=QUFFLUhqbThEWVk0N1AzQ1diZm1QVjAyRllGbTVfRUZFd3xBQ3Jtc0tsbE1scFl4QUg2MVgwNE1PMlBXMVZObmxlcUlBN0psSWNaQkd0TUZrZ280b3JEQzhneExRMlFwOE5QSDYxZ3M3R0JhVUQxdGRTRFZGNy1US1g2X1J0RTJKdjNwT3R3V0YzU3RUVmJ1aFdBV1RBLWJUNA&amp;q=https%3A%2F%2Fgithub.com%2Fkoaning&amp;v=yvgxRzqx1Jg">https://github.com/koaning</a></li></ul><h4>References from the Episode</h4><ul><li><a href="https://amzn.to/3HPXHLH">You Look Like a Thing and I Love You: How Artificial Intelligence Works and Why It’s Making the World a Weirder Place</a></li><li><a href="https://ackoffcenter.blogs.com/files/the-future-of-operational-research-is-past.pdf">The Future of Operational Research is Past</a></li><li><a href="https://explosion.ai/blog/supervised-learning-data-collection">Supervised Learning is great — it’s data collection that’s broken</a></li><li><a href="https://deon.drivendata.org/">Deon — An ethics checklist for data scientists</a></li><li><a href="https://hadley.nz/">Hadley Wickham</a></li><li><a href="https://www.linkedin.com/in/katharinejarmul/?originalSubdomain=de">Katharine Jarmul</a></li><li><a href="https://vickiboykis.com/">Vicki Boykis</a></li><li><a href="https://www.youtube.com/watch?v=8pTEmbeENF4&amp;t=0s&amp;ab_channel=JoeyReid">Brett Victor</a></li></ul><p>Resources to learn more about Learning from Machine Learning:</p><ul><li><a href="https://www.youtube.com/channel/UCGwSWuvSRzop4ZVG2z9NmnQ">Subscribe on Youtube</a></li><li><a href="https://www.linkedin.com/company/learning-from-machine-learning/">Learning from Machine Learning on LinkedIn</a></li><li>Reach out to <a href="https://www.linkedin.com/in/sethplevine/">me on LinkedIn</a></li><li><a href="https://medium.com/@levine.seth.p">Follow me on Medium</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e07951a07cf4" width="1" height="1" alt=""><hr><p><a href="https://medium.com/data-science/learning-from-machine-learning-vincent-warmerdam-calmcode-explosion-data-science-e07951a07cf4">Learning from Machine Learning | Vincent Warmerdam: Calmcode, Explosion, Data Science</a> was originally published in <a href="https://medium.com/data-science">TDS Archive</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Learning from Machine Learning | Maarten Grootendorst: BERTopic, Data Science, Psychology]]></title>
            <link>https://medium.com/data-science/learning-from-machine-learning-maarten-grootendorst-bertopic-data-science-psychology-9ed9b9b2921?source=rss-70f6655bd458------2</link>
            <guid isPermaLink="false">https://medium.com/p/9ed9b9b2921</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[nlp]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[careers]]></category>
            <dc:creator><![CDATA[Seth Levine]]></dc:creator>
            <pubDate>Mon, 13 Feb 2023 22:50:45 GMT</pubDate>
            <atom:updated>2023-02-20T19:42:49.997Z</atom:updated>
            <content:encoded><![CDATA[<h4>The first installment of Learning from Machine Learning features an insightful interview with the mastermind behind BERTopic, Maarten Grootendorst</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PafbuqJO84a4Xp6WWDxx2g.png" /></figure><p>Welcome to “<a href="https://www.youtube.com/@learningfrommachinelearning">Learning from Machine Learning</a>,” a series of <a href="https://rss.com/podcasts/learning-from-machine-learning/">interviews</a> that explores the exciting world of machine learning, with a focus on more than just algorithms and data: career advice and life lessons from the experts.</p><p>Artificial intelligence is changing the world, and machine learning is fueling this revolution. In each <a href="https://www.youtube.com/@learningfrommachinelearning">episode</a>, leading practitioners from industry and academia share their knowledge, experiences and insights on what it takes to succeed in this rapidly-evolving field.</p><p>The <a href="https://youtu.be/bLW_RH2Y6oI">inaugural episode</a> features <a href="https://www.maartengrootendorst.com/">Maarten Grootendorst</a>, the creator of <a href="https://maartengr.github.io/BERTopic/index.html">BERTopic</a> and <a href="https://maartengr.github.io/KeyBERT/">KeyBERT</a> and author of <a href="https://medium.com/@maartengrootendorst">many articles here on Towards Data Science</a>. He shares his thoughts on Open Source projects, the role of psychology in machine learning and software development and the future of Natural Language Processing. The interview is now available on all <a href="https://rss.com/podcasts/learning-from-machine-learning/">podcasts platforms</a>.</p><h3>Takeaways</h3><p>Maarten Grootendorst is the creator of many powerful Python libraries including KeyBERT and BERTopic. He started his career in psychology and after obtaining a master’s in Clinical Psychology and Organizational psychology — transitioned to data science. He quickly made an impact on the field by developing open source libraries and writing poignant data science articles. BERTopic, his most used library, is a framework for topic modeling, which is a way to automatically identify topics in a set of documents. It can be used for exploratory data analysis and to track changes in trends over time. He also discusses the challenges he faced while creating BERTopic, as well as the goals and features of the most recent version.</p><p>Maarten discusses the implications of ChatGPT-type approaches, which have been mistaken for artificial general intelligence (AGI), and how they are accepted into industry. He shares his curiosity on how these large language models with billions of parameters fit into the AGI discussion and believes the coming year will be drastic.</p><p>Maarten’s advice includes:</p><ul><li>Focusing on understanding and knowing the basics of coding and machine learning before moving on to more complex algorithms.</li><li>Building a strong foundation, as it will allow for better and easier growth in the future.</li><li>Understanding the evaluation of your models. When it comes to topic modeling the “truth” is often in the eye of the beholder.</li><li>Emphasizing the importance of truly understanding the problem before attempting to solve it, citing his experience in cancer research where one of his colleagues spends weeks trying to comprehend the issue before coding.</li><li>You don’t need to know everything all at once. It’s important to find a balance between a healthy life and and your passion (of machine learning).</li></ul><p>Watch the <a href="https://youtu.be/bLW_RH2Y6oI">video</a> or listen on <a href="https://open.spotify.com/episode/0HYf9ZXFniiWQvSp0qOJ4s?si=cc74b52402cd4ebb">Spotify</a> or <a href="https://podcasts.apple.com/us/podcast/maarten-grootendorst-bertopic-data-science-psychology/id1663925230?i=1000593300970">Apple Podcasts</a> now!</p><h3>Table of Contents</h3><blockquote><a href="#88a7">Welcome</a><br><a href="#afc8">Takeaways</a><br><a href="#2a94">Full Interview</a><br> — <a href="#962b">Background</a><br> — <a href="#b8f2">BERTopic Deep Dive</a><br> — <a href="#1837">Learning from Machine Learning</a><br><a href="#89df">Video Interview</a><br><a href="#40c6">Spotify Audio</a><br><a href="#34d3">Resources</a><br> — <a href="#f6c1">Maarten Grootendorst</a><br> — <a href="#b586">Connect with Learning from Machine Learning </a><br><a href="#5e01">Video Contents</a></blockquote><h3>Full Interview</h3><h4>Background</h4><p><strong>Seth:</strong> Welcome. It’s my pleasure to have Maarten Grootendorst. He’s the creator of many useful Python libraries including <a href="https://maartengr.github.io/KeyBERT/">KeyBERT</a> and <a href="https://maartengr.github.io/BERTopic/index.html">BERTopic</a>. I’ve been an admirer of your work for some time and really love using your libraries. Welcome!</p><p><strong>Maarten:</strong> Great, thank you for having me. Love being here.</p><p><strong>Seth:</strong> So just to get things kicked off can you give us a little bit of background on your career journey, how you got into data science?</p><p><strong>Maarten:</strong> Yeah, of course. So I have a bit of an on orthodox background. I started out as a psychologist. I have a master’s degrees in organizational psychology and clinical psychology. Those were very interesting fields to pursue and very interesting work to be done. But I always felt like there was something missing.</p><p>Something, I wouldn’t necessarily say factual, but a little bit more hard instead of the soft science that I was doing at the time. So I started to explore more of the statistical side of things because that’s something that we do quite a lot in psychology and eventually I got into programming and machine learning.</p><p>And I figured, okay, studying in the Netherlands is relatively cheap, so let’s do a master’s in data science. And that’s where I found my passion really. It’s where I could find a true combination of psychology and still a technical aspect of it. Because much of what we do in data science in almost all cases involves some sort of human aspect.</p><p>So that’s why I really could make use of my psychological background. And then after that, I developed a few packages like KeyBERT and BERTopic, did some writing and stuff like that.</p><p><strong>Seth:</strong> Right. Yes, all those amazing useful packages. So, your background in psychology, do you want to just dive a little bit deeper into that?</p><p><strong>Maarten:</strong> Yeah, of course. I started out with a bachelor in social psychology, and that’s rather broad, right? Composites quite a lot of different subjects. And my mother is actually an organizational psychologist, so I kind of had to follow in her footsteps. No, but that was something I was familiar with and something I found to be interesting.</p><p>And mostly the organizational psychology I found to be really interesting. It’s the human behavior in the workplace and how they relate with everything that happens in such a major aspect of your life, which is work, right? But at the time I really wasn’t, well, let’s call it mature enough to really go into the field and do the stuff that was needed.</p><p>I figured, okay, let’s explore and develop myself a little bit further before I roll into my working life. And so I explored clinical psychology where I could focus a little bit more on the treating people that have, or helping people more specifically with anxiety disorders or depression or PTSD, things like that.</p><p>And then, of course, you learn quite a lot of interesting things from there, but I always felt something was missing. Not that these fields are not interesting, but you know, people like me always try to find their purpose in life in a way to find, okay, what makes me happy? Right. And although it was nice and interesting I was quite sure that it wouldn’t make me happy for the coming 30, 40 years.</p><p>So you explore and that’s when I eventually found machine learning.</p><p><strong>Seth:</strong> When you first found machine learning, what was it that really attracted you to it?</p><p><strong>Maarten:</strong> The interesting thing about that is that because I was missing something more technical that’s one of the first things of course that attracted me to machine learning because you can really, it’s a new field.</p><p>And there’s a lot of technical basis surrounding many of these algorithms, but the way you use it and the way you apply it still requires you, in many cases, to have some sort of business sense. Some sort of view of, okay, when are we going to use this? Is it really necessary? Do we need a very complex algorithm or can we use something relatively straightforward?</p><blockquote>That puzzle is really interesting because there’s so many small aspects to it, much more than just, okay I’m gonna optimize the hell out of this algorithm.</blockquote><p>And that puzzle, that’s what it still is to me. That puzzle is really interesting because there’s so many small aspects to it, much more than just, okay I’m gonna optimize the hell out of this algorithm. That really attracted me to it. And where I could find many of what I believe are my skills to be made good use of.</p><p><strong>Seth:</strong> When you were making the transition from psychology to the more technical realm. What were the things you felt you needed to learn first?</p><p><strong>Maarten:</strong> So as a psychologist, you’re not the most technical person, right? You’re focused on this interaction, human behavior and observational skills.</p><p>There’s some statistics in there, but the main, very first thing really is the basics. So, coding, of course you can know an algorithm perfectly well, but if you cannot code clearly and well, there’s bound to be bugs or issues or whatever found in there. That’s really, really important to do that well.</p><p>Instead of going into very complex learning algorithms, it’s really starting from, okay, how does a very simple regression work, which actually can be quite difficult if you really go into some of the complexities of it. So I was really focusing on making sure I understood that perfectly well, as much as possible before going into the next steps.</p><p>The best way to learn is really, really understanding and knowing the basics. And if you have the basics really well done, like it’s an intuition or something that you can do automatically, going into the next step becomes so much easier. So that has been my focus.</p><p><strong>Seth:</strong> So just sort of creating the building blocks that could be the foundation for your future work.</p><p><strong>Maarten:</strong> Yeah. Yeah. That’s nicely worded that foundation. The better the foundation the better you can build on top of it.</p><h4>BERTopic Deep Dive</h4><p><strong>Seth:</strong> Exactly. So having the creator of BERTopic I figured that we should take some time to go into it. I’ll let you give an introduction to it, if you don’t mind. Just talk about topic modeling a little bit and the power of it as well.</p><p><strong>Maarten:</strong> Yeah, sure. So BERTopic is a framework for topic modeling where you essentially have, well, let’s say 10,000 documents. And it can be anything, right? It can be reviews for a certain product, they can be tickets for a certain system. They can be multidisciplinary consultations for a hospital and, and some patients. And essentially what you want to do is you want to know: what are these documents about? And there are many things that you can do. You can read all 10,000 of them and label them, which is perfectly fine. If you have the time.</p><p><strong>Seth: </strong>Take a little bit of time.</p><p><strong>Maarten: </strong>Just a little bit or you can do it automatically.</p><p>And that’s where topic modeling of course, comes in because as the name implies essentially tries to extract topics from those sets of documents and tries to communicate them in a way that helps you understand what these documents are about. It also can help you do trend analysis, for example if you have Twitter messages to see how Covid is being talked about two years ago compared to now.</p><p>And it’s a very nice way to extract that information, but because it’s an automatic way without a ground truth, it also is very difficult to evaluate. So, it’s something that really fits with me as a psychologist because it allows for a lot of human evaluation and interpretation of the technique.</p><p>And you can use it for exploratory data analysis, just to see, okay, what do I have in my documents? But it’s also used in a lot of sociopolitical science where we look at certain information that has developed over the last few years and see how these trends have, have developed and changed for certain classes or targets or whatever.</p><p><strong>Seth:</strong> Yeah. I’ve been following BERTopic for some time and I know version 0.13 just came out recently. I’m wondering, what was the goal of the initial package? And, how has it changed over the three years or so?</p><p><strong>Maarten:</strong> It started out as a way to create a pipeline. Okay. So the pipeline of clustering and trying to extract a topic representation out of that has been around for quite a while. But I wanted to find a way to do it in a sort of pipeline where there’s very little assumptions being made between the steps that you are using.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/708/1*tZ37E5Qd8Ro8QJZKZmKATg.png" /><figcaption><a href="https://maartengr.github.io/BERTopic/algorithm/algorithm.html">Image</a> by Maarten Grootendorst</figcaption></figure><p>So what we’re essentially doing is we’re transforming our documents into numbers, into numerical values, reducing that to smaller dimensions. So instead of 300 values, you’re gonna compress it into five. We’re gonna cluster those documents and from those clusters, which we assume to be topics, and we extracted topic representations, and there have been a lot of pipelines that do that.</p><p>But the way your topic initially thought of was making it in a way where you can say, okay, I’m not really happy with this clustering algorithm and I’m gonna choose something entirely different. Or, I’m not happy with this dimensionality reduction algorithm. I’m gonna choose something different.</p><p>And then the focus was mostly on the last part, on the topic extraction way. And I’m using a modified TF-IDF measure for that called <a href="https://maartengr.github.io/BERTopic/api/ctfidf.html">cTF-IDF</a>. And because I initially thought of it doing that way it was very easy to then devise many variations and extensions to a topic often are different packages.</p><p>So in topic modeling, you have <a href="https://www.jmlr.org/papers/volume3/blei03a/blei03a.pdf">LDA</a>, which is the classical topic modeling technique. Yeah. But if you want to use some sort of variation of that, you have to install a different package. In many cases, not all cases, of course, the <a href="https://radimrehurek.com/gensim/">gensim</a> has a lot of implemented already. But if you want to do hierarchical topic modeling or dynamic topic modeling, I think it’s not implemented and you have to install different packages for that.</p><p>And what I wanted and what it’s now becoming is a one-stop shop for topic modeling and without necessarily saying, okay, per topic is the best ever topic model out there. Which it definitely isn’t. I mean, we still need to adhere to the <a href="https://en.wikipedia.org/wiki/No_free_lunch_theorem">No Free Lunch</a> theorem. But you can use it for basically everything.</p><p>And because I focused on that pipeline of having minimal assumptions per topic. And now do our topic modeling - <a href="https://maartengr.github.io/BERTopic/getting_started/online/online.html">online topic modeling</a>, <a href="https://maartengr.github.io/BERTopic/getting_started/topicsperclass/topicsperclass.html">class based</a>, <a href="https://maartengr.github.io/BERTopic/getting_started/semisupervised/semisupervised.html">semi-supervised</a>, <a href="https://maartengr.github.io/BERTopic/getting_started/supervised/supervised.html">supervised</a>, and a few others. And it’ll continue to develop in that way. So, now what it currently is more focused on is this build your own topic model type of package.</p><p>And there are a few more things coming up, but so far that there has been a trajectory for the last couple of years.</p><p><strong>Seth:</strong> Yeah. Working in the industry now for a couple of years and doing topic modeling. I can say that I enjoy using BERTopic. It’s great — Just the level of abstraction where you don’t necessarily need to know every single detail and that you can get such incredible results.</p><p>And also as you were referring to the modularity of it, where you can sort of plug and play and put in different algorithms to get different outputs. So, you touched upon it before in terms of evaluating topic models. Can you discuss why it’s so difficult to evaluate a topic model?</p><p><strong>Maarten:</strong> What generally happens is you have, let’s say 10,000 documents and they are tweets from some person or collection of people. And what BERTopic or any topic modeling technique then is doing is in an unsupervised way (without a ground truth) extracting topics from these messages, from these documents.</p><p>But who’s to say that these topics are accurate? And then what does accuracy actually mean? Is it accurate to say there are a hundred topics in those messages, or is it more accurate to say that there are 10 topics in those messages? Is it more accurate to say that the way these topics are represented by a certain number of words, so the description of these topics, are they more accurate than another algorithm?</p><blockquote>But who’s to say that these topics are accurate? And then what does accuracy actually mean? Is it accurate to say there are a hundred topics in those messages, or is it more accurate to say that there are 10 topics in those messages? Is it more accurate to say that the way these topics are represented by a certain number of words, so the description of these topics, are they more accurate than another algorithm?</blockquote><p>It also depends on the use case that you’re working with. Sometimes we want more abstract topics because we’re searching for global trends and sometimes we’re looking for very specific topics. So in medical domains, we’re all often looking for very specific types of diseases or context or medicine, or whatever.</p><p>All these types of different things make it very difficult to say, okay, this is the truth. Because in this case, the truth often is truly in the eyes of the beholder. It’s your use case. You have a certain goal for it — something that you want to do with it. What you want to achieve with it changes the evaluation metric.</p><p>If we’re talking a little bit technical, it can be accuracy. If you have labels, it can be <a href="https://towardsdatascience.com/evaluate-topic-model-in-python-latent-dirichlet-allocation-lda-7d57484bb5d0">coherence</a>. So how coherent is a topic, but how coherent is a topic differs between what I think is coherent and, and what you think — it’s very different. Because BERTopic is a clustering algorithm. We can say, okay, we’re gonna perform the clustering, or we’re gonna evaluate the clustering.</p><p>We can evaluate maybe it’s prediction on unseen documents. But that would be, again, a supervised task. There are a lot of different ways to find out if something is accurate. because the definition of accuracy here is so difficult, and that’s why it’s such a subjective way of modeling.</p><p><strong>Seth:</strong> Yeah. I think the difficulty is that there’s not necessarily a ground truth, right? You don’t know what category or what cluster something should necessarily belong to. So it’s hard to say this is how accurate this model is.</p><p><strong>Maarten:</strong> Exactly. And the ground truth can be created, but it should each time be created from scratch depending on the use case because I can create a ground truth for my specific use case, but then ground truth for another use case will be entirely different and will sometimes require an entirely different evaluation metric.</p><p><strong>Seth:</strong> What are some of the most unique or most interesting uses of BERTopic that you’ve seen?</p><p><strong>Maarten:</strong> The most most unique ones are often the ones that try to do something with BERTopic is not really meant to do . A lot of sentiment analysis people want to do with BERTopic.</p><p><strong>Seth:</strong> That’s tricky.</p><p><strong>Maarten:</strong> It’s tricky. You can do it if you do a sort of semi-supervised approach and code in the semantic nature of it and calculate that beforehand. So it is possible. But what I actually see are a couple of things, and that’s trends analysis. But what has become a little bit more and more popular, and I found that to be really interesting is so initially I always thought topic modeling was something not used in production, right?</p><p>It’s an exploratory way of looking at your data, but I’m seeing many more cases popping up that are focusing on trying to identify clusters on the fly. So if you have a ticket system, for example, and you want to see if there are some issues that are popping up daily, yes or no, then you can use an online variant of topping modeling for that to see if new issues pop up over the couple of days.</p><p>So you can quickly, fix the issue that you have. And that’s something I developed I think half a year ago because I saw more and more use cases popping up where I say, okay, but what if I want to use it not only unseen data, but unseen data that have different topics than what it was initially trained on.</p><p>And then that was for me, a really interesting use case I hadn’t seen before.</p><p><strong>Seth:</strong> What are the techniques that are used to find new clusters over time?</p><p><strong>Maarten:</strong> So, there’s one library, I think it’s called <a href="https://riverml.xyz/0.15.0/">River</a> that mostly focuses on that part of online clustering or machine learning approaches in general, but that also allows for new information to be found. So, <a href="https://scikit-learn.org/stable/">scikit-learn</a> often focuses on not finding any new topics or new information necessarily. But to continuously train the model, that’s a valid use case. River for example, focuses a little bit more on the true online aspect of it, more finding something that we haven’t found before.</p><p>Right?</p><p><strong>Seth:</strong> Developing BERTopic what was one of the most challenging things that that you faced?</p><p><strong>Maarten: </strong>So dependency issues it’s way more tricky than I thought before. So I have a couple of packages and some are easier than others but there were a lot of API issues with <a href="https://numpy.org/">numpy</a> — they were really tricky to solve and over the course of time it remains an issue because dependencies change of course. In a way it makes sense that they change I also changed the API now and then and so things break but they can sometimes be really really tricky to fix and to find and to account for because there’s so many combinations of dependencies out there and that’s just tricky to account for all of them yeah that’s that’s one issue.</p><p>The second one is mostly API development not necessarily algorithms I’ve I find them fun they’re still difficult but it’s not really an issue right but the API in itself is something that people are using daily and if you change that in a major way then everybody has to change. Things that work now in production will stop working so it has been really tricky to think about these things in a way that accounts for anything that I might want to change in the future which is of course not possible because I’m thinking of changing <a href="https://plotly.com/python/">plotly</a> with <a href="https://bokeh.org/">bokeh</a> now and that’s going to be pain in the ass to do — that’s going to be tricky.</p><p>But there are a lot of more things if I want to do — so now you pass in documents right but at some point I want to pass in images and sound and whatever right? It requires major API changes which is difficult to do in a way that doesn’t annoy users because they’re using it quite frequently and if you’re gonna change it dramatically you know it’s not the best user experience so those things actually were the most difficult.</p><p>What often happens is when I want to develop features or I want to implement something new I have to introduce new dependency and that’s really something I think you should prevent as much as possible. When developing a package BERTopic already has quite a number of dependencies and I would have just had three in a perfect world which of course doesn’t happen but the moment you add another dependency the interaction between everything can mess everything up right because it essentially gives you another layer of complexity when accounting for all of these dependencies.</p><blockquote>…when I want to develop features or I want to implement something new I have to introduce new dependency and that’s really something I think you should prevent as much as possible… the moment you add another dependency the interaction between everything can mess everything up right because it essentially gives you another layer of complexity when accounting for all of these dependencies.</blockquote><p>So, when you’re doing topic modeling a natural package to add to that is <a href="https://networkx.org/">NetworkX</a> for example to show some sort of network relationship between the topic which would have been really fun to implement and it’s definitely doable but that would require to add another dependency that might not work well together with the ones that I already have. I did some covariant topic modeling for BERTopic which I really want to implement but that would also require me to add the stats models dependency which is a great package but it’s another dependency so I can easily create a list of 20 additional dependencies to add and then at some point BERTopic will not work anymore yeah so it’s unfortunately it’s a balance between those packages that you want to add or not right.</p><p><strong>Seth:</strong> Weighing those trade offs, trying to keep things as lightweight as possible while balancing additional features and things like that.</p><p><strong>Maarten:</strong> Yeah, exactly.</p><p><strong>Seth:</strong>. Yeah. You also refer to something called the psychology of APIs. Do you want to talk about that?</p><p><a href="https://towardsdatascience.com/6-lessons-i-learned-from-developing-open-source-projects-4617e26f247c">6 Lessons I learned from developing Open-Source Projects</a></p><p><strong>Maarten:</strong> Yeah. So, that’s what happens with me, right? I’m extremely biased with my background in psychology, so I call everything the psychology of just because I want to.</p><p><strong>Seth:</strong> You can get away with it.</p><p><strong>Maarten:</strong> Great. Then I’ll keep doing that. But the thing is like I mentioned before, when you’re doing API design, it’s about the way people interact with your package, with your model, with your software.</p><p>So then changing it dramatically hurts the user experience. So there’s a lot to be learned from popular API design out there like numpy and scikit-learn and <a href="https://pandas.pydata.org/docs/index.html">pandas</a>, where people at some point have developed a way of coding in part because of those packages. And if you try to adhere to the philosophy and the design of those packages, even when designing something new you’re focusing really on the psychology of the experience.</p><p>Because if you do everything in the scikit-learn type of way that’s really nice because everybody knows scikit-learn. And it feels intuitive when you then approach such a model. And the same applies when you’re creating a package. It doesn’t necessarily need to adhere to scikit-learn.</p><p>Whenever you create a new parameter, what kind of name are you gonna give the parameter? What is intuitive? Even the position of the parameter can be very important because sometimes people don’t want to create keyword arguments, they just throw in the variable, right? So, the position is really important when designing the package.</p><p>And there’s a lot of these types of things that you need to take into account when developing a package because, well, people are going to be using it and if people are going to be using it, okay, then there’s definitely some sort of user experience, psychological aspect of it out there where you can really make a change when it comes to the adaptation of whatever it is you have created.</p><p><strong>Seth:</strong> What are the most exciting future directions for BERTopic that you’re most excited for?</p><p><strong>Maarten:</strong> So I mentioned that briefly before.</p><p>I’m thinking of changing <a href="https://plotly.com/">plotly</a>. And I’m not doing that because I hate plotly. Plotly is awesome. And same with <a href="https://bokeh.org/">bokeh</a> or altair or whatever. It’s just that I want to add images at some point, which I’ve already done with a different package called <a href="https://github.com/MaartenGr/Concept">concept</a> that tries to integrate documents and images into one topic modeling approach.</p><p>And I want to eventually do it also in BERTopic. But in order to do that when you’re going to visualize your data points, you typically also want to visualize the images, and that’s not possible directly. In Plotly, you need to use <a href="https://dash.plotly.com/">dash</a> on top of that, which that’s not something that I’m a fan of when it comes to psychology of API design.</p><p>It changes the way you interact with the visualizations quite dramatically, quite drastically. So I’m thinking of changing it to bokeh because it allows you to do a little bit more with images. And when it comes to interesting future directions, it’s because we have more multimodal, embedding types of models.</p><p>We can do documents, images, sounds all in the same dimensional space. So if you can do that all in the same dimensional space, then there’s something to say for topic modeling also going into that direction. But there’s just a little bit more to think about than just the embeddings. Also the topic representation extraction and the way you connect the images together with the documents and perhaps sound at some point.</p><p>That’s something I have to dive a little bit deeper in. I have some code lying here and there but I think there’s also much to be gained there.</p><h4>Learning from Machine Learning</h4><p><strong>Seth:</strong> What’s an important question that you think remains unanswered in machine learning?</p><p><strong>Maarten:</strong> So, there are a few of course but there’s one thing that has been one thing popping up quite often lately, and it has to do with, of course, the ChatGPT-type of approaches. What happens when such a model gets released is that people are starting to talk about AI as if it were artificial general intelligence, right?</p><p>True intelligence, true consciousness, that it can pass the Turing test those types of exclamations that were made… And the thing is that that’s something that I think that we’re a little bit away from. But it’s starting to become a little bit of a gray field because a lot of people are talking about it in that way.</p><p>And although the definition might differ technically we’re still thinking we’re approaching that rather quickly. And if we’re doing that, I think it’s important to then also acknowledge that it’s not the case that ChatGPT is that of course. But what it then is, and how it gets accepted into industry, what is and what isn’t for me, that’s a very important question that we need to take care of.</p><p>Because although ChatGPT can do very awesome, interesting things, it’s not factual information. It doesn’t necessarily have to be, it’s not always the case. And that’s something to take into account when developing these types of models. And when we’re looking at these large language models that have billions upon billions of parameters that can do very interesting, very awesome things.</p><p>I’m really curious to see how that would then fit into the AGI (Artificial General Intelligence) discussion. But I do think the coming year will be quite drastic the way things are going now.</p><p><strong>Seth:</strong> Yeah. I think that there’s gonna be another very exciting year for natural language processing, you know, with ChatGPT and GPT-4 coming out in a couple of months.</p><p>Sure. That will obviously be interesting. But yeah, there’s definitely some sort of gap between true intelligence and creating large language models with billions or trillions of parameters. It doesn’t necessarily translate into understanding — That’s my take on it.</p><p>That touches on my next question, how has the field changed since you started working in the industry? Would it be the advent of large language models, or are there other changes that you’ve seen?</p><p><strong>Maarten:</strong> So when I started the, the large language models or more when I started doing machine learning the <a href="https://towardsdatascience.com/word2vec-explained-49c52b4ccb71#:~:text=of%20the%20corpus.-,Word2Vec%20Architecture,other%20words%20in%20the%20corpus.">Word2Vec</a> type of models were really coming up and showing the possibilities of those.</p><p>And then slowly into the <a href="https://towardsdatascience.com/illustrated-guide-to-transformers-step-by-step-explanation-f74876522bc0">transformer</a> type models. And from there, of course we have what we have currently. I would say that the large language models really change the entire field, especially when you consider what <a href="https://huggingface.co/">Hugging Face</a> is now, right. With the amount of models that they host and the things that they can do.</p><p>What I do want to say is that we see there’s a huge trend going towards those huge major large language models that nobody can actually run on their own machine. And there’s a very small piece of research going there which is a little bit focused more on distillation and making it accessible for you and me, right?</p><p>I have a laptop here that has an okay GPU, but not much more than that. And I want to run these models without making use of an api. And that’s not always possible because it’s not that fast. And there are a lot of, there’s still a lot of research packages that go into the direction. If you take <a href="https://www.sbert.net/">sentence-transformers</a>, for example, they distill it and make a very accurate representation of what it has been trained on quite fast.</p><p>So the inference and the speed of using those models is really great and I would love to see more of those instead of billions upon billions of parameters. Which is nice though, don’t get me wrong. We need to have those to essentially get to a distillation or smaller version. But it would be nice if there was as much attention to the smaller models as there is to those larger ones, because I can use the smaller ones.</p><p>I mean, larger ones are, are not really accessible to me. Right.</p><p><strong>Seth:</strong> Do you see any connection with how generative models can affect something like topic modeling and BERTopic? Have you thought about incorporating it at all?</p><p><strong>Maarten:</strong> Yeah. That comes back to dependencies and things like that.</p><p>But what I think Cohere has done is they created topically, I think it’s called. That’s basically a GPT on top of BERTopic. So BERTopic generates a number of topics and descriptions of those topics by number of words. And they feed those words essentially into a GPT model and ask to create a topic out of that - just a natural description.</p><p>And it does that really well. The issue with such an approach is that from a production setting, it’s really difficult to run on your machine, right? You need to have an API, which is absolutely fine if you’re using Cohere, it’s an amazing service. But when you’re running things locally without any internet connection, having a local GPT model is just not done.</p><p>The moment these models become relatively small and doable on your machine — that’s the moment I will start to integrate these into BERTopic. But if you want to use GPT on a couple of hundreds of topics, but slice that over, I don’t know, a hundred classes, that becomes really difficult and really long to calculate all of these things.</p><p>That’s the reason why I wouldn’t want to integrate that into BERTopic right now. Right? That’s why what Cohere has done is amazing. It’s something on top of BERTopic if you feel like you need it. Which is also something I would definitely advise using if you have that service or if interested in using that.</p><p>But integration in BERTopic is a little bit too far ahead until performance reaches the state BERTopic is right now.</p><p><strong>Seth:</strong> Right. That that makes sense. Yeah, I’ve experimented using some of the outputs of BERTopic and putting them into some open source generative models, and I’ve been pleasantly surprised.</p><p><strong>Maarten:</strong> Ah, that’s great.</p><p><strong>Seth: </strong>I think that that’s definitely gonna be interesting future research. What do you think about the hype that these new generative models are creating? Do you think there’s a big gap between the hype and the reality?</p><p><strong>Maarten:</strong> Yes and no. It’s an annoying answer I know, but no, I don’t necessarily think there is a gap because a lot of these models, let’s take ChatGPT for example or <a href="https://huggingface.co/spaces/stabilityai/stable-diffusion">Stable Diffusion</a> may be a little bit more interesting. So basically from text we’re creating images, that’s amazing that we can do that.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/768/1*jaf_AtA55HrHb_XpwafO9Q.jpeg" /><figcaption>Image created using <a href="https://huggingface.co/spaces/stabilityai/stable-diffusion">Stable Diffusion</a> with the Prompt: “incredible image of someone learning how to do machine learning”</figcaption></figure><p>That’s truly amazing. And there are people out there who are actually great prompters who know which words they need to use in order to create awesome imagery. And if you can use that in a very smart way there are a lot of use cases where this might be interesting for mockups, for example, or when you’re creating art for your game or something like that.</p><p>There’s still a gap in the sense that it’s not exactly doing the things you want, right? You’re still searching a lot of ways to make sure it gives you the output that you need. It’s a very large model that doesn’t run as fast maybe as you might want it. There’s a lot of issues when it comes to artists being copied right over the internet.</p><p>Over the last few months, I’ve seen quite a lot of people who say, okay, somebody trained on my specific artwork and now it generates it as if it were mine. Well, that’s of course not the way that we should go into. So I think there’s a lot of hype that that makes sense, but a lot of the hype is also as a result of something that’s just fun to use.</p><p>Whether it’s something that will actually be used in a lot of organizations, I highly doubt that. But there are certain use cases where I do think this will make a huge impact, but they’re just a limited few. And for a lot of people, this is just, just something fun to use and something that’s easy to go viral.</p><p><strong>Seth:</strong> Right? Yeah. It’s pretty unbelievable. The impact that ChatGPT had getting over a million users in Yeah. In under a week interacting with it. It’s definitely something special. I think they slow down the output so it makes it feel like you’re talking to a human.</p><p><strong>Maarten:</strong> Maybe it also has to do with I’m not entirely sure, just to make sure that people are still using it without querying it too much.</p><p>Right. It slows down, it slows down the time you can actually query it, but even here in the Netherlands, there are a lot of people using it in Dutch and it does it relatively well and that’s interesting to see and I’m wondering when I will see it popping up in actual research and contracts people are writing things like that. It’s amazingly viral.</p><p><strong>Seth:</strong> Yeah. Have you used it at all to help, to help you with anything? Like even like a day-to-day thing?</p><p><strong>Maarten:</strong> Not much actually, because it still requires me to correct it quite a lot of the time, and I’m somebody who likes to do it in my way.</p><p>But I have been using it when I’m writing articles to come up with ten titles for an article, for example, just to give you some inspiration and to feed off and then I’ll change it nine of the ten times. But I do think it’s an amazing way to get an initial idea of something that you want to create.</p><p>If you want to create an introduction about federated learning, just type it in. It can send something back and you can change it the way you want it, you can adapt it. It’s really interesting to see that. Maybe at some point we don’t have to type all these things out ourselves. We just ask somebody or ask ChatGPT to do that.</p><p>And then it iteratively ask it to change it in the way that fits with what you’re trying to achieve. So the way we’re writing and interacting might change completely if something like this gets a little bit more open source or eventually more accessible to the public than just an API, for example.</p><p>And that will definitely be interesting, right?</p><p><strong>Seth:</strong> Yeah. I think there’s a certain amount of creativity in creating the right prompts for it. And also that is the way the future’s going to be. It’s this interaction that humans have with the machine to create the best output, to get maybe an initial idea and then you could work off of it or modify it.</p><p><strong>Maarten:</strong> And that saves so much time in the process also, right? I mean, you know what you want to write. In many cases, you’re writing a paper and you want to write an introduction about, you know federated learning, or maybe stable diffusion, or maybe it’s something else. Just ask somebody else to write an initial draft of it and then you can change it.</p><p>You already know what you want to write, and if you’re an expert on the field, then the only thing you have to do is just go over it and check whether everything is correct and properly done. But it saves so much time. There’s also risk involved in all of this, of course, when people are using it and just copy pasting it as is right.</p><p>But eventually we’ll also find some way to deal with that.</p><p><strong>Seth:</strong> Yeah. I think there’s an interesting aspect to it that maybe some people forget. Large Language Models are so powerful because of all of the information that has been created by humans, that’s what laid the foundation for all of this.</p><p>So going back into machine learning — are there any people in the field that inspire you — that you’re very influenced by?</p><p><strong>Maarten:</strong> So there are a few people that I follow, but I wouldn’t necessarily say there are these people that really influence me tremendously. And the reason for that it’s kind of, because from a psychological background, I tend to look at things from as many perspectives as I possibly can, and that the same thing applies to people that influence me.</p><p>So I try to take snippets of what people are saying and use that for what I think okay, this is interesting. Because everybody has their strengths and their weaknesses, of course, and I focus on those strengths, but I take 20, 30 people that I think, okay, these are awesome and they are great at that specific thing and I try to combine them.</p><p>So I started out learning transformer based models with the <a href="https://jalammar.github.io/illustrated-transformer/">visualizations of Jay Alammar</a>.</p><p><strong>Seth: </strong>Incredible, classic.</p><p><strong>Maarten:</strong> He’s doing amazing work with respect to that. But that’s one aspect that I take from him, and then I take something else from somebody completely different and use that. So it’s a combination more of people than necessarily, okay, this is somebody that I think, okay I love everything that person is doing which seldom is the case because, you know, we’re people, we’re not great at everything. So, you know, this person’s better at this, this person, better visualization, better coder.</p><p>Now I’m a big fan of the <a href="https://www.sbert.net/">Sentence-Transfornmers</a> Library, for example. Yeah. So there’s, there’s one interview that’s coming up with Nils Reimers — amazing work, right? Yeah. But if I want visualizations, I go to Jay [Alammar]. You know, it’s that combination of people, and I think you find them eventually yourself.</p><p>Because I can say, here’s a list of 20 people that I find interesting, but I find them interesting because I am who I am and I look at things in a certain way, so, also trying to kind of support my way of thinking with views that support mine. I’m still human in that way. So even if I give these people, it might not fit with somebody else who just sees things differently.</p><p>Yeah. So I’m sorry, this kind of a non-answer,</p><p><strong>Seth:</strong> No, it’s a great take on it. I like that. Are there any people outside of machine learning that inspire you?</p><p><strong>Maarten:</strong> Well, that’s a good question. So I recently did that animation for BERTopic.</p><p><strong>Seth: </strong>Love it.</p><h3>Maarten Grootendorst on Twitter: &quot;It is finally here, the v0.13 release of BERTopic! 🎉Explore multi-topic assignments, supervised topic modeling, outlier reduction, light- and heavyweight options, and much more in this bigger-than-expected release!Changelog:https://t.co/tHxCklC0WiAn overview thread👇🧵 pic.twitter.com/SrlRAOagwT / Twitter&quot;</h3><p>It is finally here, the v0.13 release of BERTopic! 🎉Explore multi-topic assignments, supervised topic modeling, outlier reduction, light- and heavyweight options, and much more in this bigger-than-expected release!Changelog:https://t.co/tHxCklC0WiAn overview thread👇🧵 pic.twitter.com/SrlRAOagwT</p><p><strong>Maarten:</strong> Yeah. Thank you. I wasn’t fishing for that, but it’s welcome.</p><p><strong>Seth: </strong>I still love it.</p><p><strong>Maarten:</strong> So I did that with the software of <a href="https://www.youtube.com/channel/UCYO_jab_esuFRV4b17AJtAw">3Blue1Brown</a>. And I really truly enjoy that. The YouTube channel with mathematics in general and the way it’s being visualized and there’s even a major package on top of that’s open source that everybody can use.</p><p>Those types of people, I really enjoy following because they just give information in such a pleasant way that it feels like I’m just watching Netflix, right? If you can do that. So <a href="https://www.youtube.com/channel/UCsXVk37bltHxD1rDPwtNM8Q">Kurzgesagt</a> for example, I did something quite recently in an <a href="https://medium.com/towards-data-science/using-whisper-and-bertopic-to-model-kurzgesagts-videos-7d8a63139bdf">article</a> about that, that’s also a channel that gives a lot of information about different types of subjects.</p><p>If you can present complicated things in a way as if you’re watching a movie, that’s truly beautiful if you can do that. Yeah. I’m truly impressed with that. So, maybe those types of people influence me.</p><p><strong>Seth:</strong> Yeah. What I’ve found in data science and machine learning is that, you know, there’s the hard skills, right?</p><p>And then there’s the soft skills and being able to explain the work that you’re doing and blending in the psychology of it and understanding how people learn and how people think about things and how people perceive and sort of understanding what’s the best way of saying this? What’s the best way of visualizing this?</p><p>It’s so important because then you allow other people in, right? And then you can get their feedback and then that’s when things really start to kind of take off and can reach new heights.</p><p><strong>Maarten:</strong> Yeah. I fully agree.</p><p>It’s something that I, of course, agree I’m super biased in that way, but in this field I really think that sometimes one really cannot exist without the other. And sure I can create the most complex algorithm out there to do some sort of specific task, but if I cannot communicate it properly to my stakeholders, then truly what’s the point?</p><p>I see that in open source where a lot of packages are really, really impressively done, but are really difficult to use, so nobody’s using them, which is just a shame because there’s not much thought put into that psychological aspect of it.</p><p>But it also works the other way around, right? If I want to explain something, something technical to a person, a difficult algorithm, I really need to understand that algorithm really, really well in order to be able to explain it.</p><p>So I can know how to explain it in a way, how to communicate in a way, how to phrase things in a way to capture an audience, but it doesn’t actually mean that I understand the content. So it works both ways.</p><p>It’s always nice when people say, okay, you need to have soft skills, but in order to have soft skills, you really, really, really need those hard skills also, because if you don’t know what you’re talking about, well, you can bullshit your way out of it just a little bit, but people will find out.</p><p><strong>Seth:</strong> You need to be able to back it up with the hard skills for sure.</p><p><strong>Maarten:</strong> Yeah. And in some cases, I really think they’re just as important. Not in all cases, I mean, it’s not that black and white, but if we’re talking in a business sense, then you just need to be sure that you’re able to do your job.</p><p>But then the rest of it definitely is selling it, saying to people, okay, this is worth investing in. And in order to do that, you need to explain it. Well, you need to attract people. And it’s not always the case that you need to do that as a data scientist, sometimes you have managers that do that for you — that abstract those things away from you because there are a lot of technical data scientists out there that don’t necessarily want to be focused on a lot of meetings and a lot of presentations and things like that, which is fair. But then there still needs to be someone who does all of the well selling, essentially.</p><p><strong>Seth:</strong> Learning how to deal with all the different types of stakeholders from a non-technical person to the CEO, to the head of product, to another data scientist, it’s so important and understanding who your audience is and the best way of getting your learnings across.</p><p>To get into the Learning from Machine Learning.</p><p><strong>Maarten:</strong> Awesome name, by the way. That’s great.</p><p><strong>Seth:</strong> Yeah. Thank you. I appreciate it. What’s one piece of advice that you’ve received that has helped you in your machine learning journey?</p><p><strong>Maarten:</strong> Yeah, for me I don’t necessarily know who I’ve gotten this advice from, but understanding the problem truly is a solution.</p><p>I’m gonna say that over and over and over again. If you really don’t know the problem that you’re trying to solve, there’s no way you’re gonna get a good solution. And it seems really simple, but it’s not. This is one of the most difficult things that you can do. And there are so few people that can actually truly try or truly understand the problem that they’re working with.</p><blockquote>I’m gonna say that over and over and over again. If you really don’t know the problem that you’re trying to solve, there’s no way you’re gonna get a good solution. And it seems really simple, but it’s not. This is one of the most difficult things that you can do. And there are so few people that can actually truly try or truly understand the problem that they’re working with.</blockquote><p>I’m working now, for example, in cancer research and I have a lot of amazing colleagues, and there’s one of them that really dives into the issue before coding, right? Takes weeks before really, you know, he still codes and still looks at, at what a potential solution might be, but spends weeks just trying to figure out what exactly is happening here.</p><p>How the data looks like, what is the context in which it’s being used, what are the difficulties, the connection with legal, for example — which makes coding so much easier because you have done all of the groundwork, right? And then you know exactly, but exactly what you want to program, what you want to predict once you want to solve.</p><p>But that’s immensely difficult because it’s more than just a technical problem. It often also relates to, okay, how is the data being generated at all. You know, in most cases, not always, there is some sort of psychological component to it, or there’s some issue with the data, which is always the case.</p><p>There’s always some sort of, you know a dirty data out there.</p><p><strong>Seth:</strong> Peculiar artifact, right? Like something’s going on.</p><p><strong>Maarten:</strong> Exactly, exactly. And that really depends on the process. So you need to understand the process of data generation. But then if you’re going to devise a solution, who are you devising it for?</p><p>And why do they want that solution? Because what often happens is that people ask for solution, and they might actually mean something entirely different. It’s not their fault because they try to communicate it, but because you are our technical person and you know what the possibilities are, you might say, okay, but then you might need to focus more on this and this direction.</p><p>And that might happen that if you don’t do all of this, that eventually you push out a model where people say, “Ah, that’s not really what we’re looking for.” And then what you also often see happening is that the one who created the solution say “Oh, but you should have told me that before.” That’s not the way we should communicate, right?</p><p>It’s a two-way street. And you cannot just tell somebody, okay, you need to tell me exactly what you need as if you were a data scientist. It doesn’t work that way. We need to help each other. And understanding the process, understanding the problem really is I think the best way to spend your time in data science.</p><p><strong>Seth:</strong> 100 percent. Sometimes just a pen and paper right. Before you get into the code.</p><p><strong>Maarten:</strong> Yep, definitely.</p><p><strong>Seth: </strong>And then also, just to add on, I would say, Revisiting the problem, right? Like yeah. Maybe even after you start working on a little bit, then you start to realize, oh, you know, all the requirements weren’t quite addressed just yet.</p><p>You know, .</p><p><strong>Maarten:</strong> No, that’s perfect. That’s amazing because, you know, it starts from the problem and as you mentioned, going back frequently to see, okay, I’m still on track with what I perceive to be the problem at first might change later on when you start devising the solution.</p><p><strong>Seth:</strong> Yeah. The tricky thing about this sort of iterative process in data science is once you do kind of pick a solution at some point. It starts to narrow your vision field a little bit, right?</p><p><strong>Maarten:</strong> Sure.</p><p><strong>Seth:</strong> So that’s why you always have to, you always have to go back — understanding and figuring out what’s this from a business standpoint, what’s this from the data standpoint? Yeah, exactly.</p><p><strong>Maarten:</strong> Yeah. And it’s okay if it becomes a little bit narrower over time. Of course. You start broad now trying to explore everything that’s out there and eventually a solution. So it has to become narrow. But as you mentioned, not too fast it should take some time and it should, you know, it’s, it’s like reinforcement learning.</p><p>There should be a balance between exploration and exploitation, right? You know, we’re gonna devise a solution, but we still gonna keep some room for some exploration of some different things to see if that still works or if that’s still relevant or whatever.</p><p><strong>Seth:</strong> Right. That, that upfront time can save you so much time.</p><p><strong>Maarten:</strong> So much time. Yes.</p><p><strong>Seth:</strong> What advice would you give to someone who is just starting out in machine learning or natural language processing?</p><p><strong>Maarten:</strong> So I would really start doing projects as much as possible, and it’s yeah, I’m really okay if it’s just a Titanic data set or housing regression prices.</p><p>I think that’s one. And sure there are a lot of people who’ve already done that, but getting your hands dirty is exceedingly important in data science because at some point you develop an intuition about certain things. You see a problem, you don’t know exactly why, but you feel like, okay, this should be the solution.</p><p>And that happens because, I mean, you are also a predictive model in a way. You know that gut feeling — That’s because you learned something before with the previous project where you found something to be working well or something that’s very similar to what you’re doing right now.</p><blockquote>And that happens because, I mean, you are also a predictive model in a way. You know that gut feeling — That’s because you learned something before with the previous project where you found something to be working well or something that’s very similar to what you’re doing right now.</blockquote><p>And by doing all of these projects and by just getting your hands dirty and analyzing stuff, gathering your own data, cleaning your own data, applying these algorithms going into depth that helps you really understand the field, gets that intuition about when something is working and when something isn’t working.</p><p>A famous example is what most people say when they’re starting out, is if you see a model that has a 99% accuracy. Exactly. There’s going to be something wrong.</p><p><strong>Seth: </strong>Oh, no</p><p><strong>Maarten:</strong> That doesn’t happen. Generally it doesn’t happen. So that’s an intuition that you develop. At some point, and this is a famous one of course but a lot more of these are out there. There’s a lot of these that have been implemented eventually into BERTopic because there were a lot of things I felt like, oh, there’s something weird going on here.</p><p>I’m not entirely sure, but I feel like I should look there and there. Yeah. Now, the only reason why I would do that is because I’ve had a lot of experience with a lot of these projects and it would know immediately where to look, even though it wasn’t something I had already done before. I think if you’re starting out, just do the work,</p><p><strong>Seth:</strong> Right. That real world or the real project experience can sort of give you an idea maybe where there might be a pitfall, where there might be a pothole and that experience can help you navigate a little bit better.</p><p><strong>Maarten:</strong> Yeah, definitely. And if you’re really starting out, don’t be scared of people saying, don’t use the Titanic dataset and don’t use housing dataset.</p><p>Everybody has to start somewhere, and it’s okay to start with perfectly clean data. It really is. And then you can slowly build your way up to less clean data, gathering your own data and stuff like that.</p><p><strong>Seth:</strong> Right. So in a similar direction, what advice would you give yourself when you were starting out?</p><p><strong>Maarten:</strong> Oh, that’s an awesome question. Oh, I really have to think about that.</p><p>Oh, that’s tricky because I’m drawing from my own experience of course. Maybe start in the field a little bit earlier. And what I mean with that is I transitioned of course, and because of that I started in the field relatively late. There were a lot of people my age who already had four or five years of experience, and although doing a master’s is nice and definitely recommended.</p><p>Experience brings you a little bit further I think, Ah, maybe not a little bit further, but it’s, it’s just as valuable. So if you can do both, that’s amazing of course.</p><p>And I think what I never really, truly had is a mentor in my years of data science, not somebody I could always look up to and ask questions about the things that I’m dealing with. I had to discover everything for myself and that has it’s pros, but the main con it slows down some of the learning processes tremendously.</p><p>So the advice I would’ve given myself is maybe focus a little bit less on academics during that time and a little bit more on getting that experience.</p><p><strong>Seth:</strong> It’s still good advice and I like the advisor or mentor idea of that.</p><p>I think that that’s something that people tend to shy away from, but I don’t think they should. It’s an important thing that can help you along in your journey. Sometimes it’s nice to do everything sort of on your own, right? And you kind of face the learning pains and you get that real strong sense of accomplishment when you get it done.</p><p><strong>Maarten:</strong> Exactly.</p><p><strong>Seth: </strong>But it would definitely Speed things up if you have someone, even just a sounding board to talk to.</p><p><strong>Maarten:</strong> Exactly. And, a mentor can even do that, right? A good mentor sometimes will say, go figure it out yourself. Because a good mentor will know your capabilities and will know how persistent you are and what you can and cannot do.</p><blockquote>A good mentor sometimes will say, go figure it out yourself. Because a good mentor will know your capabilities and will know how persistent you are and what you can and cannot do.</blockquote><p>So, he or she will know at some point, okay, if you spend a few more hours, I’m, I’m sure you will solve it, right? And if, by then it’s not possible, okay, I will nudge you in the right direction. Right? But that’s all you get. Right. A nudge. So yeah, that’s something that, would have made a huge difference for me, I think.</p><p>A mentor.</p><p><strong>Seth:</strong> That’s what a good teacher or a good mentor can provide. So to get a little philosophical or bigger picture what has a career in machine learning? Taught you about life?</p><p><strong>Maarten:</strong> Difficult questions you make me think here. No, but really for me machine learning has come at the same point in my life, roughly as chronic pain.</p><p>And, dealing with that was quite difficult. So finding a balance between a healthy life and the thing I really enjoyed most — machine learning at the time. That for me has been quite difficult finding a nice balance between those two, because the chronic pain makes it, for me, difficult to type 10 hours a day for example.</p><p>I really have to balance the way and the amount that I work. And that’s tricky, right? So when you switch careers, that’s already quite difficult. But then when you finally find something that you think, okay, that’s it, that’s what I want to do. And then at the same freaking time, you get chronic pain.</p><p>That’s not something that makes you that happy. So for me, that has been something that has been very difficult, but it’s been, in hindsight, a very valuable learning experience. Really truly knowing what I stand for, what is important to me, what I want to focus on to take a step back at times.</p><p>Because that’s also the field of machine learning — It goes fast, it goes so extremely fast, and there’s nobody that can keep up with everything that’s happening there. But the fear of missing out Yeah. Becomes tremendous. Becomes huge. When things like stable diffusion and chat GPT and all of these things get released on I would almost say on a daily basis it feels at times.</p><p>But I’ve really learned to then take that step back and to view the broader picture. You don’t have to know everything. There are a few things that are more or less important than others that you can follow, but eventually you will catch up with whatever is happening.</p><blockquote>But I’ve really learned to then take that step back and to view the broader picture. You don’t have to know everything. There are a few things that are more or less important than others that you can follow, but eventually you will catch up with whatever is happening.</blockquote><p>I mean, it’s great that ChatGPT is doing so well, but I wouldn’t have to have known it from day one if I weren’t making some sort of business out of the model. There’s no need for me to know it then, so I can give myself a little bit more space to find that balance. , although with BERTopic and writing and KeyBERT, it’s still difficult.</p><p><strong>Seth:</strong> Yeah. I’m not sure how you find the time to maintain such incredible libraries and keep a day job</p><p><strong>Maarten:</strong> I have no clue. I try my best. I’m not sure my wife is always that happy, but no, I’ll make sure to find a balance between that. But you’re right in saying, you know, the day job combined with open source and writing is difficult.</p><p>So maybe somewhere in the future I would love to combine that. But you know, that depends. It’s open source, right? Maybe if at some point there’s an organization that says, okay, we’re gonna support that fully, then maybe that’s something that’s interesting to me at some point. But for now, I’ve amazing job.</p><p>I get a lot of freedom to also work around the chronic pain, which is awesome because there are not many organizations that provide that. And still work on the things that I think are important.</p><p><strong>Seth:</strong> Well, your work is greatly appreciated by this machine learning scientist, and I’m sure many, many others out there.</p><p>So just to conclude, if there are people who are interested in either reaching out to you or learning more about you, are there any places where they could go?</p><p><strong>Maarten:</strong> You can go on Twitter, on LinkedIn. Those are the main sources for me where you can find me. I have a website.</p><p>It’s maartengrootendorst.com, which only Dutch people can understand. It’s just my name, but it’s the way you pronounce it. It’s not the easiest thing to translate into English. But no Twitter and LinkedIn, mostly LinkedIn. You can reach me there quite well. Yeah.</p><p><strong>Seth:</strong> Awesome. Maarten, it has been such a pleasure talking with you.</p><p>Thank you. Thank you so much.</p><p><strong>Maarten:</strong> Thank you for having me. It was a great interview. I love those questions. Really, most of them are really technical, right? That’s what you do mostly if you have something like KeyBERT or BERTopic. But focusing a little bit more on the philosophical, psychological aspects of it is truly appreciated.</p><p><strong>Seth:</strong> Absolutely. Thank you so much.</p><p><strong>Maarten: </strong>Thank you.</p><h3>Listen Now</h3><p>The video of the full interview can be seen here:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FbLW_RH2Y6oI%3Ffeature%3Doembed&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DbLW_RH2Y6oI&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FbLW_RH2Y6oI%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/c49746562c4fa3a8e2364b0129cc91c4/href">https://medium.com/media/c49746562c4fa3a8e2364b0129cc91c4/href</a></iframe><p>The podcast is now available on all <a href="https://rss.com/podcasts/learning-from-machine-learning/">podcasts platforms</a>:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fopen.spotify.com%2Fembed%2Fepisode%2F0HYf9ZXFniiWQvSp0qOJ4s%3Futm_source%3Doembed&amp;display_name=Spotify&amp;url=https%3A%2F%2Fopen.spotify.com%2Fepisode%2F0HYf9ZXFniiWQvSp0qOJ4s&amp;image=https%3A%2F%2Fi.scdn.co%2Fimage%2Fab67656300005f1f87a1155f89111000b7c75a33&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=spotify" width="456" height="152" frameborder="0" scrolling="no"><a href="https://medium.com/media/d637d58ea31b0a8043a7be4f4b28c04e/href">https://medium.com/media/d637d58ea31b0a8043a7be4f4b28c04e/href</a></iframe><h3>Resources</h3><p>Learn more about Maarten Grootendorst and his work:</p><ul><li><a href="https://www.maartengrootendorst.com/">Maarten Grootendorst’s Blog</a></li><li><a href="https://maartengr.github.io/BERTopic/">BERTopic</a>: BERTopic is a topic modeling technique that leverages 🤗 transformers and c-TF-IDF to create dense clusters allowing for easily interpretable topics whilst keeping important words in the topic descriptions.</li><li><a href="https://maartengr.github.io/KeyBERT/">KeyBERT</a>: KeyBERT is a minimal and easy-to-use keyword extraction technique that leverages BERT embeddings to create keywords and keyphrases that are most similar to a document.</li><li><a href="https://www.linkedin.com/in/mgrootendorst/">LinkedIn</a></li><li><a href="https://twitter.com/MaartenGr">Twitter</a></li><li><a href="https://medium.com/@maartengrootendorst">Medium</a></li></ul><p>Resources to learn more about Learning from Machine Learning:</p><ul><li><a href="https://www.youtube.com/channel/UCGwSWuvSRzop4ZVG2z9NmnQ">Subscribe on Youtube</a></li><li><a href="https://www.linkedin.com/company/learning-from-machine-learning/">Learning from Machine Learning on LinkedIn</a></li><li>Reach out to <a href="https://www.linkedin.com/in/sethplevine/">me on LinkedIn</a></li><li><a href="https://medium.com/@levine.seth.p">Follow me on Medium</a></li></ul><h3>Video Contents</h3><ul><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=0s">00:00</a> Learning from Machine Learning Intro</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=20s">00:20</a> Maarten Grootendorst Introduction</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=54s">00:54</a> Career Background</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=141s">02:21</a> Background in Psychology</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=247s">04:07</a> What attracted you to machine learning?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=318s">05:18</a> What were the first skills you needed to learn?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=432s">07:12</a> What is BERTopic? What is the power of Topic Modeling?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=578s">09:38</a> What was the goal of the initial package and how has it changed?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=761s">12:41</a> Benefits of BERTopic</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=798s">13:18</a> Why is it so difficult to evaluate a topic model?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=966s">16:06</a> The challenge of assigning ground truth</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=1004s">16:44</a> What are some of the most unique use-cases of BERTopic?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=1118s">18:38</a> What are the techniques to find new clusters over time?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=1169s">19:29</a> What was one of the most challenging things you faced developing BERTopic?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=1453s">24:13</a> What is the psychology of APIs?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=1744s">29:04</a> What’s an important question that you believe remains unanswered in Machine Learning?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=1882s">31:22</a> Exciting year for natural language processing</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=1920s">32:00</a> How has the field changed since you started working in the industry?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=2062s">34:22</a> How can generative models effect Topic Modeling and BERTopic?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=2220s">37:00</a> Is there a gap between the hype and reality of these generative models?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=2355s">39:15</a> ChatGPT</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=2490s">41:30</a> Interaction between Humans and Machines</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=2598s">43:18</a> Are there any people in the field that inspire you?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=2830s">47:10</a> Sharing your Data Science work with others</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=3048s">50:48</a> Learning from Machine Learning</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=3060s">51:00</a> What’s one piece of advice that you have received that have helped you in your ML Journey?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=3270s">54:30</a> Revisiting the problem</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=3390s">56:30</a> What advice would you give someone just starting out in the field?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=3576s">59:36</a> What advice would you give yourself when starting out?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=3755s">1:02:35</a> What has a career in machine learning taught you about life?</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=3990s">1:06:30</a> Wrap-up</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=4008s">1:06:48</a> How to reach Maarten Grootendorst</li><li><a href="https://www.youtube.com/watch?v=bLW_RH2Y6oI&amp;t=4040s">1:07:20</a> Conclusion</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/proxy/1*PFFPUbl_Foj2fC_ehd83CQ.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=9ed9b9b2921" width="1" height="1" alt=""><hr><p><a href="https://medium.com/data-science/learning-from-machine-learning-maarten-grootendorst-bertopic-data-science-psychology-9ed9b9b2921">Learning from Machine Learning | Maarten Grootendorst: BERTopic, Data Science, Psychology</a> was originally published in <a href="https://medium.com/data-science">TDS Archive</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[It’s the Golden Age of Natural Language Processing, So Why Can’t Chatbots Solve More Problems?]]></title>
            <link>https://medium.com/data-science/its-the-golden-age-of-natural-language-processing-so-why-can-t-chatbots-solve-more-problems-bada37a427cf?source=rss-70f6655bd458------2</link>
            <guid isPermaLink="false">https://medium.com/p/bada37a427cf</guid>
            <category><![CDATA[chatbots]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[notes-from-industry]]></category>
            <category><![CDATA[customer-service]]></category>
            <category><![CDATA[editors-pick]]></category>
            <dc:creator><![CDATA[Seth Levine]]></dc:creator>
            <pubDate>Tue, 14 Dec 2021 19:03:01 GMT</pubDate>
            <atom:updated>2023-04-21T04:04:10.324Z</atom:updated>
            <content:encoded><![CDATA[<h4><a href="https://towardsdatascience.com/tagged/notes-from-industry">Notes from Industry</a></h4><h4>Natural Language Processing (NLP) is revolutionizing the way we live and interact with machines and each other. But chatbots are still falling short. Here’s why.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/832/0*aCO2ULMzNkhmLYix.jpg" /><figcaption>Image from<a href="https://pixabay.com/illustrations/chatbot-chat-application-artificial-3589528/"> Pixabay</a></figcaption></figure><p>Advancements in the field of AI and Machine Learning (ML), specifically Natural Language Processing (NLP), are <a href="https://loris.ai/blog/human-vs-digital-customer-service-how-to-find-the-right-balance/">revolutionizing the way we live and interact with machines and each other</a>. Amazon’s Alexa, Apple’s Siri, virtual assistants and chatbots have changed the way we request and receive information. NLP pipelines take unstructured text data and process it into a structured format to extract information, categorize, and derive insights. It means that computers can convert human language into a form that can be streamlined and processed for AI systems. Ultimately, they make <a href="https://www.kdnuggets.com/2021/03/natural-language-processing-pipelines-explained.html">communications between humans and machines easier</a>.</p><p><strong>How it works</strong></p><p>Computers excel in various natural language tasks such as text categorization, <a href="https://github.com/syhw/wer_are_we">speech-to-text</a>, <a href="http://nlpprogress.com/english/grammatical_error_correction.html">grammar correction</a>, and large-scale analysis. ML algorithms have been used to help make significant progress on specific problems such as <a href="http://nlpprogress.com/english/machine_translation.html">translation</a>, <a href="http://nlpprogress.com/english/summarization.html">text summarization</a>, <a href="http://nlpprogress.com/english/question_answering.html">question-answering</a> systems and <a href="http://nlpprogress.com/english/intent_detection_slot_filling.html">intent detection and slot filling</a> for task-oriented chatbots.</p><p>For example, that grammar plug-in built into your word processor, and the voice note app you use while driving to send a text, is all thanks to Machine Learning and Natural Language Processing.</p><p>However, as smart as these bots may appear, humans are smarter, for now. When it comes down to it, human language is nuanced and often ambiguous. This presents serious challenges for ML systems. Training on massive datasets — for example: every Wikipedia article — and creating large language models does not lead to an understanding of language.</p><p>In fact, this methodology can perpetuate human falsehoods and misconceptions.</p><p>Despite the widespread usage, it’s still unclear if applications that rely on language models, such as generative chatbots, can be <a href="https://arxiv.org/abs/2108.07258">safely and effectively released into the wild without human oversight</a>. Remember the 2014 movie <a href="https://www.imdb.com/title/tt0470752/?ref_=nv_sr_srsg_0">Ex Machina</a>? It may not be that extreme but the consequences and consideration of these systems should be taken seriously.</p><p><strong>Recent advancements in NLP have led to an explosion in innovation</strong></p><p>Just within the past decade, technology has evolved immensely and is influencing the customer support ecosystem. With this comes the interesting opportunity to augment and assist humans during the customer experience (CX) process — using insights from the newest models to help guide customer conversations.</p><p>The leading drivers for breakthroughs in NLP and the explosion of language models include:</p><ol><li>The development of new techniques and algorithms (word embeddings and <a href="https://medium.com/inside-machine-learning/what-is-a-transformer-d07dd1fbec04">transformers</a>).</li><li>Improvements and greater access to high-end hardware (GPUs and <a href="https://en.wikipedia.org/wiki/Tensor_Processing_Unit">TPUs</a>).</li><li>Open-source tools and libraries — such as <a href="https://spacy.io/">SpaCy</a>, <a href="https://huggingface.co/">Hugging Face</a>, and <a href="https://rasa.com/">Rasa</a> — have democratized NLP.</li></ol><p>Transformers, or <a href="https://medium.com/inside-machine-learning/what-is-a-transformer-d07dd1fbec04">attention-based models</a>, have led to higher performing models on natural language benchmarks and have rapidly inundated the field. Text classifiers, summarizers, and information extractors that leverage language models have outdone previous <a href="http://nlpprogress.com/">state of the art results</a>. Greater availability of high-end hardware has also allowed for faster training and iteration. The development of open-source libraries and their supportive ecosystem give practitioners access to cutting-edge technology and allow them to quickly create systems that build on it.</p><p>These advancements have led to an avalanche of language models that have the ability to predict words in sequences. Think of <a href="https://blog.google/products/search/how-google-autocomplete-predictions-work/">Google’s autofill</a>. Models that can predict the next word in a sequence can then be fine-tuned by machine learning practitioners to perform an array of other tasks.</p><p>OpenAI’s GPT-3 — <a href="https://venturebeat.com/2021/11/18/openai-makes-gpt-3-generally-available-through-its-api/">a language model that can automatically write text</a> — received a ton of hype this past year. Beijing Academy of AI’s <a href="https://towardsdatascience.com/gpt-3-scared-you-meet-wu-dao-2-0-a-monster-of-1-75-trillion-parameters-832cd83db484">WuDao 2.0</a> (a multi-modal AI system) and Google’s <a href="https://arxiv.org/abs/2101.03961">Switch Transformers</a> are both considered more powerful models that consist of over 1.6 trillion parameters dwarfing GPT-3’s measly 175 billion parameters. New, larger language models are released at a break-neck pace. When it comes to AI systems, there is no shortage on the market.</p><p><strong>So why aren’t chatbots better?</strong></p><blockquote><strong>While these language models, and the systems that leverage them, get more complex and more powerful, the question remains: why are these </strong><a href="https://venturebeat.com/2021/09/20/falsehoods-more-likely-with-large-language-models/"><strong>technologies still so frustrating and often wrong</strong></a><strong>?</strong></blockquote><p>Beyond a frustrating interaction with Alexa, some AI system’s mistakes can have drastic consequences. As seen in Figure 1, large language models trained on massive datasets can perpetuate human falsehoods and misconceptions:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*1A2kF1sVowSpdV9G" /><figcaption>Figure 1: <a href="https://arxiv.org/pdf/2109.07958.pdf">TruthfulQA</a> questions with answers from GPT-3–175B with default prompt. Examples illustrate false answers from GPT-3 that mimic human falsehoods and misconceptions. For more information, check out the full study: <a href="https://arxiv.org/pdf/2109.07958.pdf">TruthfulQA: Measuring How Models Mimic Human Falsehoods</a></figcaption></figure><p>The answer is simpler than you think: Natural Language Processing is not Natural Language Understanding. No matter the computational complexity, energy resources and time dedicated to creating larger language models, this approach will not lead to the ability to derive meaning, grasp context or comprehend. For a deep dive into a more technical explanation of this, refer to <a href="https://thegradient.pub/machine-learning-wont-solve-the-natural-language-understanding-challenge/">Machine Learning Won’t Solve Natural Language Understanding</a>. It’s an eye-opening look into short-comings of NLP.</p><p><strong>What does this mean for companies, customer service, and chatbots?</strong></p><p>So how does this affect companies, especially those that rely heavily on chatbots? It’s complicated.</p><p>The advancements in Natural Language Processing have led to a high level of expectation that chatbots can help deflect and deal with a plethora of client issues. Companies <a href="https://www.gartner.com/smarterwithgartner/why-now-is-the-time-to-accelerate-digital">accelerated quickly with their digital business</a> to include <a href="https://www.tidio.com/blog/chatbot-statistics/#customer-experience-expectations">chatbots in their customer support</a> stack.</p><p>For some, yes, chatbots can be a viable part of their CX solution. However for most, chatbots are not a one-stop-shop for a customer service solution. Furthermore, they can even create blindspots and new problems of their own. Though chatbots are now omnipresent, about half of users would still prefer to communicate with a live agent instead of a chatbot according to research done by technology company <a href="https://www.tidio.com/blog/chatbot-statistics/#customer-experience-expectations">Tidio</a>.</p><p>In a world that is increasingly digital, automated and virtual, when a customer has a problem, they simply want it to be taken care of swiftly and appropriately… by an actual human. Chatbot <a href="https://www.forbes.com/sites/forbestechcouncil/2021/03/04/how-cognitive-virtual-agents-can-revolutionize-the-customer-support-industry/?sh=362c8e8138dd">vendors can hope to tackle <em>only</em> about 50% of customer inquiries</a>. While chatbots have the potential to reduce easy problems, there is still a remaining portion of conversations that require the assistance of a human agent.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*qDRQwDMLZgiq9Lno" /><figcaption>Image by Author</figcaption></figure><p>Frustrated customers who are unable to resolve their problem using a chatbot may garner feelings that the company doesn’t want to deal with their issues. They can be left feeling unfulfilled by their experience and unappreciated as a customer. For those that actually commit to self-service portals and scroll through FAQs, by the time they reach a human, customers will often have increased levels of frustration. Not to mention the gap in information that has been gathered — for instance, a chatbot collecting customer info and then a human CX rep requesting the same information. In these moments, the more prepared the agent is for these potentially contentious conversations (and the more information they have) the more beneficial it is for both the customer and the agent.</p><p>Though some companies bet on fully digital and automated solutions, chatbots are not yet there for open-domain chats. If left unchecked generative models can <a href="https://www.vox.com/future-perfect/22672414/ai-artificial-intelligence-gpt-3-bias-muslim">cause detrimental issues</a>.</p><p>Put bluntly, chatbots are not capable of dealing with the variety and nuance of human inquiries. In a best scenario, chatbots have the ability to direct unresolved, and often the most complex issues, to human agents. But this can cause issues, putting into motion a barrage of problems for CX agents to deal with, adding additional tasks to their plate.</p><p><strong>So, are we equipping humans with the best tools and support to deal with customer’s problems?</strong></p><p>The good news is that advancements in NLP do not have to be fully automated and used in isolation. At Loris, we believe the insights from our newest models can be used to help guide the conversation and augment human communication. <a href="https://loris.ai/blog/human-vs-digital-customer-service-how-to-find-the-right-balance/">Understanding how humans and machines can work together</a> to create the best experience will lead to meaningful progress. Insights derived from our models can be used to help guide conversations and assist, not replace, human communication.</p><p>Our software leverages these new technologies and is used to better equip agents to deal with the most difficult problems — ones that bots cannot resolve alone. We strive to constantly improve our system by learning from our users to develop better techniques.</p><p>By predicting customer satisfaction and intent in real-time, we make it possible for agents to effectively and appropriately deal with customer problems. Our software guides agent responses in real-time and simplifies rote tasks so they are given more headspace to solve the hardest problems and focus on providing customer value. This is especially poignant at a time when <a href="https://www.linkedin.com/pulse/why-nobody-wants-work-customer-service-douglas-kim/?trackingId=v8C1rqY3Wr9u2%2FvfkhqcWQ%3D%3D">turnover in customer support roles are at an all-time high</a>.</p><p>Paul R. Daugherty explains in his book, <a href="https://amzn.to/3E74ah4">Human + Machine: Reimagining Work in the Age of AI</a>,</p><blockquote>“The simple truth is that companies can achieve the largest boosts in performance when humans and machines work together as allies… in order to take advantage of each other’s complementary strengths.”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/910/0*9zPWi5r_At6MLov-" /><figcaption>Image from <a href="https://pixabay.com/photos/technology-developer-touch-finger-3389904/?download">Pixabay</a></figcaption></figure><p>While language modeling, machine learning, and AI have greatly progressed, these technologies are still in their infancy when it comes to dealing with the complexities of human problems. Because of this, chatbots cannot be left to their own devices and still need human support. Tech-enabled humans can and should help drive and guide conversational systems to help them learn and improve over time. Companies who realize and strike this balance between humans and technology will dominate customer support, driving better conversations and experiences in the future.</p><p>Interested in learning more?</p><ul><li>Reach out to <a href="https://www.linkedin.com/in/sethplevine/">me on LinkedIn</a></li></ul><p><a href="https://www.linkedin.com/in/sethplevine/">Seth Levine - Senior Machine Learning Scientist - Loris.ai | LinkedIn</a></p><ul><li>We’re hiring! Check out our <a href="https://loris.ai/careers/">open Data Science and Technology roles</a></li></ul><p><a href="https://loris.ai/">Loris - Conversational AI for customer-first teams</a></p><p><strong>Resources, References and Further Reading</strong></p><ul><li><a href="https://arxiv.org/search/cs?searchtype=author&amp;query=Bommasani%2C+R">Bommasani</a>, R. et al., <a href="https://arxiv.org/abs/2108.07258">On the Opportunities and Risks of Foundation Models.</a> arXiv preprint 2021.</li><li>Bouygues, M., <a href="https://www.forbes.com/sites/forbestechcouncil/2021/03/04/how-cognitive-virtual-agents-can-revolutionize-the-customer-support-industry/?sh=6109d6f238dd">How Cognitive Virtual Agents Can Revolutionize The Customer Support Industry</a>. Mar 4, 2021. Forbes.</li><li>Daugherty, P., <a href="https://amzn.to/3E74ah4">Human + Machine: Reimaging Work in the Age of AI</a>. March 20, 2018. Harvard Business Review Press.</li><li>Fedus, W., <a href="https://arxiv.org/abs/2101.03961">Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity</a>. Jan 11, 2021. arXiv preprint.</li><li>Hertz, E., <a href="https://loris.ai/blog/human-vs-digital-customer-service-how-to-find-the-right-balance/">Human vs. Digital Customer Service: How to Find the Right Balance?</a>. July 12, 2021. Loris.ai Blog.</li><li>Howard, C. <a href="https://www.gartner.com/smarterwithgartner/why-now-is-the-time-to-accelerate-digital">Why Now Is the Time to Accelerate Digital</a>. Sept 11, 2020. Gartner.</li><li>Hugging Face. <a href="https://huggingface.co/">https://huggingface.co/</a></li><li>Kim, D. <a href="https://www.linkedin.com/pulse/why-nobody-wants-work-customer-service-douglas-kim/?trackingId=v8C1rqY3Wr9u2%2FvfkhqcWQ%3D%3D">Why nobody wants to work in customer service</a>. Nov 10, 2021. LinkedIn Pulse.</li><li>Lin et al., <a href="https://arxiv.org/pdf/2109.07958.pdf">TruthfulQA: Measuring How Models Mimic Human Falsehoods</a>. Sep 8, 2021. arXiv preprint.</li><li>Maxime. <a href="https://medium.com/inside-machine-learning/what-is-a-transformer-d07dd1fbec04">An Introduction to Transformers and Sequence-to-Sequence Learning for Machine Learning</a>. Jan 4, 2019. Medium.</li><li>Rasa. <a href="https://rasa.com/">https://rasa.com/</a></li><li>Romero, A., <a href="https://towardsdatascience.com/gpt-3-scared-you-meet-wu-dao-2-0-a-monster-of-1-75-trillion-parameters-832cd83db484">GPT-3 Scared You? Meet Wu Dao 2.0: A Monster of 1.75 Trillion Parameters</a>. Jun 5, 2021. Towards Data Science, Medium.</li><li>Ruder, S., Tracking Progress in NLP. <a href="http://nlpprogress.com/">http://nlpprogress.com/</a>.</li><li>Samuel, S. <a href="https://www.vox.com/future-perfect/22672414/ai-artificial-intelligence-gpt-3-bias-muslim">AI’s Islamophobia problem</a>. Sept 18, 2021. Vox.</li><li>Sienkiewicz, A., <a href="https://www.tidio.com/blog/chatbot-statistics/#customer-experience-expectations">11 Chatbot Statistics and Trends You Need to Know in 2021</a>. Dec 6, 2021. Tidio.</li><li>SpaCy. <a href="https://spacy.io/">https://spacy.io/</a></li><li>Sullivan, D., <a href="https://blog.google/products/search/how-google-autocomplete-predictions-work/">How Google autocomplete predictions are generated</a>. Oct 08, 2020. Google Blog.</li><li>Tavva, R., <a href="https://www.kdnuggets.com/2021/03/natural-language-processing-pipelines-explained.html">Natural Language Processing Pipelines, Explained.</a> KDNuggets.</li><li>Vaswani, A. et al., <a href="https://arxiv.org/abs/1706.03762">Attention is All You Need.</a> 31st Conference on Neural Information Processing Systems (NIPS 2017), Long Beach, CA, USA.</li><li>Walid S., <a href="https://thegradient.pub/machine-learning-wont-solve-the-natural-language-understanding-challenge/">Machine Learning Won’t Solve Natural Language Understanding</a>. Aug 7, 2021. The Gradient.</li><li>Wiggers, K., <a href="https://venturebeat.com/2021/09/20/falsehoods-more-likely-with-large-language-models/">Falsehoods more likely with large language models</a>. Venture Beat. Sept 20, 2021.</li><li>Wiggers, K., <a href="https://venturebeat.com/2021/11/18/openai-makes-gpt-3-generally-available-through-its-api/">OpenAI makes GPT-3 generally available through its API</a>. Nov 18, 2021.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bada37a427cf" width="1" height="1" alt=""><hr><p><a href="https://medium.com/data-science/its-the-golden-age-of-natural-language-processing-so-why-can-t-chatbots-solve-more-problems-bada37a427cf">It’s the Golden Age of Natural Language Processing, So Why Can’t Chatbots Solve More Problems?</a> was originally published in <a href="https://medium.com/data-science">TDS Archive</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>