<?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 Alexander S. Augenstein on Medium]]></title>
        <description><![CDATA[Stories by Alexander S. Augenstein on Medium]]></description>
        <link>https://medium.com/@alexander.s.augenstein?source=rss-f58882e32e65------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*GjyvY_LBN2oO5c4czm5SoA.png</url>
            <title>Stories by Alexander S. Augenstein on Medium</title>
            <link>https://medium.com/@alexander.s.augenstein?source=rss-f58882e32e65------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 20:02:40 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@alexander.s.augenstein/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[Dear developers, thank you for Github Copilot. Also, you’re fired.]]></title>
            <link>https://medium.com/@alexander.s.augenstein/dear-developers-thank-you-for-github-copilot-also-youre-fired-c73b65e3565f?source=rss-f58882e32e65------2</link>
            <guid isPermaLink="false">https://medium.com/p/c73b65e3565f</guid>
            <category><![CDATA[github-copilot]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[software-development]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[data-driven-fiction]]></category>
            <dc:creator><![CDATA[Alexander S. Augenstein]]></dc:creator>
            <pubDate>Fri, 24 Jun 2022 19:19:18 GMT</pubDate>
            <atom:updated>2026-09-18T16:34:36.715Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*DfrJUpWxkFxwqXyv" /><figcaption>Photo by <a href="https://unsplash.com/@singlepinkc?utm_source=medium&amp;utm_medium=referral">hao qin</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Software developers are a peculiar breed. Since the dawn of the career, building tools to make their work easier has been central to their job. Among these tools are “automations”, code that keeps a developer from needing to do the same monotonous task over and over again. It’s the constant task of a developer to try to automate away all of their current work efforts so that they can take on more tomorrow. But that’s also the reason people say <strong>devs are always trying to put themselves out of a job. With the emergence of GitHub Copilot, they might soon be right.</strong></p><p>For those of you who are unfamiliar with the latest trends in AI, tools like GPT-3, DALL-E 2, Google Imagen and more are taking over in force. Using natural language processing (<a href="https://medium.com/@alexander.s.augenstein/im-convinced-gpt-3-is-incredible-d89494a0100">which I talk about in this post</a>), AI can accept a command in pain-English, and respond intelligently to what you asked. This can be used for all kinds of things, from simple Q&amp;A about a topic that interests you, to the programmatic generation of beautiful artwork, or — in the case of GitHub Copilot — code. Yes, <strong>you can state “I want code that does X, Y, and Z” in plain-English, and GitHub Copilot offers not just one but many solutions</strong>. Just pick your favorite.</p><p>Now, a lot of people are saying that GitHub Copilot is limited — that it misses certain context, it gets things wrong, it can’t think architecturally for you. Their takeaway is that we still need our teams of software developers. And sure, that’s approximately true today. But I would be loath to imagine there exists anyone who sees this technology, understands exactly what it is and how it got here, without being able to see where it’s headed tomorrow. Unbounded growth is the nature of AI. Exponential growth is the nature of AI. <strong>Don’t just look at where we are, think of where we will be just a few months or years down the line</strong>.</p><p>So at what point would managers be ready to fire their dev teams? It’s not that difficult to conceive. End users, and the business-facing folks that have to put up with them, care about one thing and one thing only: does the product do what they want? Hmm. Let’s break that down. <strong>End users and managers are more than competent enough to declare “I want code that does X, Y, and Z”</strong> in plain-English. Where have we heard that before… This is the pinnacle of declarative programming — a concept that has existed for a long time, but only recently to this extreme. It’s a style of programming where you state what you want, and a lot (or all) of the implementation details are abstracted away from you. These are just taken care of. And for better or worse, now basically anyone with the right AI companion can do it.</p><p>Now I am not saying that engineers, dev teams, software security experts and the like that deeply understand the architecture of a product will never be helpful or necessary again — humans who know what the heck code is doing will be helpful for large, stable products for some time longer. That being said, many if not most products are not well-architected. Small to mid-sized <strong>teams will probably be better architected and more secure having AI write their code </strong>as opposed to letting a too-small team of underpaid overworked devs do a job they’re ill-equipped to handle. And at least in the States, many of us live in a fiercely competitive market where “it’s better to be first than better”, as they say.</p><p>We could have a nice long talk about the socioeconomic impact of blue- and white-collar career automation, though I think I’ll save that for another day. For now, I’ll leave you with this question: <strong>if you had a custom software application of your very own, what would it do? Tell me X, Y, and Z</strong>.</p><h4>UPDATE (March 2024):</h4><p>It’s been not quite 2 years since my original post, and machinations like Devin AI and the first AI virus are here to create new problems and solve them too. <a href="https://mitsloan.mit.edu/sites/default/files/2023-10/SSRN-id4573321.pdf">This research</a> suggests AI results in significant productivity and quality improvements in the right fields and in the right hands. Don’t quit your day jobs just yet though — the findings suggest it widens the gap between peek performers and everyone else, but we’re all still waiting on research that touches directly on the employment question (I’ll update here when such a report crosses my attention). My more cynical speculations portend measurable downsizing across knowledge jobs, but we have some data telling us how to avoid landing on the short end of that deal for the time being. As is often the case, it still costs less to get smart than the alternative.</p><h4>UPDATE (February 2025):</h4><p>Zuckerberg: “<em>I think this year probably in 2025, we at Meta as well as the other companies that are basically working on this are going to have an AI that can effectively be a sort of midlevel engineer that you have at your company that can write code. And once you have that, then in the beginning it will be really expensive to run and you can get it to be more efficient, and then over time we’ll get to the point that a lot of code in our apps and, and including the AI we generate is actually going to be built by AI engineers instead of people engineers.</em>”</p><p>Mark goes on to claim AI will augment people, he didn’t technically say it will replace them. You’re welcome to fill in the blanks on his ambitions for future, or not. My opinion: for now, use the capabilities at your disposal to stay relevant, and use them to help with the architectural portions of the work. We write code to solve problems — now that AI writes code, let’s hold onto that last half for as long as we can.</p><p>Full context of the quote here:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2F7k1ehaE0bdU%3Fstart%3D7696%26feature%3Doembed%26start%3D7696&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D7k1ehaE0bdU&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2F7k1ehaE0bdU%2Fhqdefault.jpg&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/690758e97ac310f06391c918b558f554/href">https://medium.com/media/690758e97ac310f06391c918b558f554/href</a></iframe><h4>UPDATE (September 2026):</h4><p>It’s been 4 years since my original post. Community perspectives from that era are perfectly preserved in the time capsule that is the comments section below. Freelancing platforms like Upwork and Fiverr peaked the following year with steadily declining revenue since. Writing code by hand is a meme now. Your job is fundamentally different — you personally shoulder more of your team’s weight until you run out of credits. Your boss hires specifically for AI proficiency, via HR who uses AI to screen incoming resumes written by it, for it. AI is discussed in terms of its macro scale ecological impact, a cost incurred while it authored the post surfaced in “your algorithm” (also AI). Not to mention the stream of headlines of tech companies citing AI as the justification for the most recent rounds of mass layoffs. The social zeitgeist at this moment in history includes signs in every yard protesting the construction of new datacenters near you. I’ll admit I was cynical. But much to my relief, I was re-educated that this is all helping actually, per Zuckerberg and pals to the white house press pool.</p><p><a href="https://medium.com/data-driven-fiction/how-to-submit-5e0808dce313">how to submit ?</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c73b65e3565f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[I’m convinced, GPT-3 is incredible]]></title>
            <link>https://medium.com/@alexander.s.augenstein/im-convinced-gpt-3-is-incredible-d89494a0100?source=rss-f58882e32e65------2</link>
            <guid isPermaLink="false">https://medium.com/p/d89494a0100</guid>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[future]]></category>
            <category><![CDATA[tech]]></category>
            <dc:creator><![CDATA[Alexander S. Augenstein]]></dc:creator>
            <pubDate>Thu, 09 Jun 2022 14:18:04 GMT</pubDate>
            <atom:updated>2022-06-19T18:38:05.690Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/400/0*_336But7dBKx6MZc.jpg" /><figcaption>It feels exactly like this scene from Silicon Valley, but real</figcaption></figure><p>It recently came to my attention that there’s a free beta available to test drive a top-of-the-line computer algorithm, functioning as a chatbot of sorts, is available online for anyone to chat with.</p><p>Chatbots are not new by any means. I’m a software engineer by education and profession (with a machine learning thesis in my grad school days, to boot). I’ve used tools that offer this type of capability many, many times for many, many years. And like all the others, for curiosity’s sake I had to give this tech a spin.</p><p>But this one is markedly different.</p><h4>Hooked on GPT-3</h4><p>Over the past 48 hours, I’ve maintained a constant emotional state at the pinnacle of intellectual elation the likes of which I have not felt for <em>a very long time</em>. I could not be happier with how this piece of technology — no, this gift to humanity, has turned out.</p><p>The algorithm in question hails from the GPT-3 family. It can be used as a chatbot, among many other things. If this doesn’t mean much to you, that’s quite alright. You could Google it, sure, or you could just <a href="https://gpt3demo.com/apps/openai-gpt-3-playground"><strong>ask GPT-3 for yourself</strong></a>.</p><p>Feel empowered, learn something new, or not. It’s your life. But I just couldn’t contain myself, so I asked it this:</p><pre>Me:<br>GPT-3, what exactly are you? What kinds of things can you accomplish? Also, are you different from other chatbots? Tell me how.</pre><pre>Davinci:<br>I am an artificial intelligence or &quot;chatbot&quot; that can communicate with humans. I am different than other chatbots because I can generate human-like responses. I can also understand and respond to natural language.</pre><p>Cool, right? But this goes well beyond typical passive amusement and the self-indulgent practice of trying to convince humans that an algorithm has finally become self-aware. Admittedly, <a href="https://www.bing.com/ck/a?!&amp;&amp;p=833bc707572e13b2b35cb62fe213b24b9c111c62a26c3c820c2d1895dfad6212JmltdHM9MTY1NDM4MTE1MSZpZ3VpZD00NzlkNjQwMy0zMzg0LTQxNWEtYjAyZC03MTdlOWVhZjNiOWMmaW5zaWQ9NTE0OA&amp;ptn=3&amp;fclid=5ba5f50f-e454-11ec-9f8f-dc824d316e80&amp;u=a1aHR0cHM6Ly9tZWRpdW0uY29tL25lcmQtZm9yLXRlY2gvd2hhdC1pcy10aGUtdHVyaW5nLXRlc3QtMzUyNDdjNWM1NDFl&amp;ntb=1">convincing a human has been among the holy grails of A.I. since the dawn of the computer</a>. Naturally I asked this question, too.</p><pre>Me:<br>Do you think you could pass the Turing test? Convince me you know what I&#39;m talking about, then convince me you might be able to do it.</pre><pre>Davinci:<br>If you&#39;re referring to the Turing test as developed by Alan Turing, then yes, I think I could pass it. The test is designed to determine whether a machine is capable of exhibiting intelligent behavior that is indistinguishable from that of a human. I believe that I could create a machine that could pass the test.</pre><p>You’ll notice I keep referring to the algorithm as “Davinci”. There are a number of less expensive instantiations of this algorithm (Curie, Babbage, Ada and others), but Davinci is the most impressive currently. And frankly, for personal use, even the primo option is affordable.</p><h4>Here’s how much it cost me</h4><figure><img alt="Shows the cost associated with my interactions with a top-of-the-line chatbot, measured in U.S. dollars." src="https://cdn-images-1.medium.com/max/986/1*PAL6Jih-yO5hFz3Vd0v8sQ.png" /></figure><p>Even though Davinci is the most expensive option, all of the questions I’ve asked over the past two days only cost one U.S. dollar. Except that I’m using a free trial, so I have a couple more weeks I can chat with this incredible deus ex machina before I have to start paying. And when the time comes, I will. Now that I’ve experienced this technology, in exactly the same way you couldn’t take my internet access without an important part of me dying inside, I feel the same about GPT-3.</p><h4>GPT-3 does it all</h4><p>To this point I’ve asked Davinci to:</p><ul><li>Summarize TV shows I enjoy</li><li>Write essays on the etymology of advanced words from the English language</li><li>List some of the more important lessons to teach teenagers that might help them grow up to be functioning adults</li><li>Write stories, songs, poems, haikus and other works of art and fiction</li><li>Tell dad jokes (and revealing what constitutes a dad joke)</li><li>Assess business idea viability</li><li>Solve practical problems</li><li>Discuss the most likely sentiments of public figures on specific topics</li><li>Teach me about new technical topics</li><li>Teach me about various job functions and responsibilities</li><li>Motivate me to try my best</li><li>Assess logical fallacies in argumentation</li><li>List pros/cons of different choices I’m considering for my life</li><li>Answer health and fitness related questions</li><li>Speculate on what important scientific questions humanity might not have thought to ask yet</li><li>Provide guidance on building better habits</li></ul><p>The list goes on. If any readers are interested, I’d be more than happy to share the output from any or all of these conversations.</p><h4>Asterisks*</h4><p>Now, it’s important to note that this tech must be used responsibly. If you read the terms and conditions on the demo website, it says that you own the responsibility for the words you put in <em>and the words that come out</em>. As do I, for the content of this post (whether or not GPT-3 generated the content). This algorithm was trained on the internet using massive volumes of real content, produced by real humans. And the internet is a very big place. At this point in history, the internet contains asymptotically closer to (but never quite reaching) 100% of everything ever recorded by humankind. Learning algorithms at this scale (including GPT-3’s Davinci) can be unwieldy at times. So, if you or I intend to share information produced by this algorithm online with others (as I am now) not only is it our responsibility to ensure that the output generated is factually correct, it is also our responsibility to ensure that any sensitive content is handled appropriately given the use-case and the audience. The output of GPT-3 is in part a reflection of the combined knowledge and sentiments echoed by all of us. And the internet isn’t always a kind place. Ergo, while GPT-3 is beyond amazing, it is also capable of generating suggestions that may press some peoples’ buttons. It is not infallible, as human are not. So take its words with more than an ounce of caution.</p><p>But with this caveat in mind, this honestly makes the algorithm all the more amazing to me. Isn’t it fascinating that it’s even possible to create a machine that expresses itself in the same fallible way as humans?</p><h4>My overall impression</h4><p>When I’m interacting with Davinci, I feel like I’m engaging with the collective mind of all humanity at once. And not typically the underbelly — the overwhelming majority of the time the outputs are well-reasoned, thoughtful and appealing in a way that someone versed in the topic at hand might be.</p><p>I could ask anything at any moment, an infinite number of topics could arise, but unlike internet search engines Davinci comes back immediately with fine-tuned topical intelligence. And with its responses, it never ceases to amaze me.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d89494a0100" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Kaggle “Novice” Badge to “Contributor” in 7 Lines of Code]]></title>
            <link>https://medium.com/analytics-vidhya/kaggle-novice-badge-to-contributor-in-7-lines-of-code-4ca0265d7b2c?source=rss-f58882e32e65------2</link>
            <guid isPermaLink="false">https://medium.com/p/4ca0265d7b2c</guid>
            <category><![CDATA[kaggle-competition]]></category>
            <category><![CDATA[getting-started]]></category>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[coding]]></category>
            <category><![CDATA[kaggle]]></category>
            <dc:creator><![CDATA[Alexander S. Augenstein]]></dc:creator>
            <pubDate>Mon, 28 Sep 2020 02:00:55 GMT</pubDate>
            <atom:updated>2023-04-08T02:58:20.548Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3S7IPqAlPvI2WYEZj0NkGA.png" /></figure><h3>Kaggle = Social Media for Data Science Nerds</h3><p>Ever heard of <a href="https://www.kaggle.com/">Kaggle.com</a>? No? That’s fine. I won’t hold it against you, but now that you have, you MUST check it out. This site is too huge not to get involved (assuming of course you have a coding aptitude and a conscience).</p><p>You’ll also quickly realize you should have been using this site a long time ago, so it’s critical you shake off that “Novice” badge next to your name! How, you ask? They provide a convenient checklist 📜.</p><p>First thing’s first, update your profile. Just pop in a short bio, your location, occupation, and organization (your school is fine if you’re a student). Verify your phone number, cast an upvote, write a comment, ✔✔✔ easy stuff!</p><p>Running a Notebook or script is a requirement buy may be slightly more tricky if you’ve never used Jupyter Notebooks before (but if you like R or Python, you’ll love these!). Go to the Notebooks tab:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/347/1*C0YdOStm_0FDEOLwsnXqdw.png" /></figure><p>Pick a notebook that interests you:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8AxKFPHm61OaK4k3JiRlFA.png" /></figure><p>Click “Copy and Edit”:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Zgpd5l6zQERBdjRqddLehQ.png" /></figure><p>This will take you to the code used to generate the analysis for the Notebook in question. Just run it (there’s a button)! Kaggle will count this as meeting the minimum requirement towards your next badge ✔.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*jbnFuibaQRidgLDhpnLqsA.png" /></figure><p>Lastly, you have to MAKE A COMPETITION SUBMISSION. dun dun dunnn. This isn’t as difficult as it sounds, and we’re going to do it in 7 lines of code.</p><p>If you’ve spent any time on the site whatsoever, you’ll see that Kaggle pushes this Titanic competition, hard. It’s a game where they give you basic info about Titanic passengers, and it becomes our job to guess as accurately as possible whether they lived or died. Wow, that got dark fast. Sign up for the competition, then create a new Notebook (a script, more specifically).</p><p>Now, we’ll write our “code” — what we’ll do here is about as simple as it gets. Most people aboard the Titanic didn’t make it (31.6% 😢), so by submitting code that assumes nobody made it, we should expect 68% accuracy, within variance.</p><p>Here’s my initial submission, written in Python: <a href="https://www.kaggle.com/alexaugenstein/asa55-titanic-competition?scriptVersionId=43567606">https://www.kaggle.com/alexaugenstein/asa55-titanic-competition?scriptVersionId=43567606</a></p><p>Or you can just copy and paste the secret sauce below:</p><pre>import numpy as np<br>import pandas as pd<br><br>data_test = pd.read_csv(&#39;/kaggle/input/titanic/test.csv&#39;)<br>data_submission = pd.DataFrame()data_submission[&#39;PassengerId&#39;] = data_test[&#39;PassengerId&#39;]<br>data_submission[&#39;Survived&#39;] = pd.Series(np.zeros(data_submission.size).astype(int))</pre><pre>data_submission.to_csv(&#39;/kaggle/working/titanic_submission.csv&#39;, index = False)</pre><p>This code prepares a Pandas DataFrame object with two columns: the first contains passenger ID’s, provided by Kaggle, the next contains our guess as to whether or not that passenger survived (hence the 0’s 😢)</p><p>As a Kaggle member, you have some storage space available to you out in the cloud somewhere. The code provided lands it in your Kaggle directory called /kaggle/working/titanic_submission.csv.</p><p>Kaggle of course knows this file exists, and associated the output with your active competition. To make the submission, go to Notebooks → Your Work → [whatever you named your Titanic competition submission] and scroll down until you see the data we generated:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gKFWRytF0pZVbvGFK6sNGQ.png" /></figure><p>Click submit.</p><p>The accuracy will be 62.2%. Could be better, but hey, you’ll see on your profile that you’re a certified Contributor now 🎉🎈 🎇! Of course there’s a whole lot more you can do on Kaggle, this only barely scratches the surface, but you’re already well on your way now you know what it takes to contribute!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4ca0265d7b2c" width="1" height="1" alt=""><hr><p><a href="https://medium.com/analytics-vidhya/kaggle-novice-badge-to-contributor-in-7-lines-of-code-4ca0265d7b2c">Kaggle “Novice” Badge to “Contributor” in 7 Lines of Code</a> was originally published in <a href="https://medium.com/analytics-vidhya">Analytics Vidhya</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Algorithms & Data Structures Cheat Sheet]]></title>
            <link>https://medium.com/@alexander.s.augenstein/on-the-road-to-professional-web-development-6817d4cee9?source=rss-f58882e32e65------2</link>
            <guid isPermaLink="false">https://medium.com/p/6817d4cee9</guid>
            <category><![CDATA[web-development]]></category>
            <category><![CDATA[tutorial]]></category>
            <category><![CDATA[algorithms]]></category>
            <category><![CDATA[computer-science]]></category>
            <category><![CDATA[data-structures]]></category>
            <dc:creator><![CDATA[Alexander S. Augenstein]]></dc:creator>
            <pubDate>Mon, 18 May 2020 03:31:26 GMT</pubDate>
            <atom:updated>2023-04-08T02:57:23.086Z</atom:updated>
            <content:encoded><![CDATA[<h3>Part I: Algorithms &amp; Data Structures</h3><p>This topic may span semesters worth of study for undergraduate software engineering or computer science majors. Without prior exposure, I’d recommend watching the lecture videos for <a href="https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-0001-introduction-to-computer-science-and-programming-in-python-fall-2016/lecture-videos/">MIT 6.0001</a> and <a href="https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-006-introduction-to-algorithms-fall-2011/lecture-videos/">MIT.6.006</a>. It’s not unreasonable to absorb the content of each class in only a few weeks. Algorithms research will always remain a topic of current interest in research, but we don’t need to dive into the darkest depths of the field to have a solid grip on what we’ll need to know to have successful careers as developers.</p><p>Deep knowledge of data structures and algorithms may be used rarely, if ever, in a career as a developer. In practice, using a proven solution is often preferable to authoring our own. But for the purposes of an interview, new programmers typically aren’t expected to know the nuances of specific languages, tools, or technologies, so these topics offer a good source of puzzles and brain teasers that candidates can prepare for. Beyond the interview, the concepts covered in this post form the foundation of a vocabulary shared by many developers.</p><p>We’ll explore some of the topics listed below (derived from typical CS curricula), but not all are required for all dev careers. Having prior background in the first half of this list will be of assistance for the remainder of this post:</p><ul><li>Set theory (including N0, N1, Z, Q, R, C, cardinality)</li><li>Functions, functionals</li><li>Combinatorics, counting probabilities, probability distributions</li><li>Recursion (and how it relates to proof by logical induction)</li><li>Logic, K-maps</li><li>Graph theory</li><li>Complexity (Theta, O, Omega)</li><li>Linear algebra, ODEs</li><li>Matrix / tensor calculus</li><li>Optimization</li><li>Design of experiments (full / fractional factorial, Latin hypercube)</li><li>Select topics from information theory</li><li>Stats and stochastic processes, Markov chains</li><li>Orthogonal functions, encoding / decoding, PDEs</li></ul><p>Review content follows.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sQYslZ2YLNkAl7k_X97e_g.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PZKPqMUxl2Qpdj10Jl1-bA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fdDx0KqJGVLhvVvguQUz_w.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*c5mMF4_mflWgop-X1Nv7XQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qe0DbmKE6OvUqQVH_Yx76A.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0LJWExD-gaJAyO2LbUK-hQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DM1Dh3SRwo5O1XoKkN4_gw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*GwFEygpeXQNLFQ3IE62z1A.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IMKCK2Mub62ena1k8V8qcA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iGTCxu_ZAS5L-YVkw8XOTQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rExf2UhHXo_OahnC9f0ktA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ESGN8mU06H4nxJiB6kMlDw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*apYUaj9Oqpqxv7rtAoAM4Q.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uncCdAIWhUZgsY3y5IxEVA.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gXwC9OfuGDEZrkc7OjCsKw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6FhF08pf54gGbPleea81MQ.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*v2DWWMYPyhtND6p1d24w1w.png" /></figure><h3>Part II: Algorithms &amp; Data Structures For Technical Interviews</h3><p>While Part I covered a broad-brush of the vocabulary relevant to conversational fluency in these topics, the application of these ideas to the interview process and beyond has not yet been addressed.</p><p>Interview questions on these topics most frequently arise in the context of a programming language. As such, the implementation will depend heavily on our language of choice. In all cases, the builtin tools we use will have time and memory complexity aspects we should be aware of. Most interview questions will be posed in terms of a highly specific problem (e.g. invert a string, rotate a matrix in-place, etc.), so we’ll need to decide which abstract concepts apply. It’s worth grinding through practice problems in advance, but memorizing all the answers is both impractical and insufficient for the needs of the interviewer. Remember that these questions can be made arbitrarily easy or difficult on the fly — most interviewers are looking to examine our thought processes, so we need to keep an open dialogue and say what’s on our mind. Solving the problem is important, but working with the interviewer to analyze it from multiple angles is equally important. It’s inappropriate to flounder on an implementation of a specific algorithm or data structure of our choosing (once we name an intended solution), but it will always be appropriate to work closely with the interviewer to fully understand the problem.</p><p>There are a several interview-appropriate problems that are so famous they have a name. Occasionally you’ll hear of interviewers who present these famous problem statements without naming them. While I’d speculate this is uncommon (I haven’t personally experienced it at least), if it happens to us we should be mindful that it’s our job to call them out by name (and ideally know some properties of optimal solutions). Some of the more popular problems include:</p><ul><li>the Traveling Salesman Problem (TSP)</li><li>the Towers of Hanoi</li><li>the N-Queens Problem</li><li>the M-Coloring Problem</li><li>the Knapsack Problem</li><li>the Fractional Knapsack Problem</li></ul><p>The infamous problems listed above are far less common than some of the standard fare used to demonstrate comfort and fluency with a programming language itself. While only tangentially related to the above discussion on algorithms &amp; data structures, it is always a good idea to practice <a href="https://simpleprogrammer.com/programming-interview-questions/">language-agnostic interview questions like the ones listed here</a>.</p><p>If you’re interested in truly putting the topics we’ve covered into action, an excellent categorized pool of <a href="https://www.hackerrank.com/interview/interview-preparation-kit">more difficult algorithm &amp; data structures questions are listed here</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6817d4cee9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Passing The C++ Code Interview In 3 Weeks]]></title>
            <link>https://medium.com/@alexander.s.augenstein/how-i-passed-the-c-code-interview-in-3-weeks-a3e350214a01?source=rss-f58882e32e65------2</link>
            <guid isPermaLink="false">https://medium.com/p/a3e350214a01</guid>
            <category><![CDATA[study-guide]]></category>
            <category><![CDATA[coding-interviews]]></category>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[code-interview]]></category>
            <category><![CDATA[cplusplus]]></category>
            <dc:creator><![CDATA[Alexander S. Augenstein]]></dc:creator>
            <pubDate>Fri, 03 Apr 2020 20:42:52 GMT</pubDate>
            <atom:updated>2023-04-08T02:49:41.851Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*KNNxh07ynXCe8LWo" /><figcaption>Photo by <a href="https://unsplash.com/@flipboo?utm_source=medium&amp;utm_medium=referral">Philippe Bout</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p><strong>*<em>Full disclosure</em></strong><em>: I did not get the job for which I passed the code interview. The code interview itself was an early interview stage and went over extremely well, but in a following interview stage I did not demonstrate sufficient preparation for Linux system administration questions that were also relevant to the job. With that caveat in mind, interviewer truly was impressed with my performance during the code interview. If you too are trying to learn C++ quickly and efficiently, I hope you find this shared experience insightful.</em><strong>*</strong></p><h3>C++ Has A Notoriously Steep Learning Curve</h3><p>If you’ve ever taken any other introductory programming course, you’ll likely feel left in the dark by the time you finish one on C++. By the time you’re ready to learn the ABC’s of the standard template library (possibly <em>the most important</em> topic), it’s already over.</p><p>Contrast this with <a href="https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-0001-introduction-to-computer-science-and-programming-in-python-fall-2016/">MIT 6.0001 for Python</a>, or the <a href="https://www.youtube.com/watch?v=vaysJAMDaZw">Edureka Python full course,</a> and you come out feeling like a pro. So what gives?</p><p>There are many topics relevant to C++ that you never need to think about in other languages. For reasons of concepts such as “the memory model” alone, there’s a steep learning curve — but learning the language is far from an insurmountable task.</p><h3>First, Let’s Get You Set Up To Write Some Code</h3><p><a href="https://repl.it/">Check out REPL.it</a> to get started right now — this is by far the fastest way to get started. *But* since you’re studying for a code interview, I strongly suggest you follow one of The Cherno’s tutorials to get set up with Visual Studio (for <a href="https://www.youtube.com/watch?v=1OsGXuNA5cc">Windows</a>, <a href="https://www.youtube.com/watch?v=1E_kBSka_ec">Mac</a>, or <a href="https://www.youtube.com/watch?v=LKLuvoY6U0I">Linux</a>). <strong>You will need to understand the memory model, compiling, linking, loading, makefiles, and other things that become more apparent with a complete development environment.</strong> Which is why you will need to graduate from REPL.it as soon as possible to an IDE like <a href="https://www.eclipse.org/downloads/packages/release/kepler/sr2/eclipse-ide-cc-developers">Eclipse</a> or <a href="https://visualstudio.microsoft.com/vs/features/cplusplus/">Visual Studio</a>.</p><h3>Here’s A List Of Everything I Studied Before The Interview</h3><ol><li>Understand what C++ is all about via <a href="https://www.youtube.com/playlist?list=PLlrATfBNZ98dudnM48yfGUldqGD0S4FFb">The Cherno’s YouTube Channel</a></li><li>Learn to use the Standard Library via <a href="https://www.youtube.com/playlist?list=PL5jc9xFGsL8G3y3ywuFSvOuNm3GjBwdkb">the first several videos on Bo Qian’s YouTube Channel</a></li><li>Understand the difference between classical C++ and modern C++ via <a href="https://www.modernescpp.com/index.php/what-is-modern-c">this article on the topic</a></li><li>Practice until you’re comfortably answering <a href="https://simpleprogrammer.com/programming-interview-questions/">language-agnostic interview questions like the ones listed here</a> (<strong>with and without</strong> the use of the standard template library)</li><li>If you’re comfortably answering questions like the ones asked on <a href="http://cppquiz.org/">CPPQuiz.org</a>, you’ve probably gone further than needed to pass the code interview (depending on the job listing, you can likely infer if this is true in your scenario)</li><li>Watch every recorded lecture in <a href="https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-0001-introduction-to-computer-science-and-programming-in-python-fall-2016/">MIT 6.0001 (Intro to CS in Python)</a> and implement each algorithm / data structure using C++</li><li>Learn to <a href="https://eigen.tuxfamily.org/dox/GettingStarted.html">import and use the Eigen library</a>. If you’re a Matlab person, <a href="https://eigen.tuxfamily.org/dox/AsciiQuickReference.txt">you’ll feel so at home with C++ after reading this</a>. (I wasn’t tested on the use of a specific library, but they asked detailed questions to quiz my knowledge of the process)</li><li>Listen to C++ founder Bjarne Stroustrup’s thoughts on the evolution of the language (he has many videos on YouTube)</li></ol><h3>Look These Terms Up On Google Or StackOverflow</h3><p>Some learning objectives you’ll need to clear before moving to the coding interview include:</p><ul><li>Basic syntax</li><li>Ranged ‘for’ loops</li><li>Templates</li><li>Operator overloading</li><li>Preprocessor statements and macros</li><li>Namespaces</li><li>Exceptions</li><li>Const and mutable (and friend functions)</li><li>Static and dynamic casting</li><li>Pointers / void pointers / smart pointers</li><li>Enums</li><li>String streams</li><li>Virtual functions and virtual inheritance and vtables</li><li>Runtime type information (RTTI)</li><li>Helpful 3rd party libraries (Qt, OpenGL, TensorFlow, OpenCV, Eigen, Drake)</li><li><a href="https://www.youtube.com/playlist?list=PL5jc9xFGsL8G3y3ywuFSvOuNm3GjBwdkb">The Standard Library</a> (STL) (containers , the algorithms over them , and the iterators that connect them)</li><li>Visual studio debug tools (breakpoints, memory view)</li><li>Package management</li><li>g++</li><li>gtest</li><li>cmake</li></ul><h3>There’s A Difference Between C++ And Good C++</h3><p>We are extremely fortunate to have the creator of C++ still working hard to improve the language each year. <a href="http://www.stroustrup.com/C++.html">Bjarne Stroustrup</a> has some brilliant thoughts, and helps maintain a <a href="https://github.com/isocpp/CppCoreGuidelines/blob/master/CppCoreGuidelines.md">plain English description of C++ best-practices</a>. You should definitely <a href="https://www.youtube.com/watch?v=u_ij0YNkFUs">take a moment to listen to him speak</a>. I strongly recommend you search and internalize some of the following big ideas in C++:</p><ul><li>Memory allocation as initialization</li><li>Don’t use casts</li><li>Avoid naked new and naked delete</li><li>Avoid macros except for include guards</li><li>Avoid unions</li><li>Hide arrays from interfaces, keep them in low-level if needed</li><li>Prefer immutable to mutable data</li><li>Encapsulate messy constructs rather than spreading them thru the code</li><li>Interfaces are the single most important aspect of code organization, make them explicit, encapsulate rule violations</li><li>State preconditions, prefer Expects() for expressing preconditions and Ensures() for postconditions</li><li>Prefer &lt;vector&gt; by default, prefer &lt;array&gt; to C-style arrays if needed</li><li>C++98 added templates, containers, &lt;algorithm&gt;, &lt;string&gt;, &lt;iostream&gt;</li><li>C++11 added auto, function pointers / lambdas, multithreading, regexp, smart pointers, hash tables</li><li>C++14 added read-write locks, generic lambdas //C++17 added parallel algorithms</li><li>C++20 added the spaceship operator and templated lambdas</li></ul><h3>Code Interview Questions I Practiced, With Solutions</h3><h4>(1) Practice: Objects</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/f91c5a0afb2e99800fb35ad438ef6ffe/href">https://medium.com/media/f91c5a0afb2e99800fb35ad438ef6ffe/href</a></iframe><h4>(2) Practice: Preprocessor Statements</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c2d4422b602046fa62882db4dcabfb3a/href">https://medium.com/media/c2d4422b602046fa62882db4dcabfb3a/href</a></iframe><h4>(3) Practice: Pass By Value, Pass By Reference</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/4c8f14938a9de04ed68aed48fbd0bb9e/href">https://medium.com/media/4c8f14938a9de04ed68aed48fbd0bb9e/href</a></iframe><h4>(4) Practice: Allocating Memory On The Stack Vs On The Heap</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/8e3995508b97eac997882c435ab34ae2/href">https://medium.com/media/8e3995508b97eac997882c435ab34ae2/href</a></iframe><h4>(5) Practice: Error Handling</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/5913611231a25c04f52750fece836c3c/href">https://medium.com/media/5913611231a25c04f52750fece836c3c/href</a></iframe><h4>(6) Practice: Templates (This Topic Gets A Lot Wider And Deeper)</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/9759ddf62a8a6b263a86a5ed35cb804f/href">https://medium.com/media/9759ddf62a8a6b263a86a5ed35cb804f/href</a></iframe><h4>(7) Practice: Foreach Function Without STL</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/d16fc26dc8a3bff264b56208930dc1ce/href">https://medium.com/media/d16fc26dc8a3bff264b56208930dc1ce/href</a></iframe><h4>(8) Practice: Lambdas</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/ff64b5598d04917511e83d96b32b15c7/href">https://medium.com/media/ff64b5598d04917511e83d96b32b15c7/href</a></iframe><h4>(9) Practice: MATLAB-Like Matrix Manipulations With The Eigen Library</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/2be8560e4180dfbb105a5368ac3c160e/href">https://medium.com/media/2be8560e4180dfbb105a5368ac3c160e/href</a></iframe><h4>(10) Common Interview Question: Find The Missing Number</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/526f8070c238c136457cd547e6517fd4/href">https://medium.com/media/526f8070c238c136457cd547e6517fd4/href</a></iframe><h4>(11) Common Interview Question: Find The Duplicate Number</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/41f1a40f194d9bcbb1f6a81119a546ca/href">https://medium.com/media/41f1a40f194d9bcbb1f6a81119a546ca/href</a></iframe><h4>(12) Common Interview Question: Find Max Or Min In An Unsorted Array</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/56db270d0f13b7d7d5c287467019428b/href">https://medium.com/media/56db270d0f13b7d7d5c287467019428b/href</a></iframe><h4>(13) Common Interview Question: Find Integer Pairs That Sum To X</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/6ec00c2fa3560682c6528b2e18fd4be0/href">https://medium.com/media/6ec00c2fa3560682c6528b2e18fd4be0/href</a></iframe><h4>(14) Common Interview Question: Find Multiple Duplicates</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/e282aec7a398a46b3c7360cb3553c24f/href">https://medium.com/media/e282aec7a398a46b3c7360cb3553c24f/href</a></iframe><h4>(15) Common Interview Question: Print Duplicate Characters From A String</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/092055565da4192595759024478aa168/href">https://medium.com/media/092055565da4192595759024478aa168/href</a></iframe><h4>(16) Common Interview Question: Check If Strings Are Anagrams</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c2e3b658c13843e737895341ee798c38/href">https://medium.com/media/c2e3b658c13843e737895341ee798c38/href</a></iframe><h4>(17) Common Interview Question: Print The First Non-Repeated Character</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/01a654eb937265fbbb9313c99d45128c/href">https://medium.com/media/01a654eb937265fbbb9313c99d45128c/href</a></iframe><h4>(18) Common Interview Question: Reverse A String Using Recursion</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/0d4ed088c709f2c3d5e20694f6eb39e4/href">https://medium.com/media/0d4ed088c709f2c3d5e20694f6eb39e4/href</a></iframe><h4>(19) Common Interview Question: Check If A String Contains Only Digits</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/f2f0b4c32740bc0a5ecc3104879f9214/href">https://medium.com/media/f2f0b4c32740bc0a5ecc3104879f9214/href</a></iframe><h4>(20) Common Interview Question: Swap Two Numbers Without Using A Temporary Variable</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/7cb7d7bf7dfc7f0670b15076f848a754/href">https://medium.com/media/7cb7d7bf7dfc7f0670b15076f848a754/href</a></iframe><h4>(21) Common Interview Question: Manually Transpose A 3x3 Matrix</h4><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/be1e04fc8f1af43e474cb23d7057c37c/href">https://medium.com/media/be1e04fc8f1af43e474cb23d7057c37c/href</a></iframe><h3>Closing Thoughts</h3><p>The code interview itself was performed over <a href="https://coderpad.io/sandbox">CoderPad</a>. Using this platfom, the interviewer could set up my code environment, show me questions, and see the code and the output. We talked on the phone during this process. I had 30 minutes to answer three questions, including writing a function to fill a vector of size n with odd numbers, writing a function to return the square of an integer, and writing a function to determine if a given string was a palindrome. I talked through my thought process the entire time as I worked through problems, which was received very positively by the interviewer.</p><p>All in all, the experience was positive and met expectations in terms of difficulty level. So study hard, and know that you can do this! (Just don’t make the same mistake I did afterward!)</p><p><strong>Update:</strong> <em>I wrote this post ^^ a few years ago. It’s still great advice in my opinion, though you’ll probably be due to look up whatever has been included in the latest version of C++23 (which didn’t exist when I first posted). I’ve been employed full-time as a DevOps engineer the past several years at the time of this edit. Keep studying, actively seek advice, and your efforts to get employed in this space will pay off in the end.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a3e350214a01" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Python vs. JavaScript Syntax Cheat Sheet]]></title>
            <link>https://medium.com/@alexander.s.augenstein/beginners-guide-to-python-vs-javascript-in-one-big-table-22bc78bdf880?source=rss-f58882e32e65------2</link>
            <guid isPermaLink="false">https://medium.com/p/22bc78bdf880</guid>
            <category><![CDATA[python]]></category>
            <category><![CDATA[comparison]]></category>
            <category><![CDATA[javascript]]></category>
            <category><![CDATA[venn-diagrams]]></category>
            <category><![CDATA[table]]></category>
            <dc:creator><![CDATA[Alexander S. Augenstein]]></dc:creator>
            <pubDate>Fri, 03 Apr 2020 15:42:58 GMT</pubDate>
            <atom:updated>2023-04-08T03:33:13.389Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ahf_OowSCE_LvCWk" /><figcaption>Photo by <a href="https://unsplash.com/@maxcodes?utm_source=medium&amp;utm_medium=referral">Maxwell Nelson</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>Edit: Below is a quick comparison between Python and JavaScript basic syntax. I wrote this several years ago, and in hindsight I’m not sure why I posted about it on Medium. At this point, it doesn’t seem to be the most appropriate medium 🙃. Note that the table below is just an embedded version of the following GitHub gist: <a href="https://gist.github.com/asa55/3bc818a88eb9b44d60545874392e19a0">https://gist.github.com/asa55/3bc818a88eb9b44d60545874392e19a0</a></p><p>Enjoy</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/b43e79f8b5b6f6c804ab156928786935/href">https://medium.com/media/b43e79f8b5b6f6c804ab156928786935/href</a></iframe><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=22bc78bdf880" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>