<?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 Lye Jia Jun on Medium]]></title>
        <description><![CDATA[Stories by Lye Jia Jun on Medium]]></description>
        <link>https://medium.com/@lyejiajun?source=rss-9e666ce9a7ab------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*5rXfjHy79q38whe_QO8-MA.png</url>
            <title>Stories by Lye Jia Jun on Medium</title>
            <link>https://medium.com/@lyejiajun?source=rss-9e666ce9a7ab------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Fri, 20 Nov 2026 02:28:27 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@lyejiajun/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[Demystifying LLMs: A Quick Summary to Andrej Karpathy’s Intro to LLM Course]]></title>
            <link>https://pub.towardsai.net/demystifying-llms-a-quick-summary-to-andrej-karpathys-intro-to-llm-course-6816111dc809?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/6816111dc809</guid>
            <category><![CDATA[llm]]></category>
            <category><![CDATA[large-language-models]]></category>
            <category><![CDATA[ai-governance]]></category>
            <category><![CDATA[ai-safety]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Wed, 03 Jan 2024 12:01:26 GMT</pubDate>
            <atom:updated>2024-01-03T12:01:26.830Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*V-uYzff6tmznyZOsll_z6w.jpeg" /></figure><p>ICYMI, <a href="https://karpathy.ai/">Andrej Karpathy</a> (Ex-Senior Director of AI at Tesla, Current OpenAI Technical Member) recently made a concise yet comprehensive video on the basics of Large Language Models (LLM).</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FzjkBMFhNj_g%3Fstart%3D1381%26feature%3Doembed%26start%3D1381&amp;display_name=YouTube&amp;url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DzjkBMFhNj_g&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FzjkBMFhNj_g%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/e81d99d9d363814e803d5d6589fb67c5/href">https://medium.com/media/e81d99d9d363814e803d5d6589fb67c5/href</a></iframe><p>In this article, I’ll <strong>summarize the key takeaways </strong>of Andrej Karpathy’s Intro to LLM course in under 10 minutes.</p><h4>Table of Contents</h4><ul><li><strong>LLM is actually just two files</strong></li><li><strong>LLM neural network (high-level details)</strong></li><li><strong>Fundamentally, LLM will try to imitate (and it is imperfect)</strong></li><li><strong>Building an LLM requires just two steps</strong></li><li><strong>LLM Scaling Law and Why It Matters</strong></li><li><strong>The capabilities of LLM and where it’s going</strong></li><li><strong>LLM Security</strong></li></ul><h3><strong>LLM is actually just two files</strong></h3><p>Fundamentally, to run an LLM, we just need two files</p><ol><li>A parameters file (~140GB)</li><li>A script to run those parameters (it can be written in C or other programming languages)</li></ol><p>No internet connection is required.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/831/1*6SjiR9SIX-QMbypXqkF2LA.png" /></figure><h3>LLM Neural Network (High-Level Details)</h3><h4>The complexity comes from getting the parameters file, and the intuition is that we’re “compressing” the internet</h4><p>Here are some numbers about one of the leading open-source LLM, <a href="https://ai.meta.com/llama/">LLama 2–70b</a> by META</p><ul><li>Meta AI scientists took a chunk of the internet (roughly 10 terabytes or 10,000 gigabytes of data) by crawling and scraping the internet.</li><li>They then procured GPU clusters of roughly 6,000 powerful GPUs and ran the training process for ~12 days</li><li>The total cost of training this model is approximately $2 million.</li><li>We eventually get a “zipped” parameters file of roughly ~140GB<br>(we have a compression ratio of almost 100x, hence the intuition of “compressing” the internet)</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/966/1*CwXGPqnLLg8R4QEkSC9Nzg.png" /></figure><h4>The neural network of LLM simplified: next word prediction task</h4><ul><li>Fundamentally, the neural network of the LLM does one specific task: <strong>next-word prediction.</strong></li><li>Given the context of the first 4 words, “cat sat on a,” <strong>the model predicts that the most likely next word is “mat.”</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/947/1*E84-nyc985AZvzzeLK9jRg.png" /></figure><h4>The next word prediction task forces the neural network to learn about the world</h4><ul><li>Here’s a Wikipedia page of <a href="https://en.wikipedia.org/wiki/Ruth_Handler">Ruth Handler</a>, the inventor of Barbie Doll.</li><li>One great value of the next-word-prediction task is that <strong>it forces the neural network to learn about the world.</strong></li><li>In the underlined red text, we see that there are a lot of facts about Ruth; <strong>by feeding this entire Wikipedia article to the neural network, the network inevitably learns a lot about Ruth, thereby gaining “knowledge.”</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/763/1*o8tOiHBFF8BQkJDFs3tlSA.png" /></figure><h4>Using the neural network to do model inference ⇒ generate next word and re-feeding to model to generate the following word</h4><ul><li>There is an “auto-regressive” nature of the Decoder used in the transformer architecture, which powers the LLM.</li><li>Given a prompt, the LLM guesses the best next word, then fits this best next word back to itself to generate the next following best word, and the process repeats.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*3QGC1x5e_8Z33ZzV.png" /></figure><h3>Fundamentally, an LLM will try to imitate (and it is imperfect)</h3><ul><li>On the left, we see the LLM “dreaming” java code that looks correct, but may not actually work.</li><li>In the middle, we see the LLM “dreaming” Amazon products: the ISBN number may be of the correct length and format, but the corresponding product most likely won’t exist.</li><li>On the right, we see the LLM “dreaming” a Wikipedia article about a fish. Here, even though it is “dreaming,” the information is somehow still largely correct without repeating the text verbatim of the original training data, highlighting that LLM contains “knowledge.”</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/991/1*mPja4rYZ0Ba9FauTH7XLwA.png" /></figure><h4>We know LLM fundamentally uses the Transformer architecture but we still don’t fully know how LLM stores knowledge</h4><ul><li>We can optimize and improve the LLM neural network as a whole, iteratively to make it better at predicting the next word (i.e. better performance for LLM).</li><li>However, we don’t know how individual parameters or neurons collaborate: this is a field called interpretability that is still growing.</li><li>A notable illustration of the peculiar behavior in large language models (LLMs) is the <strong>“reverse curse”</strong> phenomenon. For instance, if we query the LLM about the identity of Tom Cruise’s mother, it can accurately respond with “Mary Lee.”</li><li>Yet, intriguingly, when the question is reversed to inquire who the son of Mary Lee is, the LLM struggles to provide the correct answer, despite having the relevant information.</li><li>This shows how knowledge could be <em>“one-dimensional”</em> for LLM and how LLM stores knowledge is something we do not have full understanding of.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/950/1*_41VRgT_-WJXUopEiejh1g.png" /></figure><h3>Fundamentally, Building an LLM Requires Just Two Steps</h3><p><strong>Step 1: Pretraining (to get an internet document generator)</strong></p><ul><li>As mentioned previously, we first scrape the internet and get a large amount of textual data.</li><li>We’ll use this data to train our neural network to optimize for better performance in the next-word-prediction task.</li><li>We’ll then get a “document generator software” that can spit out text.</li></ul><p>However, the conventional LLM we all see today isn’t merely a document generator software; we want an assistant that is actually useful and could help us with a wide array of tasks rather than spit out useless text.</p><p><strong>Step 2: Finetuning (to get actual AI assistants)</strong></p><ul><li>In step 2, we would hire people and give them precise labeling instructions (i.e., answer this question while ensuring your answer is factual and harmless…, etc.)</li><li>There are also human-machine collaborations, like sampling parts of different LLM answers to form the most effective answer overall to facilitate a more efficient labeling process.</li><li>This process is drastically lower-cost than the pre-training stage, and upon many iterations, we’d obtain our desired assistant LLM model.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/894/1*5A-_dq4TSBIeKw-3ilWWIg.png" /></figure><h3>LLM Scaling Law and Why It Matters</h3><ul><li>The <strong>LLM Scaling Law</strong> refers to the relationships that describe how a system behaves as it scales in size or other relevant parameters.</li><li>In general, the trend seems to <strong>suggest that more parameters and more data are leading to better performance</strong> (thus allowing us to gain more intelligence “for free” without much algorithmic improvement)</li><li>This matters because <strong>improvement in LLM performance transfers over to many other NLP tasks (as proven during the evaluation process)</strong>, thus creating a more useful LLM.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/758/1*1PysJ-yMSjPEqvaZmfen7w.png" /></figure><h3>Capabilities of LLM and Where It’s Going</h3><h4>Current Capability: Tool Use and Multimodality</h4><p>State-of-the-art LLMs today like ChatGPT are already using tools and are multimodal: <strong>they can leverage web search, calculator functions, computer vision, speech-to-text, and text-to-speech features to enhance user experience.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*H5RHmIgLE5aIDbZfgJpiqA.png" /></figure><h4>Future Capability 1: System 2 Thinking</h4><ul><li>In the book, <a href="https://en.wikipedia.org/wiki/Thinking,_Fast_and_Slow">Thinking, Fast and Slow</a>, the concepts of <strong>System 1</strong> and <strong>System 2</strong> types of thinking were mentioned.</li><li>In essence, the <strong>System 1 type of thinking is more intuitive, quick, and automatic; the System 2 type of thinking is slower, more rational, and serves more complex contemplation.</strong></li><li>Currently, LLM only has <strong>System 1 type of thinking. </strong>The dream is to eventually equip LLM with <strong>System 2</strong> <strong>type of thinking, where it could take 15–30 minutes to think about a problem, and iterate through all possible options like a tree of thoughts, before thoughtfully sharing their answer (and thought process) to a question.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/790/1*xhFQ9KdkeLWPOqnqOhOBEA.png" /></figure><h4>Future Capability 2: Self-Improvement</h4><ul><li>How can LLM surpass human intelligence?</li><li>Taking reference from a popular AI model (<a href="https://en.wikipedia.org/wiki/AlphaGo">AlphaGo</a>) developed by Google Deepmind to play the board game Go, we see that we can <strong>train an AI model that surpasses human performance by creating a reward system to reward the AI model whenever it wins a game.</strong></li><li>By playing thousands and thousands of games and tweaking parameters based on the reward function, <strong>AlphaGo eventually defeated the best Go player in the world.</strong></li><li>The dream here is to <strong>find such reward criteria to allow LLM to self-improve at exponential rates.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/929/1*WKEh72qJnUmBddDJNYORfQ.png" /></figure><h4>Future Capability 3: Customized LLM Models</h4><ul><li>There is likely no single reward criteria that can facilitate LLM to self-improve at exponential rates for <strong>all tasks.</strong></li><li>The future of LLM, however, could be having a series of specialized LLMs that excel at their niche: <strong>the launch of the ChatGPT app store is a testament to OpenAI moving towards such a vision.</strong></li><li>In the future, <strong>LLMs may even have the opportunities to communicate with each other to leverage each other’s strengths and niches.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/947/1*4yDV61VHeZS4A3__gkyebw.png" /></figure><h3><strong>LLM Security</strong></h3><h4>Jailbreaking</h4><ul><li>In the context of Large Language Models (LLMs), “jailbreaking” is the <strong>process of carefully engineering prompts to exploit model biases and generate outputs that may not align with their intended purpose.</strong></li><li>For example, asking LLM how to create dangerous weapons.</li><li>While LLMs would typically be designed with some security mechanisms, <strong>attackers are constantly finding new ways to override them.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/698/1*m7aSvhCy1CXbybLcwn4a8Q.png" /></figure><h4>Prompt Injection</h4><ul><li>Prompt injections involve <strong>bypassing filters or manipulating the LLM using carefully crafted prompts that make the model ignore previous instructions or perform unintended actions.</strong></li><li>These vulnerabilities can lead to unintended consequences, including data leakage, unauthorized access, or other security breaches.</li><li>One example of an attack is when <strong>Bing chat shows a fraud link, which is generated because a Bing search result contains a prompt injection attack that instructed Bing to show the fraud link.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/692/1*ozbyf9hSOqMO9Ip0vzCJKg.png" /></figure><h4>Data Poisoning and Backdoor Attack</h4><ul><li>Data Poisoning refers to <strong>intentionally manipulating the training data of a machine-learning model to influence its future behavior</strong></li><li>Used with data poisoning, a backdoor attack refers to how the attacker introduces <strong>a specific trigger into the training data.</strong> When this trigger is present in a prompt or query, <strong>it causes the model to produce a predetermined response or behavior.</strong></li><li>An undesirable situation could be that if ML engineers scrape the internet, which contains such malicious data, <strong>LLMs could be susceptible to these backdoor attacks.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/907/1*7VyvmzWNXDKWqSGS4fYZnQ.png" /></figure><p>And… that’s it! This is the quick summary of Andrej Karpathy’s Introduction to LLM course.</p><p>I hope you enjoyed it and found some value. 🚀</p><p>Cheers, and I’ll catch you in the next article!</p><p><em>I am a computing undergraduate in Singapore actively exploring the space of AI Safety, Startups, and Venture Capital. I write about </em><a href="https://www.linkedin.com/pulse/introduction-ai-governance-journal-lye-jia-jun/"><em>AI Governance</em></a><em>, technical stuff, and productivity matters. If these interests you, do </em><a href="https://www.linkedin.com/in/lyejiajun/"><em>follow me</em></a><em> for more insightful pieces. Cheers!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6816111dc809" width="1" height="1" alt=""><hr><p><a href="https://pub.towardsai.net/demystifying-llms-a-quick-summary-to-andrej-karpathys-intro-to-llm-course-6816111dc809">Demystifying LLMs: A Quick Summary to Andrej Karpathy’s Intro to LLM Course</a> was originally published in <a href="https://pub.towardsai.net/">Towards AI</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Comprehensive Guide to Stakeholder Analysis in AI Governance (Part 2)]]></title>
            <link>https://pub.towardsai.net/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-2-6b9039649317?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/6b9039649317</guid>
            <category><![CDATA[stakeholder]]></category>
            <category><![CDATA[ai-governance]]></category>
            <category><![CDATA[the-ai-governance-journal]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ai-ethics]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Fri, 01 Dec 2023 22:01:47 GMT</pubDate>
            <atom:updated>2023-12-01T22:01:47.830Z</atom:updated>
            <content:encoded><![CDATA[<h4>Understanding Interests, Influences, and Impacts for Key Stakeholders in AI Governance</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5RVUMGwuFal7m--ykvc_rQ.png" /></figure><p><em>*Author Note: This article is written as part two of the Comprehensive Guide to Stakeholder Analysis in AI Governance. Read Part 1 </em><a href="https://pub.towardsai.net/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-1-b8a464755f9f"><em>here.</em></a></p><p>Welcome back to the continuation of our<strong> ‘A Comprehensive Guide to Stakeholder Analysis in AI Governance.’</strong> If you’re just tuning in, <strong>I highly recommend starting with the </strong><a href="https://pub.towardsai.net/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-1-b8a464755f9f"><strong>first part</strong></a><strong> of this two-part series </strong>to understand the <strong>key stakeholders</strong> we’ve identified in the space of AI Governance.</p><p><strong><em>🚀 Stay tuned to the end of this article for practical, actionable takeaways from this comprehensive AI Governance stakeholder analysis guide!</em></strong></p><h3>Brief Recap of Part 1 of The Guide</h3><h4>In Part 1, we explored the following</h4><ol><li>The foundations of stakeholder analysis, its significance, and methodologies.</li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/630/0*OKXVlhjewQ0bPtdS.png" /></figure><p>2. The SCIM framework for the <strong>Stakeholder Analysis Process.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/464/0*vfw--pceUl_RyQzx.png" /></figure><p>3. Stakeholder Analysis application in AI Governance by identifying the key stakeholders in the space, step “S” in the SCIM framework.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/630/0*Fv7A5NUgKRmGXFg0.png" /></figure><h4>In Part 2, we will continue with the subsequent steps of the stakeholder analysis process</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/464/1*XzeY4eR1CAF5Mie3SSGi1w.png" /></figure><ul><li>With our identification of the key stakeholders in AI Governance, <br>we will learn how to <strong>analyze, categorize, and prioritize these stakeholders</strong>.</li><li>Beyond that, we’ll delve into the methods to <strong>develop and implement effective engagement strategies</strong>, continuously monitor these relationships, and make adjustments as needed.</li><li>Our guide will conclude with a comprehensive wrap-up, emphasizing the key <strong><em>ACTIONABLE </em>takeaways you can apply in the field of AI governance.</strong> Ready to dive deeper? Let’s embark on this next phase.</li></ul><h3>Table of Content</h3><h3>Part 1</h3><p>(ICYMI: part 1 of the guide is available <a href="https://pub.towardsai.net/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-1-b8a464755f9f">here</a>)</p><ul><li>Stakeholder Analysis</li><li>What is Stakeholder Analysis?</li><li>Why Stakeholder Analysis?</li><li>How Can We Perform Stakeholder Analysis?</li><li>Applying Stakeholder Analysis to AI Governance</li><li>Identification of Stakeholders</li></ul><h3>Part 2</h3><ul><li><strong>Categorize and Prioritize Stakeholders</strong></li><li><strong>Develop and Implement Engagement Strategies</strong></li><li><strong>Monitor and Adjust</strong></li><li><strong>Actionable Takeaways</strong></li><li><strong>Conclusion</strong></li></ul><p>Recap that we have a total of <strong>9 stakeholders in AI Governance</strong> belonging to <strong>four distinct categories.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/692/1*K0hkZW3z8fSlk0pjoHT2Dg.png" /></figure><p>After our identification of the key stakeholders in AI Governance, we need to <strong><em>categorize</em></strong> the stakeholders as part of our <strong>SCIM</strong> framework.</p><p>💡Remember SCIM: Stakeholder Identification; <strong>Categorize and Prioritize</strong>;<strong> </strong>Implement Engagement Strategies; Monitor and Adjust.</p><h3>Analyze, Categorize, and Prioritize Stakeholders:</h3><blockquote>Group stakeholders based on their influence, interest, or impact on the project. Common categories include internal, external, primary, and secondary stakeholders.</blockquote><blockquote>Determine which stakeholders have the most significant impact or influence, and consider tools like power/interest grids or stakeholder maps to visualize this information.</blockquote><h4>Power-interest Grid</h4><p>The power-interest grid is a tool used in stakeholder management to <strong>categorize stakeholders based on their influence and interest</strong> in a project.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*s5Yge4mYnVRENmCBA2OAcQ.png" /></figure><p><strong>High Power, High Interest (Manage Closely):</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/580/1*bmmpXl2JzZSYr_17n-S-yA.png" /></figure><ul><li>Stakeholders in this quadrant possess both <strong>significant influence over and keen interest in the project</strong>. It’s vital to <strong>frequently engage and collaborate</strong> with these key players, ensuring their inputs are incorporated and concerns are addressed to gain their continuous support.</li></ul><p><strong>High Power, Low Interest (Keep Satisfied)</strong>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/583/1*OVNNiz1M_kfEq4sql-J6QQ.png" /></figure><ul><li>These stakeholders have considerable power but <em>might not be deeply invested in the project’s minute details</em>. It’s essential to keep them informed about <strong>major milestones and decisions</strong>, ensuring their satisfaction while avoiding overwhelming them with unnecessary details.</li></ul><p><strong>Low Power, High Interest (Keep Informed)</strong>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/583/1*ZYKFLXDkRRGB_pMFP9lBEA.png" /></figure><ul><li>Stakeholders here may not wield significant influence, but they’re <strong>very interested in the project’s progress</strong>. Regular communication tailored to their concerns and interests ensures they remain informed and positively engaged.</li></ul><p><strong>Low Power, Low Interest (Monitor)</strong>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/585/1*bW7qco9FGKKs_D0eGpcVdg.png" /></figure><ul><li>Stakeholders in this quadrant have <em>minimal influence and interest</em>. It’s still important to monitor their stance and keep them generally informed, but t<strong>hey typically require less intensive engagement than the others.</strong></li></ul><h4>Categorizing Stakeholders in AI Governance</h4><p>Given our 9 core stakeholders, we can create a <strong>Power-Interest Grid</strong> as follows:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*y8AJ2vmKijvUN-FNHjTVxw.png" /></figure><p>Let us explore the process of categorizing and prioritizing these stakeholders.</p><h4><strong>High Power, High Interest (Manage Closely):</strong></h4><p>The three core stakeholders in AI Governance placed in the <strong>Manage Closely</strong> basket are the key international and governmental bodies because <strong>these entities often have high power and high interest in AI governance</strong>; they are <em>directly responsible</em> for setting standards and regulations and <em>have vested interests</em> in doing so.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/666/1*AP21fnl2yJj-vz26DHjNIA.png" /></figure><p>Overall, Their decisions can<strong> significantly impact the development and application of AI.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/341/1*VqLayDBb7aT0Iv3EqB6hqQ.png" /></figure><p><strong>Governments/Regulatory Bodies:</strong> Possessing the power to legislate and regulate, these bodies exert <strong><em>direct control</em></strong> over actions within their jurisdiction. Their influence on AI Governance is<strong><em> </em>immediate and broad within their territories.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/250/1*8e4074mR0VzoX2yn79_WYg.png" /></figure><p><strong>International Bodies</strong>: Influencing global norms, these bodies can <strong>impact <em>numerous nations and corporations</em></strong> in the AI Governance space<em>. </em>Their broad scope and need for consensus among member nations sometimes mean they <strong><em>operate at a powerful, albeit slower pace.</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/207/1*rMSuPXWZCNUQzbZMxQcmdQ.png" /></figure><p><strong>Other Industry Associations and Professional Bodies</strong>: They set <strong><em>industry standards and have a collective voice that impacts practices</em></strong>, such as those in the field of AI safety. They are driven by the interests of their members and often have a more specialized focus than broader entities.</p><h4><strong>High Power, Low Interest (Keep Satisfied)</strong>:</h4><p>In the <strong>Keep Satisfied</strong> bracket, we have research and academic institutions They are crucial for <strong>advancing AI knowledge and informing policy through research.</strong> While they <em>may not directly enforce regulations</em>, their insights and research <strong>can significantly influence</strong> the field’s direction.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/666/1*_R25htnN3_ODqHv84evzUA.png" /></figure><p>Given their <strong>strong focus on advancing technologies and science</strong>, however, their direct interest in AI Governance may not be high.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/200/1*hi_p9k67rG8ws73XJdCiYQ.png" /></figure><p><strong>Research and Academic Institutions</strong>: These entities are centers of knowledge and innovation. Their influence is based on their <strong>academic credibility</strong>, and they often <strong><em>prioritize knowledge dissemination and research</em></strong> over commercial, political interests, or AI Governance.</p><h4><strong>Low Power, High Interest (Keep Informed)</strong>:</h4><p>End-users and ethicists are often <strong>highly interested in the safe </strong>and proper governance of AI. They are <strong>directly impacted by AI and have a vested interest in demanding AI safety.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/942/1*3iPcDQZkoqfXYxbAVi2ImQ.png" /></figure><p>Despite their strong interest, however, they often have <strong>low or limited power in influencing AI Governance policies</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/329/1*rvfZeD0uePEV1P1k6mmXOQ.png" /></figure><p><strong>End-Users</strong>: Consumers are directly affected by AI and their preferences and demands can shape the direction of AI. Though weak alone,<em> their power comes from collective demand.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/204/1*mWtGLc3bcVp7IMz6Hfl_ow.png" /></figure><p><strong>Ethicists and Philosophers</strong>: Offering moral and ethical perspectives, their voices can shape the broader discourse around technology and its societal implications, even if they lack the direct influence of other entities.</p><h4><strong>Low Power, Low Interest (Monitor)</strong>:</h4><p>The average AI Engineer typically <strong><em>prioritizes advancing their technical knowledge</em></strong> while corporations and investors are <strong>primarily interested in the monetary gain from AI.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/940/1*5SUhwWphbYoG7iVv87sDiQ.png" /></figure><p>AI Governance <strong><em>is often only important to them to the degree that they need to focus on compliance and regulatory requirements</em></strong>; a typical AI engineer, investor, or corporation generally would not have vested interests nor power to influence AI Governance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/175/1*j-3D07TbFNZotkY8kBvkHQ.png" /></figure><p><strong>Corporations</strong>: Driving the commercialization of technology, they have resources and market presence that can scale innovations, <strong><em>largely guided by profitability and market demands. </em></strong>AI Governance, however, is not their primary focus.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/137/1*h4iOgVQPr_eRyhYL6hzfNQ.png" /></figure><p><strong>Investors</strong>: While they hold financial power, they primarily influence the direction of AI to the degree of deciding which projects get greenlit. <strong><em>Their primary interest is monetary return on investments</em></strong>, not AI Governance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/151/1*xW7VrvyQnwEuoE7UeoYzXA.png" /></figure><p><strong>AI Engineers</strong>: The technical experts behind innovations, their power lies in their knowledge and skills, but <strong><em>they might lack the interest, platform, or collective voice</em></strong> to drive large-scale change in AI Governance.</p><p>Now that we have categorized our stakeholders based on the <strong>Power-Interest grid, we can move to the next step of developing engagement strategies.</strong></p><h3>Develop and Implement Engagement Strategies</h3><blockquote>Create tailored communication and engagement plans for each stakeholder group to address their specific needs, concerns, and expectations.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/306/1*ymGJVbsfERYTQuX7l5yt2A.png" /></figure><p>Consider the following <strong>approaches that you (or your organization) can take to engage with the various stakeholders</strong>:</p><p><strong>High Power, High Interest (Manage Closely):</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/794/1*EzJp-o-qt7B2j5qBc7kZMQ.png" /></figure><ul><li>🏛️ <strong>Governments/Regulatory Bodies:</strong> Regular meetings and consultations to ensure alignment with current regulations and anticipate future legislative changes. Offer training sessions or informational seminars about the latest AI advancements. Transparent reporting and proactive disclosure of AI initiatives and impacts, especially regarding ethical considerations and societal implications.</li><li>🌍 <strong>International Bodies: </strong>Attend international conferences and participate in working groups or committees. Collaborate on developing international standards or guidelines for AI. Publish white papers, position statements, or research findings to inform and influence international norms and standards.</li><li>🏢 <strong>Other Industry Associations and Professional Bodies: </strong>Join industry-specific roundtables and work closely with these bodies to shape industry standards and best practices. Offer partnership opportunities for research or pilot projects. Share best practices, case studies, and technological advancements with these associations through journals, conferences, and seminars.</li></ul><p><strong>High Power, Low Interest (Keep Satisfied):</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/798/1*9wYUQ9WNiMcBnKkqwVu2xQ.png" /></figure><ul><li>🏫 <strong>Research and Academic Institutions:</strong> Fund and collaborate on research projects. Offer internships or scholarships to students and establish academic chairs or centers of excellence. Publish joint research findings, conduct workshops, and co-host academic conferences. Engage in knowledge exchange programs.</li></ul><p><strong>Low Power, High Interest (Keep Informed):</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/797/1*7kTusKRShfCha1H7ZAg3ww.png" /></figure><ul><li>👨 <strong>End-Users:</strong> Gather feedback through surveys, focus groups, or beta testing. Ensure user-friendly AI solutions and address their concerns promptly. Regular product updates, user manuals, FAQ sections, and hosting community forums for open dialogue.</li><li>🦉 <strong>Ethicists and Philosophers: </strong>Invite them to participate in ethical review boards or committees overseeing AI projects. Host or sponsor ethical debates or forums. Publish ethical considerations and reflections regarding AI projects. Engage in open dialogues and debates.</li></ul><p><strong>Low Power, Low Interest (Monitor):</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/799/1*tlY8ThdqpM2MrAplrephQQ.png" /></figure><ul><li>💼 <strong>Corporations: </strong>Establish strategic partnerships, co-innovation labs, or joint ventures. Regularly review contractual agreements and explore opportunities for collaboration. Conduct quarterly business reviews, share market insights, and organize joint marketing or product launch events.</li><li>💵 <strong>Investors: </strong>Regular investor meetings, financial updates, and clear demonstration of business growth and ROI. Quarterly financial reports, annual general meetings, and investor relations communications.</li><li>👨‍💻 <strong>AI Engineers: </strong>Provide ongoing training, workshops, and opportunities for engineers to attend conferences or further their education. Foster an inclusive environment where they can voice their opinions and concerns. Internal newsletters, regular team meetings, and feedback sessions.</li></ul><h3>Monitor and Adjust</h3><p>Finally, you need to ensure continuous monitoring of stakeholder engagement for AI Governance success.</p><blockquote>Continually monitor stakeholders’ responses and feedback. Adapt engagement strategies as needed to ensure ongoing support and alignment.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/316/1*lM8ryhVJypRBxC37lIRavA.png" /></figure><p><strong>Periodically review feedback and engagement metrics</strong> to gauge the effectiveness of communication and collaboration strategies. As the AI landscape evolves, <strong>stakeholders’ interests and concerns will inevitably shift</strong>.</p><p>Adapt engagement tactics accordingly to <strong>maintain alignment, trust, and mutual benefit</strong>, ensuring that AI governance remains responsive, inclusive, and dynamic.</p><p>If you made it this far, congratulations! 🎉 We have completed the <strong>Complete Stakeholder Analysis of Key Stakeholders in AI Governance.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/515/1*wTHul25feb1m_I2dBDVmAg.png" /></figure><p>Now that we have completed our <strong>SCIM framework</strong>, let us explore the <strong>key takeaways </strong>from our analysis.</p><h3>Actionable Takeaways 🚀</h3><p>To ensure the successful implementation of AI Governance through comprehensive stakeholder analysis, focus on the following:</p><ol><li><strong>Empower the End-Users</strong>: While technical and regulatory stakeholders are essential, don’t underestimate the insights from end-users. They often offer real-world feedback that can lead to more ethical and efficient AI applications.</li><li><strong>Stay Updated</strong>: Regularly review international standards, guidelines, and research from organizations like the OECD and UNESCO to ensure compliance with global best practices.</li><li><strong>Prioritize Ethical Decision-making</strong>: Always factor in the ethical implications of AI. Seeking advice from multiple stakeholders can provide valuable perspectives on making the most informed choice for complex moral dilemmas.</li></ol><h3>Conclusion</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*W9s4jf62bXJVHPB2jHaCTw.png" /></figure><p>Effective governance in the AI landscape hinges on comprehensive stakeholder management, <strong>addressing the myriad of ethical, societal, and operational challenges</strong>. Stakeholder analysis ensures that all voices are considered, risks are minimized, and<strong> collective objectives align with AI initiatives.</strong></p><p>As AI technologies increasingly influence our world, it’s imperative to maintain robust, adaptable governance. Stakeholders, far from being mere observers, are active participants, <strong>collaboratively guiding AI towards a responsible, ethical, and universally beneficial future.</strong></p><p>I hope you found value in this piece. Cheers!</p><h3>Acknowledgement &amp; References</h3><ul><li><a href="https://oecd.ai/en/ai-principles">OECD AI Principles</a></li><li><a href="https://www.forbes.com/advisor/business/what-is-stakeholder-analysis/">Forbes: “What Is A Stakeholder Analysis? Everything You Need To Know”</a></li><li><a href="https://unesdoc.unesco.org/ark:/48223/pf0000380455">UNESCO Recommendation on the Ethics of Artificial Intelligence</a></li></ul><p><em>Thank you for journeying with me through my introduction to “The AI Governance Journal”. As we delve deeper into the realms of AI governance, I invite you to join the conversation, challenge the status quo, and champion responsible AI. For a holistic understanding, </em><a href="https://www.linkedin.com/in/lyejiajun/"><em>follow me</em></a><em> for more insightful AI Governance pieces. Stay curious and stay informed.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=6b9039649317" width="1" height="1" alt=""><hr><p><a href="https://pub.towardsai.net/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-2-6b9039649317">A Comprehensive Guide to Stakeholder Analysis in AI Governance (Part 2)</a> was originally published in <a href="https://pub.towardsai.net/">Towards AI</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Comprehensive Guide to Stakeholder Analysis in AI Governance (Part 1)]]></title>
            <link>https://pub.towardsai.net/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-1-b8a464755f9f?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/b8a464755f9f</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ai-governance]]></category>
            <category><![CDATA[the-ai-governance-journal]]></category>
            <category><![CDATA[stakeholder]]></category>
            <category><![CDATA[ai-ethics]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Mon, 30 Oct 2023 18:01:38 GMT</pubDate>
            <atom:updated>2023-12-11T01:55:16.153Z</atom:updated>
            <content:encoded><![CDATA[<h4>Identification of Key Stakeholders in AI Governance</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*103B31k6t8Fb86eHMtImbQ.png" /></figure><p>In the previous edition of <a href="https://www.linkedin.com/newsletters/the-ai-governance-journal-7110822650079318017/"><strong><em>The AI Governance Journal</em></strong></a>, we covered the <a href="https://www.linkedin.com/pulse/12-core-principles-ai-governance-lye-jia-jun">12 Core Principles of AI Governance</a>. These principles serve as a compass to guide us through ethical dilemmas.</p><blockquote>Have you ever wondered who this compass is meant for?</blockquote><p>Effective AI Governance is deeply rooted in <strong>People</strong>. Central to this is a comprehensive <strong>stakeholder analysis</strong> — a process that delves into the <strong>myriad interests, concerns, and influences </strong>of those involved in or affected by AI.</p><p>In this two-part series, we embark on a journey to unravel these complex interactions, offering insights on <strong>identifying, engaging, and balancing the wide mix of stakeholder interests.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/216/1*PrWSASCkklVt_OC8uwNr4w.png" /></figure><p>This first part will lay down the foundation and provide <strong>an overview of the stakeholder analysis process</strong>. Our journey will progress into its application in AI, leading to the <strong>identification of pivotal stakeholders in AI Governance.</strong></p><p>In Part 2, we’ll delve deeper into <strong>analyzing, categorizing, and prioritizing stakeholders </strong>in AI Governance, and more.</p><p>Without further ado, Let us embark on the first phase of this insightful guide.</p><h3>Table of Content</h3><h4>Part 1</h4><ul><li><strong>Stakeholder Analysis</strong></li><li><strong>What is Stakeholder Analysis?</strong></li><li><strong>Why Stakeholder Analysis?</strong></li><li><strong>How Can We Perform Stakeholder Analysis?</strong></li><li><strong>Applying Stakeholder Analysis to AI Governance</strong></li><li><strong>Identification of Stakeholders</strong></li></ul><h4>Part 2</h4><ul><li>Categorize and Prioritize Stakeholders</li><li>Develop and Implement Engagement Strategies</li><li>Monitor and Adjust</li><li>Actionable Takeaways</li><li>Conclusion</li></ul><h3>Stakeholder Analysis</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/368/0*cB8r-bhPRS9-PtQu.png" /></figure><h4>What is Stakeholder Analysis?</h4><p>Stakeholder analysis is a process that involves<strong> identifying and analyzing the individuals, groups, or entities that have an interest</strong> in a particular project or business activity.</p><p>This is the first step in <strong>stakeholder management.</strong></p><h4>Why Stakeholder Analysis?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/523/0*oNu0zIGRj4vmCLRD.png" /></figure><ul><li>Fundamentally, stakeholder analysis plays a pivotal role in ensuring the <strong>successful execution and adoption of initiatives.</strong></li><li>It allows organizations to identify, understand, and address the <strong>diverse interests, needs, and concerns of various stakeholders</strong>, fostering an environment of <strong>cooperation, support, and mutual benefit.</strong></li><li>By facilitating enhanced communication and engagement, stakeholder analysis aids in <strong>mitigating risks, resolving potential conflicts, and aligning stakeholders’ objectives with the project goals.</strong></li><li>This alignment not only paves the way for smoother implementation but also boosts the project’s credibility and acceptance, <strong>driving optimal outcomes and sustainable growth.</strong></li></ul><p>💡 <strong>Fun Fact: </strong>The concept of stakeholder theory<strong> </strong>traces its roots back to R. Edward Freeman — an American philosopher and professor of business administration — in the 1980s when it was initially developed to address morals and values in managing an organization.</p><h4>How Can We Perform Stakeholder Analysis?</h4><p>Conducting a stakeholder analysis involves <strong>four steps</strong> to effectively identify, understand, and manage stakeholders’ expectations and influences.</p><p>Here’s a step-by-step guide:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/724/1*0dTQ8UgG_S5oRUX-RiH0Jg.png" /></figure><p><strong>1. Identify Stakeholders </strong>📋<strong>:<br></strong>Start by listing all potential stakeholders, including individuals, groups, or entities affected by or interested in the project.</p><p><strong>2. Categorize and Prioritize Stakeholders </strong>🏷️<strong>:</strong></p><ul><li>Group stakeholders based on their influence, interest, or impact on the project. Common categories include internal, external, primary, and secondary stakeholders.</li><li>Determine which stakeholders have the most significant impact or influence, and consider tools like <strong>power/interest grids</strong> or stakeholder maps to visualize this information.</li></ul><p><strong>3. Develop and Implement Engagement Strategies </strong>💡<strong>:<br></strong>Create tailored communication and engagement plans for each stakeholder group to address their specific needs, concerns, and expectations.</p><p><strong>4. Monitor and Adjust </strong>🔄<strong>:<br></strong>Continually monitor stakeholders’ responses and feedback. Adapt engagement strategies as needed to ensure ongoing support and alignment.</p><p>💡 Consider the following acronym to easily remember this stakeholder analysis process — <strong>SCIM</strong>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/516/1*8yWoOnQZ3ERt-DEMWXOB2w.png" /></figure><p>By following these steps, organizations can effectively manage stakeholders, leading to <em>increased support, reduced risks, and a higher likelihood of success.</em></p><h3>Applying Stakeholder Analysis to AI Governance</h3><h4>1. Identification of Stakeholders</h4><blockquote>Start by listing all potential stakeholders, including individuals, groups, or entities affected by or interested in the project.</blockquote><p>There are <strong>9 key stakeholders surrounding AI Governance, belonging to 4 distinct groups.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*h6n-T2bZy-QdbWv3KBnhGQ.png" /></figure><p><strong>Group 1: Regulatory and Standard-Setting Entities:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/254/1*e0waCrtyHYTqXSOW4_qTdw.png" /></figure><ul><li><strong>Governments/Regulatory Bodies:</strong> Governments and regulatory agencies are responsible for establishing and enforcing rules and regulations, ensuring the ethical and safe development and application of AI technologies within their jurisdictions.<br>(Examples include the US Government, the Singapore Government, etc.)</li><li><strong>International Bodies:</strong> These organizations promote cross-border cooperation and standardization, addressing global challenges and opportunities associated with AI technologies and ensuring consistent international standards.<br>(Examples include the UNESCO, European Union, etc.)</li><li><strong>Other Industry Associations and Professional Bodies:</strong> These groups set professional standards, ethical guidelines, and best practices for the AI industry, acting as intermediaries between technology innovators and regulatory bodies.<br>(Examples include the IEEE, Singapore Computer Society, etc.)</li></ul><p><strong>Group 2: AI Development and Application Entities:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/228/1*1Pvnjhu97kaFfwDZROex_g.png" /></figure><ul><li><strong>Corporations:</strong> These are the businesses and companies that develop, deploy, and utilize AI technologies, navigating the regulatory and ethical landscape to bring innovative AI products and services to the market.</li><li><strong>AI Engineers:</strong> These professionals are at the forefront of designing, creating, and refining AI technologies. Their technical and ethical choices significantly influence the AI landscape.</li><li><strong>Research and Academic Institutions:</strong> These institutions are sources of foundational and applied AI research, innovation, and insights into the capabilities and broader implications of AI technologies.</li></ul><p><strong>Group 3: Primary Financial Stakeholders:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/220/1*bnYOdqAj8fTZ1NAw_vlh_Q.png" /></figure><ul><li><strong>Investors:</strong> Individuals or groups who provide financial resources to AI initiatives and companies. They are keenly interested in the success, ethical deployment, and regulatory compliance of AI technologies to ensure a return on investment.</li></ul><p>💡 <strong>Did You Know? </strong>Since 2014, funding for AI tech startups has surged, reaching a whopping $38 billion in 2021 alone. (Source: <a href="https://www.statista.com/statistics/621468/worldwide-artificial-intelligence-startup-company-funding-by-year/">Statista Market Report</a>)</p><p><strong>Group 4: Ethical and Societal Impact Advisers:</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/426/1*O3waZURl-ypyHaCoOvsiMA.png" /></figure><ul><li><strong>Ethicists and Philosophers:</strong> They explore the ethical and moral implications of AI technologies and advise on navigating complex ethical issues to promote responsible AI development and application.</li><li><strong>End-Users: </strong>This group comprises individuals or entities using or affected by AI technologies in their daily lives or operations, providing essential insights into the real-world impacts, ethics, and effectiveness of AI applications.</li></ul><h3>Anticipating the Next Phase: Dive Deeper in Part 2 of ‘A Comprehensive Guide to Stakeholder Analysis in AI Governance’</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/313/1*sSEGBiLza-6OebIao3LDtw.png" /></figure><p>We’ve now delved into the very essence of stakeholder analysis, starting from its definition, its importance, its methodologies, and <strong>how it intertwines with AI governance.</strong></p><p>However, as we’ve started identifying the stakeholders, the next steps are crucial in shaping our approach to AI governance. <strong><em>What strategies do we implement? How do we adjust and monitor stakeholder relationships effectively?</em></strong></p><p>All these pivotal topics, from a<strong>nalyzing and categorizing stakeholders to developing precise engagement strategies and monitoring adjustments,</strong> will be the core of our upcoming segment.</p><blockquote><strong>Stay tuned for </strong><a href="https://medium.com/towards-artificial-intelligence/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-2-6b9039649317"><strong>Part 2 of this guide</strong></a>, where we’ll not only delve into these essential facets but also wrap up our discussion with a comprehensive conclusion with practical <strong>Actionable Takeaways</strong>.</blockquote><p>The journey into AI governance and stakeholder analysis is vast, and the insights that await in the next part are indispensable. Keep an eye out!</p><p><em>*Author Note: Part 2 of the Comprehensive Guide to Stakeholder Analysis in AI Governance is out. Read Part 2 </em><a href="https://medium.com/towards-artificial-intelligence/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-2-6b9039649317"><em>here.</em></a></p><h3>Acknowledgement &amp; References</h3><ul><li><a href="https://oecd.ai/en/ai-principles">OECD AI Principles</a></li><li><a href="https://www.forbes.com/advisor/business/what-is-stakeholder-analysis/">Forbes: “What Is A Stakeholder Analysis? Everything You Need To Know”</a></li><li><a href="https://unesdoc.unesco.org/ark:/48223/pf0000380455">UNESCO Recommendation on the Ethics of Artificial Intelligence</a></li><li><a href="https://www.statista.com/statistics/621468/worldwide-artificial-intelligence-startup-company-funding-by-year/">Funding of AI Tech Startup, Statista 2021</a></li></ul><p><em>Thank you for journeying with me through my introduction to “The AI Governance Journal”. As we delve deeper into the realms of AI governance, I invite you to join the conversation, challenge the status quo, and champion responsible AI. For a holistic understanding, </em><a href="https://www.linkedin.com/in/lyejiajun/"><em>follow me</em></a><em> for more insightful AI Governance pieces. Stay curious and stay informed.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b8a464755f9f" width="1" height="1" alt=""><hr><p><a href="https://pub.towardsai.net/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-1-b8a464755f9f">A Comprehensive Guide to Stakeholder Analysis in AI Governance (Part 1)</a> was originally published in <a href="https://pub.towardsai.net/">Towards AI</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The 12 Core Principles of AI Governance]]></title>
            <link>https://pub.towardsai.net/the-12-core-principles-of-ai-governance-a17fd26643c7?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/a17fd26643c7</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[the-ai-governance-journal]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai-ethics]]></category>
            <category><![CDATA[ai-governance]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Sat, 07 Oct 2023 12:01:37 GMT</pubDate>
            <atom:updated>2023-10-07T13:02:39.991Z</atom:updated>
            <content:encoded><![CDATA[<h4>Balancing Ethics and Innovation: An Introduction to the Guiding Principles of Responsible AI</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bDXOrUX7TdqoPzdiYZRFiw.png" /></figure><p>Sarah, a seasoned AI developer, found herself at a moral crossroads. One algorithm could maximize efficiency but at the cost of privacy. The other safeguards personal data but lacks speed.</p><blockquote><em>If you were Sarah, which algorithm would you choose?</em></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/417/1*JqHHo1xXV41H9WnCNn-GZA.png" /></figure><p>These kinds of decisions are made daily. Yet, <strong>we seldom ever pause to consider frameworks of thinking</strong> in order to <strong>make the most<em> informed decision</em>.</strong></p><h3>The Need for an AI Governance Framework</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/243/1*HGnC8nnI5HzuSOmN4Jfoag.png" /></figure><p>The truth is that <strong>making such decisions requires more than technical expertise</strong>; it calls for a <strong>framework to balance efficiency with ethical considerations</strong>. This is where AI Governance steps in, offering guidelines to <strong>reconcile the competing priorities of performance and ethics.</strong></p><p>The core principles of AI Governance aren’t just theoretical constructs; <strong>they’re practical guides that can help professionals like Sarah make informed decisions</strong>.</p><p>In the following sections, we’ll explore the <strong>12 Core Principles of AI Governance</strong>, illustrating how they can be <strong>instrumental in resolving the complex dilemmas that punctuate the AI landscape.</strong></p><h3>1. Accountability</h3><blockquote>AI actors should be accountable for the proper functioning of AI systems and for the respect of AI ethics and principles, based on their roles, the context, and consistent with the state of art.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/283/1*OsM-CkaF3iINEWWqRWkzqg.png" /></figure><p>Accountability means that <strong>there should be mechanisms in place to identify and mitigate biases, errors, and unintended consequences</strong>, and ensure that the <strong>legal and ethical responsibilities are clearly defined and upheld.</strong></p><p>Accountability aims to answer the question: <strong><em>Who should be held accountable for the good (or bad) decisions that AI systems make?</em></strong></p><h3>2. Auditability</h3><blockquote>AI systems should allow interested third parties to probe, understand, and review the behaviour of the algorithm through the disclosure of information that enables monitoring, checking, or criticism.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/327/1*I1j04t79jPaXWr-JS2A8FA.png" /></figure><p>Auditability means that AI systems must be <strong>designed and operated</strong> in a way that allows <strong>their decisions, data, and operations to be inspected, evaluated, and verified by independent third parties.</strong></p><p>Auditability aims to answer the question: <strong><em>How can we ensure that AI systems are operating as intended and adhering to legal and ethical guidelines?</em></strong></p><h3>3. Explainability</h3><blockquote>AI systems should be designed such that we have the ability to assess the factors that led to the AI system’s decision, its overall behaviour, outcomes, and implications</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/253/1*20y0O4iJ1zLy0NTKs4-dYw.png" /></figure><p>Explainability means that the <strong>internal mechanisms of AI systems </strong>and the <strong>relationships between input data and decisions should be understandable.</strong></p><p>Explainability aims to answer the question(s): <strong><em>What are the factors influencing the AI system’s decision, and how are they utilized in the decision-making process?</em></strong></p><h3>4. Fairness</h3><blockquote>AI systems should be designed to treat all individuals and groups fairly and not discriminate against specific cohorts.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/282/1*fWXC2pNBMSliMjnDh39E4A.png" /></figure><p>Fairness means <strong>designing and operating AI systems in a way that is impartial, just, and equitable</strong>, ensuring that decisions made do not favor one group over another due to bias, prejudices, or stereotypes.</p><p>Fairness aims to answer the question: <strong><em>How are the AI systems designed to be fair and impartial to all individuals and groups?</em></strong></p><h3>5. Robustness</h3><blockquote>AI system should be resilient against tampering and manipulation and function reliably and effectively under unforeseen circumstances</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/281/1*eniKicxSXSDyKp_fhKl0QA.png" /></figure><p>Robustness means that AI systems should be constructed to <strong>withstand tampering and manipulation</strong>, ensuring they<strong> continue to operate reliably and effectively</strong> even under <strong>unexpected and challenging conditions.</strong></p><p>Robustness aims to answer the question: <strong><em>How resilient is the AI system to various types of operational and environmental changes and challenges?</em></strong></p><h3>6. Security</h3><blockquote>AI systems and related infrastructure should be secure against attacks and be able to maintain confidentiality, integrity, and availability</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/289/1*jkRRMnr0-C_lYM7Tqd388w.png" /></figure><p>Security means that the AI systems and their <strong>supporting infrastructures are fortified</strong> against various forms of attacks, ensuring the <strong>confidentiality, integrity, and availability of data and systems.</strong></p><p>Security aims to answer the question: <strong><em>How well are we able to protect systems and infrastructures to withstand attacks and ensure confidentiality, integrity, and availability?”</em></strong></p><h3>7. Safety</h3><blockquote>AI systems must be designed to be safe and not cause (physical) harm to humans.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/272/1*8j0exFKf0hd0rnsyytKGRg.png" /></figure><p>Safety means that AI systems should be designed with <strong>significant importance placed on human well-being</strong>, ensuring they <strong>do not pose risks or cause physical harm to individuals.</strong></p><p>Safety aims to answer the question: <strong><em>How safe and risky are the AI systems?</em></strong>”</p><h3>8. Transparency</h3><blockquote>AI systems and their operators should ensure that the processes, decisions, and impacts of AI applications are clear, understandable, and accessible to affected parties and stakeholders.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/281/1*gpjSXz-KeJVfvQVQdfnGSg.png" /></figure><p>Transparency means that there is <strong>openness and clarity in sharing information</strong> about the <strong>AI system’s design, operation, and decision-making processes.</strong></p><p>Transparency aims to answer the question: <strong><em>Is clear and accessible information about the AI system available to all stakeholders?</em></strong></p><h3>9. Reproducibility</h3><blockquote>AI results should be reproducible; others should be able to replicate outcomes given the same inputs and conditions.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/285/1*H_HDYpb6Z1eBagGt0bjBjQ.png" /></figure><p>Reproducibility means the <strong>AI’s decisions and outputs can be consistently replicated using the same data and conditions.</strong></p><p>Reproducibility aims to answer the question: <strong><em>Can the AI system’s decisions be reproduced under the same conditions?</em></strong></p><h3>10. Human Agency and Oversight</h3><blockquote>AI systems should support human autonomy and decision-making; humans should have the ability to intervene, oversee, and take control of AI systems.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/192/1*qAdTLoKpQIbIjh1trgIx-Q.png" /></figure><p>Human Agency and Oversight mean that <strong>AI should not undermine human autonomy and capacity for decision-making</strong>, and <strong>humans should remain in control.</strong></p><p>Human Agency and Oversight aims to answer the question: <strong><em>To what extent do humans have control over and the ability to intervene in the AI system’s decisions?</em></strong></p><h3>11. Privacy and Data Governance</h3><blockquote>AI systems should respect privacy, data protection, and security norms, ensuring individuals’ data are handled with utmost integrity.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/260/1*iQYjSZwdCifuKff_XxefzA.png" /></figure><p>Privacy and Data Governance means that personal data is <strong>protected, and systems are designed and operated in ways that respect individual privacy.</strong></p><p>Privacy and Data Governance aims to answer the question: <strong><em>How does the AI system ensure the protection of individual privacy and data?</em></strong></p><h3>12. Inclusive growth, sustainable development, and well-being</h3><blockquote>AI systems should be designed to foster universal prosperity, equity, and environmental health, ensuring benefits for all individuals, societies, and generations while protecting ecological balance.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/268/1*KpZTcmyKuI5oSp63mgei5Q.png" /></figure><p>Inclusive growth, sustainable development, and well-being means that AI systems should be<strong> designed and deployed to benefit all individuals and communities, promoting equity, sustainability, and overall well-being.</strong></p><p>Inclusive growth, sustainable development, and well-being aim to answer the question: <strong><em>How does the AI system contribute to the equitable economic growth, environmental sustainability, and well-being of individuals and communities?</em></strong></p><h3>Utilizing AI Governance Principles</h3><p>Remember Sarah? The seasoned AI developer who needs to choose between two algorithms that prioritize either efficiency or privacy.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/338/1*UoVI4Q50AXrG7GkqQCttJg.png" /></figure><p><strong>If Sarah were aware of the core principles of AI Governance</strong>, she’d chosen the <strong>algorithm that protects the users’ privacy</strong>, in accordance with the <strong>Privacy and Data Governance</strong> principle.</p><p>Of course, this scenario is a straightforward instance where the <strong>core AI Governance principles are not in conflict with each other</strong>. In reality, principles often <strong>intersect and conflict.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/193/1*_buqTa2faOxgMAEsXVdjZA.png" /></figure><p>The conflicts among AI Governance principles <strong>aren’t a flaw but a feature</strong>, compelling us to <strong>scrutinize, debate, and refine the role of AI in our society continually</strong>. It is in this dynamic space, where <strong>principles meet practice,</strong> that the <strong>true value of AI Governance is revealed.</strong></p><h3>Conclusion: The Importance of Principles</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/652/1*awliaExLiSjjC-aRSosg3Q.png" /></figure><p>In the ongoing development of AI, these 12 Core AI Governance Principles serve as <strong>foundational elements, providing direction and clarity</strong>. They help ensure that AI is <strong>developed and utilized for the benefit of all, fostering innovation and improving human well-being while balancing ethical demands.</strong></p><p>In the upcoming pieces, we will <strong>deep dive into each core principle</strong>, bringing to life <strong>real-world case studies </strong>and <strong>evaluating how different organizations follow (or violate) the core AI Governance principles.</strong></p><p>I hope you found this foundational piece insightful.</p><p>Cheers!</p><h3>Acknowledgement &amp; References</h3><ul><li><a href="https://www.pdpc.gov.sg/-/media/Files/PDPC/PDF-Files/Resource-for-Organisation/AI/SGModelAIGovFramework2.pdf">Singapore Model AI Governance Framework 2.0</a>; <a href="https://aiverifyfoundation.sg/">AI Verify Foundation</a></li><li><a href="https://oecd.ai/en/ai-principles">OECD AI Principles</a></li><li><a href="https://digital-strategy.ec.europa.eu/en/library/ethics-guidelines-trustworthy-ai">European Commission, “Ethics Guidelines for Trustworthy AI,” Digital Strategy</a></li><li><a href="https://standards.ieee.org/wp-content/uploads/import/documents/other/ead_v2.pdf">IEEE Global Initiative on Ethics of Autonomous and Intelligent Systems</a></li><li><a href="https://unesdoc.unesco.org/ark:/48223/pf0000380455">UNESCO Recommendation on the Ethics of Artificial Intelligence</a></li></ul><p><strong><em>*Other principles such as inclusivity, accuracy, progressiveness, and human centricity &amp; well-being are not included as they were not commonly established among various sources.</em></strong></p><p><em>Thank you for journeying with me through my introduction to “The AI Governance Journal”. As we delve deeper into the realms of AI governance, I invite you to join the conversation, challenge the status quo, and champion responsible AI. For a holistic understanding, </em><a href="https://www.linkedin.com/in/lyejiajun/"><em>follow me</em></a><em> for more insightful AI Governance pieces. Stay curious and stay informed.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a17fd26643c7" width="1" height="1" alt=""><hr><p><a href="https://pub.towardsai.net/the-12-core-principles-of-ai-governance-a17fd26643c7">The 12 Core Principles of AI Governance</a> was originally published in <a href="https://pub.towardsai.net/">Towards AI</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Introduction to The AI Governance Journal]]></title>
            <link>https://medium.com/@lyejiajun/introduction-to-the-ai-governance-journal-4b8188e3cad8?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/4b8188e3cad8</guid>
            <category><![CDATA[the-ai-governance-journal]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai-governance]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[ai-ethics]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Fri, 22 Sep 2023 03:17:12 GMT</pubDate>
            <atom:updated>2023-12-11T01:57:15.952Z</atom:updated>
            <content:encoded><![CDATA[<h4>Charting the Course for Responsible AI Futures</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KXdsojhgKTsuKgJIFExc5g.jpeg" /></figure><p><em>Imagine a world where machines make decisions for us, where their choices shape economies, influence laws, and touch every facet of our lives — whether we like it or not.</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/866/1*3bxXecHX-TDcvHBk9FoS_A.png" /></figure><p>In many ways, <strong><em>that future is already here</em></strong>.</p><p>As we stand at this critical juncture, the pressing question arises: how do we ensure these technologies act in our best interest?</p><h3>Welcome to “The AI Governance Journal”</h3><p>Four years ago, I created my first-ever <a href="https://www.youtube.com/watch?v=_wF7eDu6wdQ&amp;ab_channel=LyeJiaJun">machine-learning project to detect malicious websites</a>. Since then, I’ve witnessed the rapid acceleration of AI technologies.</p><p>With each new development, a recurring thought lingers: <em>“What could possibly go wrong?”</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/448/1*EHXvQIb0JbC4WfBuxuxBqw.jpeg" /></figure><p>In this series, we will delve deep into the core of this question, exploring the intersection of AI and humanity. This isn’t just a blog series; it’ll serve as a guide through the maze of artificial intelligence, its potential, and its challenges.</p><p>As the possibilities of AI expand, so should the responsibility of ensuring that AI is implemented ethically and responsibly. This is the essence of <strong>AI Governance.</strong></p><h3>Defining AI Governance</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/349/1*YslxIl4GfMHvlRvwNuR_ug.png" /></figure><p>AI Governance refers to <strong>the principles, policies, and practices by which AI systems are managed</strong>. It ensures the <strong>adoption of AI in ethical and responsible ways.</strong></p><p><strong>It encompasses the rules and standards set by various entities,</strong> such as governmental bodies, industry groups, and individual organizations. These rules ensure that AI systems operate <strong><em>ethically, transparently, and in the best interests of society.</em></strong></p><h3>Why AI Governance?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/192/1*AWSOaIJJRf19j1tfVIFEdQ.png" /></figure><p>Fundamentally, AI Governance allows us to address the following concerns:</p><ol><li><strong>Ethical Considerations</strong>: AI systems, especially those leveraging deep learning, can unintentionally perpetuate biases found in training data. Without proper oversight, these biases can harm marginalized groups and perpetuate inequality.</li><li><strong>Transparency and Accountability</strong>: As AI takes on more critical roles, understanding how decisions are made becomes crucial. When an algorithm makes a wrong decision, <em>who should be held accountable?</em></li><li><strong>Safety and Reliability</strong>: As AI applications become more complex, ensuring they function as intended — without causing unforeseen harm — must become a top priority.</li></ol><h3>What’s Next?</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/1*Um8ZBSN--egwbYx0q_w_pQ.jpeg" /><figcaption><a href="https://pixabay.com/photos/forest-path-rocks-trail-438432/">Image from SplitShire from Pixabay</a></figcaption></figure><p>In upcoming discussions on AI Governance, we’ll dive deep into multiple vantage points, <strong>encompassing perspectives from governments, businesses, and everyday individuals.</strong></p><p>Our journey will shed light on prevalent machine-learning issues surrounding AI biases and spark conversations on ethical dilemmas. By exploring insightful case studies, I hope to <strong>bring real-world implications to the forefront.</strong></p><p>Additionally, we’ll stay abreast of the latest advancements, challenges, and solutions in the realm of AI governance.</p><p><strong>Be prepared for a thought-provoking ride.</strong></p><h4>#TheAIGovernanceJournal Article Series</h4><ul><li><a href="https://www.linkedin.com/pulse/12-core-principles-ai-governance-lye-jia-jun/?trackingId=i4h%2BRwFBT56JZLzy%2FEm5Yw%3D%3D"><strong>The 12 Core Principles of AI Governance</strong></a><strong> (Featured in </strong><a href="https://pub.towardsai.net/the-12-core-principles-of-ai-governance-a17fd26643c7"><strong>Towards AI</strong></a><strong>)</strong></li><li><a href="https://www.linkedin.com/pulse/comprehensive-guide-stakeholder-analysis-ai-part1-lye-jia-jun-aphgc/?trackingId=ynsWTiNuRgKbbfcsXT8zxg%3D%3D&amp;lipi=urn%3Ali%3Apage%3Ad_flagship3_publishing_post_edit%3BZewI6nm8TnGrWJhoHd0lsQ%3D%3D"><strong>A Comprehensive Guide to Stakeholder Analysis in AI Governance (Part 1)</strong></a><strong> (Featured in </strong><a href="https://pub.towardsai.net/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-1-b8a464755f9f"><strong>Towards AI</strong></a><strong>)</strong></li><li><a href="https://www.linkedin.com/pulse/comprehensive-guide-stakeholder-analysis-ai-part2-lye-jia-jun-zuloc"><strong>A Comprehensive Guide to Stakeholder Analysis in AI Governance (Part 2)</strong></a><strong> (Featured in </strong><a href="https://medium.com/towards-artificial-intelligence/a-comprehensive-guide-to-stakeholder-analysis-in-ai-governance-part-2-6b9039649317"><strong>Towards AI</strong></a><strong>)</strong></li></ul><p><em>Thank you for journeying with me through my introduction to “The AI Governance Journal”. As we delve deeper into the realms of AI governance, I invite you to join the conversation, challenge the status quo, and champion responsible AI. For a holistic understanding, </em><a href="https://www.linkedin.com/in/lyejiajun/"><em>follow me</em></a><em> for more insightful AI Governance pieces. Stay curious and stay informed.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4b8188e3cad8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[12 Lessons I Learned From Writing 12 Articles In 12 Months]]></title>
            <link>https://medium.com/illumination/12-lessons-i-learned-from-writing-12-articles-in-12-months-2b3811a25dc1?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/2b3811a25dc1</guid>
            <category><![CDATA[growth]]></category>
            <category><![CDATA[self-improvement]]></category>
            <category><![CDATA[crystalizemylearning]]></category>
            <category><![CDATA[writing]]></category>
            <category><![CDATA[learning]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Sun, 25 Dec 2022 16:41:22 GMT</pubDate>
            <atom:updated>2022-12-26T03:32:09.286Z</atom:updated>
            <content:encoded><![CDATA[<h4>The finale of my <a href="https://medium.com/@lyejiajun/introducing-the-crystalizemylearning-movement-1240d57ab9ff"><strong><em>#CrystalizeMyLearning</em></strong></a><strong><em> writing movement. And </em></strong>yes, I wrote 12 articles in the past 12 months.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*z_KI5RHNg2jrFYqM.jpg" /><figcaption><a href="https://pixabay.com/photos/notepad-glasses-travel-planning-1130743/">Image by DariuszSankowski via Pixabay</a></figcaption></figure><p>At the start of 2022, I founded the <a href="https://medium.com/@lyejiajun/introducing-the-crystalizemylearning-movement-1240d57ab9ff"><strong><em>#CrystalizeMyLearning</em></strong></a><strong><em> </em></strong>writing movement to <strong>learn through writing.</strong></p><p>Upon founding the movement, I described <strong>three tangible goals</strong> I hope to achieve.</p><ol><li><strong>Think more critically and become more creative</strong></li><li><strong>Build a community of avid learners</strong></li><li><strong>Share my knowledge with the wider community</strong></li></ol><p>Fast forward to today, I have written twelve full articles over twelve months.</p><p><strong>The results</strong></p><ul><li><strong>Did I become a more creative thinker?</strong> Heck yes.</li><li><strong>Was I able to find passionate learners who share similar interests? </strong>Of course!</li><li><strong>Did my community find value in my articles? </strong>I hope so!</li></ul><h4><strong>20,000 words — the approximate total number of words in my past twelve articles</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*NkWZxwGkAXj-6ub9.jpg" /><figcaption><a href="https://pixabay.com/photos/paper-writing-old-antique-write-623167/">Image by Nile via Pixabay</a></figcaption></figure><p>I hope my writing of <strong>12 articles in 12 months totaling around ~20,000 total</strong> <strong>words </strong>written highlights how I probably at least learned a thing or two during this tremendous journey!</p><p>Actually, it’s more than a thing or two.</p><p>In this article, I reflect on the <strong>twelve lessons</strong> I’ve learned in this year-long writing journey.</p><p><strong><em>**Serious author note: This article is NOT written by ChatGPT!</em></strong></p><h4>Lesson 1 — Finding Sources of Inspiration</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*fKqxgsjdT2_N-Q6Y.jpg" /><figcaption><a href="https://pixabay.com/photos/sparkler-lightbulb-spray-candle-4629347/">Image by Myriams-Fotos via Pixabay</a></figcaption></figure><ul><li>Among my twelve articles, I wrote about topics such as <a href="https://medium.com/@lyejiajun/5-tips-to-excel-in-polytechnic-c628d750f47c">education</a>, <a href="https://medium.com/@lyejiajun/everything-i-have-learnt-about-mental-health-crystalizemylearning-feb-2022-a576104a32">mental health</a>, <a href="https://medium.com/@lyejiajun/python-flask-a-powerful-and-flexible-web-app-framework-crystalizemylearning-mar-2022-6890e39e6c1e">technical frameworks</a>, <a href="https://pub.towardsai.net/4-cognitive-biases-in-ai-ml-systems-a208ac87177d">psychology</a>, and even the <a href="https://medium.com/geekculture/meta-layoff-what-tech-workers-businesses-and-founders-can-learn-from-this-19acac036e6f">latest global affairs like the Meta tech layoff</a>.</li><li>You may wonder what is the source of my inspiration. The answer: <strong>podcasts, books, and articles.</strong></li><li>It wasn’t easy writing about such a disparate range of topics, but by reading more books, listening to more podcasts, and keeping up with daily affairs through articles, it became much easier to find interesting topics to write about.</li></ul><h4>Lesson 2— Discovering &amp; Learning New Technologies</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*l4SnAki4Dpx4QrORGqYGJA.jpeg" /><figcaption><a href="https://pixabay.com/photos/circles-abstract-space-design-6928543/">Image by xresch via Pixabay</a></figcaption></figure><ul><li>There are clear benefits to consuming content broadly. You’d discover new technologies like I did.</li><li>Through writing, I not only found new technologies but also gain a deeper appreciation of the “whys” behind technology as I needed to research intently before writing about a subject. I could understand <a href="https://medium.com/geekculture/understanding-the-emotions-behind-the-human-language-34938a1fb5c7"><strong>Natural Langauge Processing</strong> more profoundly through a project I built</a>, and <a href="https://levelup.gitconnected.com/what-the-heck-are-jwt-and-oauth-2-0-865312ed2aff">deep-dive into the intricacies of the <strong>OAuth 2.0 authorization standard</strong></a>.</li><li>It was super interesting to deep-dive into new technologies as you may sometimes notice fundamental principles shared across different technological domains — I highly recommend this intellectually stimulating experience!</li></ul><h4>Lesson 3— Distilling Technical Insights In Layman Terms</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*f2L4U_U_hyJ7cvhj.jpg" /><figcaption><a href="https://pixabay.com/photos/drawing-plan-design-sketch-project-933207/">Image by tookapic via Pixabay</a></figcaption></figure><ul><li>Through my best efforts, I tried to make my technical and <a href="https://en.wiktionary.org/wiki/devlog">‘devlog-style’</a> articles easier to read for laymen.</li><li>I’m not able to confidently say that every layman could easily understand my “<a href="https://medium.com/codex/becoming-a-metaverse-developer-de398a4702cb"><strong>Becoming A Metaverse Developer</strong></a><strong>” </strong>article where I shared about Unity and C# Game Programming, but hey, at least two of my friends found it helpful to kickstart their game programming projects.</li><li>I’m sure this skill of distilling technical subjects to laymen will come in handy in the future, so I’m super excited for when that time comes!</li></ul><h4>Lesson 4 — Dealing With Writer’s Block</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*YejvQljxP72_Ae-DSzhwYA.jpeg" /><figcaption><a href="https://pixabay.com/photos/pen-ball-point-paper-crumpled-3301338/">Image by steve_a_johnson via Pixabay</a></figcaption></figure><ul><li>If you haven’t sat down to write write (yes — like ‘write, you know, like actual writing), you are warned against this nightmare for all writers:<strong> Writer’s Block.</strong></li><li><strong>Writer’s Block</strong> refers to a period of creative slowdown or a specific moment when your mind goes instantly blank, completely shutting down your ability to write.</li><li>During such occasions, I learned to either take a break from writing or do the classic <a href="https://en.wikipedia.org/wiki/Stream_of_consciousness"><strong><em>stream-of-consciousness writing</em></strong></a><strong><em> </em></strong>(a.k.a brain dump) method. Give it a shot when you’re stuck during a writing session next time. It’ll honestly help to spark more creativity!</li></ul><h4>Lesson 5— Revisiting My In(complete) Work</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*o2cCTPUaKMdd-QXk.jpg" /><figcaption><a href="https://pixabay.com/photos/philatelist-stamp-collection-stamp-1844081/">Image by qimono via Pixabay</a></figcaption></figure><ul><li>The best way I’ve found to improve a piece of writing is simply to <strong>revisit it with a fresh mind </strong>so I can spot all the errors I couldn&#39;t before.</li><li>The duration before I revisit my work can range from <strong>hours to days, and maybe even weeks!</strong></li><li>To ensure my writing is clear and concise, I’d sometimes even go for a run to clear my mind before revisiting my work to refine it. Trust me when I say this works — and it works really well.</li></ul><h4>Lesson 6 — The Power of Habits</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*y6qIbk5MkHMmvqRn.jpg" /><figcaption><a href="https://pixabay.com/photos/stone-zen-white-spa-rock-316225/">Image by PublicDomainPictures via Pixabay</a></figcaption></figure><ul><li>Are there times in my life when I felt so extremely overwhelmed by the busyness of my work and life that I didn’t feel like writing another article? You bet!</li><li>However, I didn’t rely on “motivation” to write one article consistently each month. I relied on <strong><em>habits — the habits of writing consistently.</em></strong></li><li>Though difficult initially, once <strong>I conscientiously built the habit of writing by simply sitting down and ‘forcing’ myself to write on numerous occasions</strong>, I slowly grew my writing muscles. By building such habits, I gradually found it easier to write and has fewer of those ‘ughh’ writing moments.</li></ul><h4>Lesson 7— Ditch Grammarly, Learn Proper Grammar</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*TR2ZLIjMRapZGDaS.jpg" /><figcaption><a href="https://pixabay.com/photos/english-english-language-language-2724442/">Image by BiljaST via Pixabay</a></figcaption></figure><ul><li>I’m just kidding. Don’t do that. (As in, don’t ditch Grammarly. But please learn proper grammar!) (this is probably a grammatically incorrect statement but whatever)</li><li>Though it was tempting to <strong>delegate Grammarly as my catch-all error buddy</strong>, I still tried to (re)learn grammar while I’m writing — especially knowing how important it is.</li><li>I’m not an English teacher but I’m happy to say that I’ve at least learned <strong>several grammatical rules like the proper usage of colons, parenthesis, and em-dashes</strong>. Fun fact: there are different writing standards that may advocate different usage of punctuation.</li></ul><h4>Lesson 8 — In and Outs of LinkedIn &amp; Medium</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*YlSi87o0LUVDjlLCwGNx_w.png" /></figure><ul><li>I published all of my articles on both <a href="https://medium.com/@lyejiajun">Medium </a>and <a href="https://www.linkedin.com/in/lyejiajun/">LinkedIn</a>.</li><li>Needless to say, I learned about the quirks of each platform, such as the <strong>(in)ability to add headers, image captions, and dividers.</strong> <strong>Though it may seem minor, I believe they are critical to making your writing more readable.</strong></li><li>Also, since writing this article, LinkedIn has improved its writing platform so there’s no need for me to use the hack-ish quoted triple asterisk (***) divider workaround anymore:”)<br>(if you ever read my LinkedIn article on mobile devices, this is the reason why there is a weird triple asterisk in my articles)</li></ul><h4>Lesson 9 — Writing Etiquette for Publications</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*VBtVpwOxjhgg61rgG3Pjlw.jpeg" /><figcaption><a href="https://pixabay.com/photos/book-etiquette-manners-emily-post-2752591/">Image by MorningbirdPhoto via Pixabay</a></figcaption></figure><ul><li>Do you know that there are <strong>best writing practices that differ from publication to publication?</strong></li><li>Though I was initially surprised too, I eventually understood why. <strong>Different publication caters to different audiences</strong>, thus the content you produce may need to vary based on length, writing style, etc. to suit those specific audiences.</li><li>I think such writing standards can be beneficial because readers could easily know what types of articles to expect in a publication. For writers, <strong>such standards serve as a helpful guideline too.</strong></li></ul><h4>Lesson 10 — Gained Newfound Appreciation for Fields Like Psychology</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*zjVNoOf-f-Fhpj4xPauI7Q.png" /><figcaption><a href="https://pixabay.com/illustrations/mind-brain-mindset-perception-544404/">Image by johnhain via Pixabay</a></figcaption></figure><ul><li>In an attempt to <strong>step out of my comfort zone</strong>, I explored fields outside of computer science through this writing movement.</li><li>In several articles, I shared about <a href="https://medium.com/@lyejiajun/everything-i-have-learnt-about-mental-health-crystalizemylearning-feb-2022-a576104a32">mental health</a>, <a href="https://medium.com/illumination/3-counter-intuitive-ideas-to-challenge-your-thinking-e31882eb83a9">psycholog</a>y, and even the <a href="https://pub.towardsai.net/4-cognitive-biases-in-ai-ml-systems-a208ac87177d">cognitive science behind machines and biases</a>.</li><li><strong>By researching, learning, and introspecting on ideas that are foreign to me</strong>, I undoubtedly gained a deeper appreciation for different fields of study such as cognitive science, psychology, and neuroscience — super dope!</li></ul><h4>Lesson 11 — Gaining Self Awareness</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*wkY-L6WutLxSyZQq76mszw.jpeg" /><figcaption><a href="https://pixabay.com/photos/psychology-psyche-mask-wire-frame-2706899/">Image by geralt via Pixabay</a></figcaption></figure><ul><li>There are two parts to this. Firstly, I’m more aware of my preferred writing style. My articles tend to be lengthier because <strong>I firmly believe that context is important for most lessons I’m trying to share.</strong></li><li>Such awareness made me curious to try shorter, more concise forms of writing. Thus sometimes you may find some shorter 3–4 minute length of articles I’ve written.</li><li>Also, I’m more aware of the <strong>mental processes, habits, and environment which provides the optimal mental clarity for me to craft a great piece of writing.</strong></li></ul><h4>Lesson 12 — Becoming More Open-Minded</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*vTv1I0llAw3DdnSs.png" /><figcaption><a href="https://pixabay.com/vectors/open-mind-thoughts-thinking-brain-6570815/">Image by mohamed_hassan via Pixabay</a></figcaption></figure><ul><li>During the earlier stage of writing, I’d often seek advice from my friends regarding the content I’m about to write.</li><li>I see the value of being open-minded because I truly received so much wonderful feedback, <strong>like bolding words more mindfully, adding more images so the article is more interesting to read, and more.</strong></li><li>As I conclude my final lesson, I hope to give a huge shout-out to my long-time supporters: Pei Qi, Razan, Jeffery, Ian, and Veron!<br>Thank you guys so much! 🤗🤗</li></ul><h3>That’s a wrap! (and closing thoughts)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/959/1*d-6R78-7ovDT5eKxN7x9Nw.jpeg" /></figure><p>This article officially marks the end of the <a href="https://medium.com/@lyejiajun/introducing-the-crystalizemylearning-movement-1240d57ab9ff">#CrystalizeMyLearning movement</a> where I learn through writing.</p><p>It has been a <strong>crazy twelve months of juggling between military service, volunteering work, and writing one article each month, amongst many other things.</strong></p><p>To all my readers, I sincerely thank you for your support. Even though I may step away from formal writing for a while — mostly to recharge — <strong>I have learned so much during the past twelve months, and I’ll definitely continue creating more content in the future.</strong></p><p>I hope you found this article interesting. If any of my writing resonates with you, I’d love to connect with you on <a href="https://www.linkedin.com/in/lyejiajun/">LinkedIn</a>.</p><p>Thank you so much for reading and following me on this year-long journey! I’ll catch you in my (not-so-soon) next article. Cheers!!! 😁😁</p><h3>In Case You Missed It: Here’s the Entire #CrystalizeMyLearning Article Collection</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/471/0*IEYiT-Ur6wWdmbZi.png" /></figure><p><strong>January Article: </strong><a href="https://medium.com/@lyejiajun/5-tips-to-excel-in-polytechnic-c628d750f47c">5 Tips to Excel In Polytechnic | #CrystalizeMyLearning (Jan 2022</a>) (<a href="https://medium.com/sgexams/5-tips-to-excel-in-polytechnic-c628d750f47c">Featured in SGExams</a>, <a href="https://fourblindmicesg.com/5-tips-to-excel-in-polytechnic/">FourBlindMice</a>, and <a href="https://www.polyparcelsg.com/school/5-tips-to-excel-in-polytechnic/">PolyParcel</a>)</p><p><strong>February Article: </strong><a href="https://medium.com/@lyejiajun/everything-i-have-learnt-about-mental-health-crystalizemylearning-feb-2022-a576104a32">Everything I have Learnt About Mental Health | #CrystalizeMyLearning (Feb 2022)</a> (Featured in I’m Friendly Co and Deepstash App)</p><p><strong>March Article: </strong><a href="https://medium.com/@lyejiajun/python-flask-a-powerful-and-flexible-web-app-framework-crystalizemylearning-mar-2022-6890e39e6c1e">Python Flask — A Powerful And Flexible Web App Framework | #CrystalizeMyLearning (Mar 2022) (Featured in Level Up Coding Medium)</a></p><p><strong>April Article: </strong><a href="https://medium.com/codex/becoming-a-metaverse-developer-de398a4702cb">Becoming A Metaverse Developer | #CrystalizeMyLearning (April 2022) (Featured in CodeX Medium)</a></p><p><strong>May Article: </strong><a href="https://levelup.gitconnected.com/i-published-my-first-chrome-extension-on-the-chrome-web-store-b20f0f75cb03">I Published My First Chrome Extension on the Chrome Web Store! | #CrystalizeMyLearning (May 2022) (Featured in Level Up Coding Medium)</a></p><p><strong>June Article:</strong> <a href="https://medium.com/geekculture/understanding-the-emotions-behind-the-human-language-34938a1fb5c7">Understanding The Emotions Behind The Human Language | #CrystalizeMyLearning (June 2022) (Featured in Geek Culture Medium)</a></p><p><strong>July Article:</strong> <a href="https://levelup.gitconnected.com/what-the-heck-are-jwt-and-oauth-2-0-865312ed2aff">What the heck are JWT and OAuth 2.0? | #CrystalizeMyLearning (July 2022) (Featured in Level Up Coding Medium)</a></p><p><strong>August Article:</strong> <a href="https://medium.com/illumination/3-ways-to-turn-knowledge-into-action-3978282c7493">3 Ways To Turn Knowledge Into Action | #CrystalizeMyLearning (August 2022) (Featured in Illumination </a>and Deepstash App)</p><p><strong>September Article:</strong> <a href="https://medium.com/illumination/3-counter-intuitive-ideas-to-challenge-your-thinking-e31882eb83a9">3 Counter-Intuitive Ideas To Challenge Your Thinking | #CrystalizeMyLearning (September 2022) (Featured in Illumination</a> and Deepstash App)</p><p><strong>October Article:</strong> <a href="https://pub.towardsai.net/4-cognitive-biases-in-ai-ml-systems-a208ac87177d">4 Cognitive Biases In AI/ML Systems | #CrystalizeMyLearning (October 2022) (Featured in Towards AI</a> and Deepstash App)</p><p><strong>November Article:</strong> <a href="https://medium.com/geekculture/meta-layoff-what-tech-workers-businesses-and-founders-can-learn-from-this-19acac036e6f">Meta Layoff: What Tech Workers, Businesses, and Founders Can Learn From This | #CrystalizeMyLearning (November 2022) (Featured in GeekCulture)</a></p><p><strong>December Article: </strong><a href="https://medium.com/illumination/12-lessons-i-learned-from-writing-12-articles-in-12-months-2b3811a25dc1">12 Lessons I Learned From Writing 12 Articles In 12 Months | #CrystalizeMyLearning (December 2022, Finale Edition) (Featured in Illumination)</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2b3811a25dc1" width="1" height="1" alt=""><hr><p><a href="https://medium.com/illumination/12-lessons-i-learned-from-writing-12-articles-in-12-months-2b3811a25dc1">12 Lessons I Learned From Writing 12 Articles In 12 Months</a> was originally published in <a href="https://medium.com/illumination">ILLUMINATION</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Meta Layoff: What Tech Workers, Businesses, and Founders Can Learn From This]]></title>
            <link>https://medium.com/geekculture/meta-layoff-what-tech-workers-businesses-and-founders-can-learn-from-this-19acac036e6f?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/19acac036e6f</guid>
            <category><![CDATA[meta]]></category>
            <category><![CDATA[crystalizemylearning]]></category>
            <category><![CDATA[founders]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[business]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Mon, 28 Nov 2022 13:02:01 GMT</pubDate>
            <atom:updated>2022-11-29T08:43:53.247Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*DAUDN4YOLvPSgh9a.jpg" /><figcaption><a href="https://pixabay.com/illustrations/meta-logo-3d-social-network-6946620/">Image by Artapixel via Pixabay</a></figcaption></figure><p><a href="https://www.nytimes.com/2022/11/09/technology/meta-layoffs-facebook.html">On 9th November 2022, Meta announced a massive layoff of more than 11,000 employees</a>. The layoff, possibly influenced by Twitter’s massive employee cutback after Elon Musk’s acquisition, was Meta&#39;s first significant job cut since its 18-year history.</p><p>What are the implications of this layoff? Are we expecting another dotcom bubble burst? What can individuals, businesses, and founders learn from this?</p><p>We’ll be exploring all of that and more in this article.</p><h3>Table of Content</h3><ul><li><strong>About The Meta Layoff</strong></li><li><strong>The Implications For Tech Workers</strong></li><li><strong>What Businesses Can Learn From This</strong></li><li><strong>The Case for Founders</strong></li><li><strong>Summary and Key Takeaways</strong></li></ul><h3>About The Meta Layoff</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*gwklX9DNcBVp1TCz.jpg" /><figcaption><a href="https://pixabay.com/illustrations/mobile-smartphone-app-networks-1087845/">Image by geralt via Pixabay</a></figcaption></figure><p>Before we delve into the specifics of the layoff, it is imperative that we have a <strong>big-picture overview</strong> of the situation.</p><p><strong>So why did such a massive layoff happen at Meta?</strong></p><p>In the words of CEO Mark Zuckerberg:</p><blockquote><em>Today I’m sharing some of the most difficult changes we’ve made in Meta’s history. I’ve decided to reduce the size of our team by about 13% and let more than 11,000 of our talented employees go. We are also taking a number of additional steps to become a leaner and more efficient company by cutting discretionary spending and extending our hiring freeze through Q1.</em></blockquote><p>Based on the words of Meta CEO Mark Zuckerberg, it seems like Meta’s layoff was driven by the goal to <strong>reduce expenses and become a more (cost)efficient company.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*c8TOjYrh-bVHpBYJ.jpg" /><figcaption><a href="https://pixabay.com/photos/financial-analytics-blur-business-2860753/">Image by 6689062 via Pixabay</a></figcaption></figure><p>Looking at objective metrics, that decision makes perfect sense in terms of protecting the company’s bottom line. Since Facebook rebranded itself into Meta, the company has reported its <strong>second-ever quarterly drop in revenue and saw profit cut in half from the prior year.</strong></p><p>Much of the revenue decline, however, was not merely due to the company’s strategic shift towards building the Metaverse. <strong>It was a mixture of post-pandemic conditions, rising interest rates, and other factors that affected the tech industry as a whole.</strong></p><h3>The Implications For Tech Workers</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*cmqpZNHX_ILmcK66.jpg" /><figcaption><a href="https://pixabay.com/photos/software-developer-web-developer-6521720/">Image by Innovalabs via Pixabay</a></figcaption></figure><p>Now we have the big-picture in mind, we can address the elephant in the room —<strong> does this layoff means that the end for tech workers is soon to come?</strong> Well, it seems like it isn’t that straightforward.</p><ul><li>The huge layoff is mostly limited to some massive organizations like Meta and Twitter while <strong>some smaller companies are seemingly unaffected.</strong></li><li>The likely explanation is that the layoff is largely driven by the <strong>massive hiring during the pandemic boom</strong> as demand for digital services increases. Naturally, <strong>larger companies would conduct a bigger hiring spree than the smaller ones.</strong></li><li>As the pandemic condition starts to improve, there is a shift in market demand, <strong>leading to a decline in consumers’ desires for various digital services. </strong>This means some tech workers are fired <strong>because they are no longer required in a handful of companies, not because they are no longer valuable in the entire market.</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*dEEVyXXw5iKD9_hf.jpg" /><figcaption><a href="https://pixabay.com/illustrations/web-network-globe-continents-3850917/">Image by geralt via Pixabay</a></figcaption></figure><p>In the post-pandemic world, <strong>the roles that tech workers have to undertake are ever-expanding.</strong> While there is increasingly more technological automation that could replace lower-entry tech jobs — such as machine learning data labelers — <strong>the need for workers to build and maintain such technology is still on the rise.</strong></p><p>This is especially true for industries that are slow to implement automation.</p><p><strong>What can we learn as tech workers?</strong></p><ol><li>Tech workers <strong>will not be fully replaced</strong> — especially not anytime soon.</li><li>It pays <strong>to embrace lifelong learning and adapt to the current dynamic environment.</strong></li><li>Depending on your technical role, you may find yourself in a<strong> rapidly progressing environment</strong> or <strong>one that is deteriorating</strong> —<strong> it is up to you to pivot.</strong></li></ol><h3>What Businesses Can Learn From This</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/552/1*mfiEf8NSZFRgxJZyVp-VYg.png" /><figcaption><a href="https://pixabay.com/vectors/store-online-ecommerce-shopping-4156934/">Image by Megan_Rexazin from Pixabay</a></figcaption></figure><p>The massive layoff of Meta provides great lessons for businesses. Some also speculate that the layoff has to do with Meta’s shift in business focus.</p><ul><li>Indeed, the strategic business shift from Meta to focus on the Metaverse did cause a lot of controversies.</li><li>With many article titles such as <strong><em>“Meta: Why The Metaverse Will Probably Fail” </em></strong>and <strong><em>“3 Reasons Meta Platforms Is Not the Future of the Metaverse”</em></strong>, coupled with the recent layoff, it can appear that the future of Meta is bleak.</li><li>However, <strong>just how much do we understand the vision behind <em>the Metaverse</em>?</strong></li></ul><p>It may be worth noting that similar comments were made about Amazon during its early days, only to end with <strong>Warren Buffett admitting his misjudgment of Amazon.</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/448/1*l1bFEXOFgMmW_9wwiVM3Rg.jpeg" /></figure><blockquote>“I was too dumb to realize what was going to happen…But I did not think [Jeff Bezos] was going to succeed on the scale that he has.” — Warren Buffett</blockquote><p><strong>Regarding Business Opportunities…</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*pWViciss0Ck8YI6w.jpg" /><figcaption><a href="https://pixabay.com/photos/doors-choices-choose-decision-1690423/">Image by qimono from Pixabay</a></figcaption></figure><ul><li>Even legendary investors like Warren Buffett have moments of misjudgment.</li><li>Learning from this, it is helpful to understand that the technology surrounding the Metaverse — like VR/AR headgear and blockchain — is still far from being mature.</li><li>Without sufficient strong reasons why the Metaverse would fail, it may be desirable to maintain a more neutral (or perhaps slightly optimistic) outlook so that your <strong>business can leverage the opportunities when they do arise.</strong></li></ul><p><strong>Regarding The Layoff…</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/1*pucBOIJirKa7usfog8HvgQ.jpeg" /><figcaption><a href="https://pixabay.com/photos/we-are-hiring-hiring-recruitment-2578901/">Image by islandworks on Pixabay</a></figcaption></figure><ul><li>As mentioned earlier, the pandemic has accelerated e-commerce and other online ventures, <strong>increasing the need for organizations to rapidly hire more tech workers.</strong></li><li>However, knowing that the post-pandemic era would ultimately come, we should try to build a <strong>healthy, conservative hiring standard.</strong></li><li>One way to do so is by <strong>creating flexible short-term job contracts.</strong></li></ul><h3><strong>The Case for Founders</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/254/1*OCbsalW4VR3nKP4CNo9P-g.png" /></figure><p>Not surprisingly, Meta’s story and layoff is also a great case study for founders — in particular, we can learn about <strong>how founders should make product decisions.</strong></p><ul><li>The founder of SavvyCal (a software company with 7-digit annual revenue), Derrick Reimer, shared his thoughts about making difficult product decisions in a startup podcast.</li><li>There, he cautioned founders of <strong>becoming too myopic in heeding the market advice </strong>— be it from investors or customers — and echoing the thought that one should always keep a <strong>“founder filter”.</strong></li></ul><p><strong>The “Founder Filter”</strong></p><blockquote>“Being a brand new baby founder… I’m going to just build everything people want… Just build what your customers want, and then you’ll hit product market fit. Because I don’t think that’s actually true. You have to have that filter — that founder filter — to keep that vision in line”</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*fjeopzsC98C_PYJo.jpg" /><figcaption><a href="https://pixabay.com/photos/color-black-white-filter-hand-2246333/">Image by yusufk53 on Pixabay</a></figcaption></figure><p>Going against traditional advice, the “Founder Filter” advice may seem <strong>highly unintuitive</strong>. However, it seems to be how companies like Apple come up with outstanding innovations like iPhone.</p><p>For founders, then, it may be wise to consider striking a desirable balance between listening to your customer and maintaining your “<strong>founder filter”.</strong></p><h3>Summary and Key Takeaways</h3><p>Is Meta on the brink of collapsing? Or will it rise above its competitor and create a new future for all of humanity?</p><p>It seems like only time will tell. But in the meanwhile, <strong>let us — tech workers, businesses, and founders — learn from its journey and do better, be better.</strong></p><h4>Key Takeaways</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*qs0y5-9EbaiHdAf3.jpg" /><figcaption><a href="https://pixabay.com/photos/vintage-map-compass-atlas-coins-4896141/">Image by Ghinzo on Pixabay</a></figcaption></figure><p><strong>For Tech Workers</strong></p><ul><li>embrace lifelong learning and adapt to the current dynamic environment.</li><li>identify if you are in a progressive environment or a deteriorating environment, and pivot accordingly.</li></ul><p><strong>For Businesses</strong></p><ul><li>remain neutral (or perhaps slightly optimistic) about novel technologies so you can leverage the opportunities when they do arise.</li><li>try to build a healthy, conservative hiring standard, especially during volatile periods like the pandemic.</li></ul><p><strong>For Founders</strong></p><ul><li>learn to strike a balance between listening to the market and maintaining your “founder filter”.</li></ul><p>I hope you gleaned some new insights from this article. Do drop a follow if you like what you’re reading :)</p><p>I’ll catch you in the next article. Cheers!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=19acac036e6f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/geekculture/meta-layoff-what-tech-workers-businesses-and-founders-can-learn-from-this-19acac036e6f">Meta Layoff: What Tech Workers, Businesses, and Founders Can Learn From This</a> was originally published in <a href="https://medium.com/geekculture">Geek Culture</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[4 Cognitive Biases In AI/ML Systems]]></title>
            <link>https://pub.towardsai.net/4-cognitive-biases-in-ai-ml-systems-a208ac87177d?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/a208ac87177d</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[cognitive-science]]></category>
            <category><![CDATA[crystalizemylearning]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[software-development]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Tue, 25 Oct 2022 16:01:29 GMT</pubDate>
            <atom:updated>2022-10-25T16:01:29.995Z</atom:updated>
            <content:encoded><![CDATA[<h4>In the ideal world, machines are unbiased and objective. However, the humans who engineer these machines are inherently biased.</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*rSUBSiq9bauuUBny.jpg" /><figcaption><a href="https://pixabay.com/illustrations/brain-chip-neurons-machine-learning-6010961/">Image by chenspec via Pixabay</a></figcaption></figure><p>With the rapid proliferation of advanced technology, more and more systems are now equipped with artificial intelligence and machine learning algorithms.</p><blockquote>However, can we objectively say that these systems are truly fair and unbiased?</blockquote><p>In this article, I share 4 cognitive biases that AI/ML systems have and how machines could be full of biases because the <strong>engineers who built them are inherently imperfect.</strong></p><h3>Selection Bias</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*b8SuNmD2fUKM4ZVV.jpg" /><figcaption><a href="https://pixabay.com/illustrations/cubes-choice-one-yellow-light-2492010/">Image by qimono via Pixabay</a></figcaption></figure><ul><li><strong>Selection bias</strong> refers to the selection of training/testing data that is <strong>not representative of the entire population.</strong></li><li><strong>Example:</strong> An engineer chooses the <strong>first 100 volunteers</strong> who responded to his email as his training data</li><li><strong>The problem:</strong> The <strong>first 100 respondents</strong> may be more enthusiastic about a product or study than the <strong>last 100 respondents</strong>. By explicitly choosing the first 100 respondents, the engineer<strong> introduces unfairness in his data collection method.</strong></li><li><strong>The solution:</strong> select a <strong>random sample</strong> of 100 users from your pool of email respondents instead of the first 100.</li></ul><h3>Reporting Bias</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*BU47f01wBOYkIZlq.jpg" /><figcaption><a href="https://pixabay.com/photos/board-chalk-feedback-review-3699978/">Image by athree233 via Pixabay</a></figcaption></figure><ul><li><strong>Reporting bias</strong> refers to<strong> </strong>people’s (conscious and unconscious)<strong> tendencies to suppress the information they report</strong>.</li><li><strong>Example:</strong> Many amazon products <strong>have more 5-star and 1-star reviews than 2, 3, or 4-star reviews</strong> because <strong>people who have extreme experiences (either positive or negative) are more likely to post a review than those who have neutral experiences</strong>.</li><li><strong>The problem:</strong> An engineer that uses online reviews as their primary data source may create an AI model that is <strong>great at detecting extreme sentiments but not so much at detecting more neutral, subtle sentiments.</strong></li><li><strong>The solution:</strong> Consider <strong>broadening the data collection scope to account for the underrepresented data.</strong></li></ul><h3>Implicit Bias</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*uWKrIXi9uYs2B65k.jpg" /><figcaption><a href="https://pixabay.com/illustrations/girl-woman-female-silhouette-young-4064552/">Image by Merio via Pixabay</a></figcaption></figure><ul><li><strong>Implicit bias</strong> refers to people’s <strong>unconscious tendencies to make assumptions or associate stereotypes with others.</strong></li><li><strong>Example:</strong> An engineer who is often bitten by dogs believes that dogs are more aggressive than cats, even though that <strong>may not be scientifically true.</strong></li><li><strong>The problem:</strong> The engineer would believe that the <strong>ground truth</strong> is “dogs = aggressive” and thus <strong>fine-tuned her AI model to label dogs as being more aggressive than cats.</strong></li><li><strong>The solution:</strong> As implicit bias is <strong>unconscious to an individual</strong>, having <strong>multiple engineers code the AI</strong> and <strong>establishing proper peer review procedures</strong> would <strong>reduce the occurrence of such biases.</strong></li></ul><h3>Framing Bias</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/566/1*K9MG0LVwpJ-BWVgjabCZpg.png" /></figure><ul><li><strong>Framing bias</strong> refers to<strong> </strong>people’s<strong> tendency to be influenced based on how information is presented.</strong></li><li><strong>Example:</strong> An engineer who sees a <strong>dark and dull</strong> website for a product believes that the product must have poor sales, ignoring the actual positive sales number of the product.</li><li><strong>The problem:</strong> In designing the AI algorithm, the engineer may<strong> take into account subjective variables,</strong> such as the color of a website, <strong>instead of focusing on objective metrics.</strong></li><li><strong>The solution:</strong> <strong>Avoid subjective</strong> (usually qualitative) data and <strong>prioritize objective, factual data</strong> instead.</li></ul><h3>Summary and Best Practices</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/543/0*oQsLizWImscaZahj.jpg" /><figcaption><a href="https://pixabay.com/illustrations/artificial-intelligence-ai-robot-2228610/">Image by Seanbatty via Pixabay</a></figcaption></figure><p>A quick recap of the 4 cognitive biases in AI/ML systems</p><ul><li><strong>Selection Bias (selecting data not representative of the entire population)</strong></li><li><strong>Reporting Bias (people’s tendency to underreport information)</strong></li><li><strong>Implicit Bias (people’s unconscious tendencies to assume)</strong></li><li><strong>Framing Bias (people’s tendency to be affected by how information is presented)</strong></li></ul><p>The 5 best practices to enhance objectivity in AI/ML systems</p><ul><li><strong>Always select a random sample (instead of the first or last hundred data points)</strong></li><li><strong>Verify your data sources through comparison with other data sources</strong></li><li><strong>Assign more (diverse) engineers to develop the AI/ML system</strong></li><li><strong>Establish proper peer review procedures to cross-examine logic and unconscious bias</strong></li><li><strong>Prioritize objective and factual data over subjective (usually qualitative) data.</strong></li></ul><p>As you can see, humans’ perceptions are <strong>deeply flawed and our imperfection may trickle down into the systems we build. </strong>By <strong>acknowledging such biases,</strong> however, we can <strong>optimize the AI systems</strong> we build and <strong>make them more fair and objective.</strong></p><p>I hope you learned something new today. If you like what you’re reading, do drop a clap or a follow!</p><p>I’ll catch you in the next article. Cheers!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a208ac87177d" width="1" height="1" alt=""><hr><p><a href="https://pub.towardsai.net/4-cognitive-biases-in-ai-ml-systems-a208ac87177d">4 Cognitive Biases In AI/ML Systems</a> was originally published in <a href="https://pub.towardsai.net/">Towards AI</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[3 Counter-Intuitive Ideas To Challenge Your Thinking]]></title>
            <link>https://medium.com/illumination/3-counter-intuitive-ideas-to-challenge-your-thinking-e31882eb83a9?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/e31882eb83a9</guid>
            <category><![CDATA[life]]></category>
            <category><![CDATA[self-improvement]]></category>
            <category><![CDATA[crystalizemylearning]]></category>
            <category><![CDATA[productivity]]></category>
            <category><![CDATA[critical-thinking]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Sun, 25 Sep 2022 13:45:04 GMT</pubDate>
            <atom:updated>2022-09-25T13:45:04.853Z</atom:updated>
            <content:encoded><![CDATA[<h4>When intuition fails, how will you react?</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*xHzXEUyODzdD1pwA.jpg" /><figcaption><a href="https://pixabay.com/zh/illustrations/artificial-intelligence-brain-4389372/">Image by geralt from Pixabay</a></figcaption></figure><p>A <a href="https://www.healthline.com/health/how-many-thoughts-per-day">study</a> showed that the average person <strong>has more than 6000 thoughts every day.</strong> That is approximately 4 thoughts every minute, or 1 thought every 15 seconds.</p><p>Despite having thousands of thoughts every day, we hardly could claim ourselves to be deeply creative and innovative.</p><blockquote>Indeed, it is the <strong>quality of our thoughts that counts — not the quantity.</strong></blockquote><p>In this article, I will share <strong>3 unintuitive ideas to challenge your thinking and encourage you to think outside the box.</strong></p><h3>Idea #1 — Hot water turns to ice faster than cold water (Science)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*GH1kt1N_L2pFMPfx" /><figcaption><a href="https://unsplash.com/photos/s5DR64vWQtw">Image by Priscilla Du Preez from Unsplash</a></figcaption></figure><p>Known as the <a href="https://en.wikipedia.org/wiki/Mpemba_effect"><strong>Mpemba effect</strong></a>, the phenomenon that <strong>hot water freezes faster than cold water </strong>was named after <a href="https://en.wikipedia.org/wiki/Erasto_B._Mpemba"><strong>Erasto Mpemba</strong></a><strong>.</strong></p><h4>The Story</h4><ul><li>In 1963 when <strong>Erasto </strong>was a student at Magamba Secondary School, he wanted to sell ice cream to earn some pocket money.</li><li>In the urgency to prepare the ice cream quickly, he <strong>skipped the cooling phase of the preparation</strong> and <strong>placed the hot milk mixture directly into the freezer.</strong></li><li>To his surprise, he observed that <strong>his hot milk mixture froze faster than his classmates’ cooled milk mixture.</strong></li><li>The observation ultimately spun into a research paper and was termed the <strong>Mpemba effect.</strong></li></ul><h4>The Controversy</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*HkB0KN6ElzSs8KIK" /><figcaption><a href="https://unsplash.com/photos/g-fm27_BRyQ">Image by Andrew George from Unsplash</a></figcaption></figure><ul><li>Logically speaking, it makes sense that <strong>cold water freezes faster than hot water since it can skip the cooling process that the hot water needs to undergo.</strong></li><li>With this logic in mind, the Mpemba effect seems <strong>highly unintuitive</strong>.</li><li>Indeed, this phenomenon has caused <strong>significant controversy</strong> in the scientific community.</li><li><strong>Although some very reputable scientific bodies have debunked this theory, the validity of this effect is still largely disputed, </strong>with some attributing the strange phenomenon being observed <strong>due to impurities in the mixture.</strong></li></ul><p>If you’re a chemist, you may feel underwhelmed or disappointed by this idea. <strong><em>“Isn’t this just a failed experiment then?”.</em></strong></p><p>If you took some time to explore the controversy online,<strong> </strong>however,<strong> you may discover a treasure trove of interesting theories proposed by science fanatics that tries to prove this phenomenon right.</strong></p><h3>Idea #2 — More than one average (Math)</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*5kUsoBnMjESvjnxc.jpg" /><figcaption><a href="https://pixabay.com/zh/photos/libra-old-scale-weigh-out-old-2878196/">Image by moritz320 from Pixabay</a></figcaption></figure><p>What is the <strong>average</strong> age of 3 boys aged 10, 15, and 20? Well, you can simply add all their ages up 10 + 15 + 20 = 45 and divide it by 3 — giving you the answer of 15.</p><h4>Now, imagine this weird, complicated scenario:</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*B2clRzGEzsidnKS476_MGg.jpeg" /></figure><p>You have <strong>3 types of storybooks </strong>— <strong>Book A, Book B, and Book C</strong> — of <strong>varying quantities</strong>. You need to find the <strong>average number of words in the storybooks.</strong></p><ul><li>Each Book A has an <strong>average of 100 words/book (quantity = 5, total words = 500)</strong></li><li>Each Book B has an <strong>average of </strong>2<strong>00 words/book (quantity = 10, total words = 2000)</strong></li><li>Each Book C has an <strong>average of 500 words/book (quantity = 15, total words = 7500)</strong></li><li>The total number of words is 500 + 2000 + 7500 = 10000</li><li>The total number of books is 5 + 10 + 15 = 30</li></ul><h4><strong>How would you find the average number of words in the storybooks?</strong></h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/845/1*MesYg8wu6n8Ii6x6tZ3pdw.png" /></figure><p>Here are your TWO options:</p><ol><li><strong>Get the average of the averages (i.e. average of the sum of averages in Book A, B, and C).</strong> <strong>(100 + 200 + 500) / 3 = ~267 words per book</strong></li><li><strong>Get the average by dividing the total number of words by the total number of books.</strong> <strong>10,000 / 30 = ~333 words per book</strong></li></ol><p>Even if you’re not a mathematician, you may be inclined to pick the 2nd option. Perhaps surprisingly to some, <strong><em>both approaches are actually correct.</em></strong></p><h4>The Reason</h4><p><strong>Method 1 — ‘average of the averages’</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*FYNUifVBV3lpI4NVbsWt4g.jpeg" /></figure><ul><li>The term <strong><em>‘average of the averages’ </em></strong>is often frowned upon by mathematicians.</li><li>However, in this case, the term <strong><em>‘average of the averages’ </em></strong>is merely a disguise. The calculation of the <strong><em>‘average of the averages’</em></strong> actually leads to the calculation of <strong>Simple Average, </strong>a perfectly valid way of calculating the average.</li><li>Not to get too <strong><em>math-y</em></strong> here, simple average merely means that you’re taking the average of a particular dataset without considering the <strong>weight (or in this case, the quantity)</strong> of each item group.</li></ul><p><strong>Method 2 — total words / total number of books</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/424/1*4osHpBmFyC9On1dUsPimuA.jpeg" /></figure><ul><li>This is the classic <strong>‘weighted average’</strong> approach and is likely what most people use to calculate averages.</li><li>This kind of average takes into consideration the <strong>weight</strong> (<strong>or quantity</strong>) of each item group.</li></ul><p>Both averages are <strong>perfectly valid</strong> and one may be more suitable than the other depending on the context.</p><p><strong><em>Surprised? Well… the learning point here isn’t about math. It’s about realizing how we could adopt a shift in perspective and still derive the right answer.</em></strong></p><h3>Idea #3 — The Bystander Effect (Psychology)</h3><p><strong><em>Are you more likely to receive help in a room with 3 people, or a room with 30 people? </em></strong>The answer may be surprising to some.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*1vWiW1h82R4ikuLT.jpg" /><figcaption><a href="https://pixabay.com/photos/pedestrians-people-busy-movement-400811/">Image by B_Me from Pixabay</a></figcaption></figure><p><strong>According to the Bystander Effect:</strong></p><blockquote>The probability that a person in distress <strong>will be helped by a bystander is inversely proportional to the number of</strong> bystanders present.</blockquote><p>To put it simply, this means <strong>the more bystanders are present</strong>, <strong>the less likely someone in distress will receive help.</strong></p><h4>Background of The Bystander Effect</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/284/1*xn3Kwar7QmXhgajRQlhUVw.jpeg" /></figure><ul><li>This strange paradox of human behavior was first formulated by psychologists <strong>John Darley and Bibb Latané.</strong></li><li>This social psychological theory was first prompted by a dark incident where it was wrongly reported that <strong>38 bystanders watched passively</strong> during an attack on a victim.</li><li>The theory is then demonstrated and popularized in 1968 and many social scientists soon launch <strong>a series of experiments to test the theory.</strong></li><li>Today, this theory is <strong>widely established and taught in many Psychology classes.</strong></li></ul><h4>Factors causing the Bystander Effect</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*GZcVbbxJIhqBKu7JfWoYEw.jpeg" /></figure><p>There are two major factors that may explain the Bystander Effect.</p><ol><li><strong>Diffusion of Responsibility</strong></li></ol><figure><img alt="" src="https://cdn-images-1.medium.com/max/320/1*xgk0IGrgNFjvRhXOe7I27w.png" /><figcaption><a href="https://pixabay.com/vectors/network-people-business-icon-5508173/">Image by mohamed_hassan from Pixabay</a></figcaption></figure><ul><li><strong>“Someone else will do it” (i.e. someone else will help the person in distress)</strong></li><li>In this case,<strong> the presence of others makes individuals feel less responsible to take action.</strong></li><li>Here, the <strong>responsibility is thought to be shared among those who are present.</strong></li></ul><p><strong>2. Social Influence</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*ny0Sevn0IUQUWked6A9n9w.jpeg" /><figcaption><a href="https://pixabay.com/illustrations/we-the-people-self-quarantine-4983077/">Image by ooceey from Pixabay</a></figcaption></figure><ul><li><strong>“I will monitor how others act before I do anything”</strong></li><li>As humans, we tend to <strong>subconsciously</strong> <strong>adjust our behaviors to fit our current environment.</strong></li><li>In this case, when <strong>those around us fail to react to the person in distress,</strong> we often take that as a <strong>signal that nothing is wrong.</strong></li></ul><p>Both <strong>diffusion of responsibility</strong> and <strong>social influence</strong> makes it <strong>less likely for a person to receive help</strong> in a room with<strong> 3 people</strong> than in a room with <strong>30 people.</strong></p><p>The <strong>Bystander Effect</strong>, a counter-intuitive theory that <strong>the more people are present, the less likely you’ll receive help,</strong> is yet another idea that challenges conventional logic.</p><h3>Summary</h3><p>That’s it! Do you feel weird, surprised, ambivalent, or even confused? If you did, then I’m glad that I have sparked some thoughts in your mind.</p><p><strong>Recap of Our 3 Counter-Intuitive Ideas</strong></p><ul><li><strong>Idea #1 — Hot water turns to ice faster than cold water (Science)</strong></li><li><strong>Idea #2 — More than one average (Math)</strong></li><li><strong>Idea #3 — The Bystander Effect (Psychology)</strong></li></ul><p>I hope you have experienced <strong>some level of paradigm shift</strong> through learning about these thought-provoking ideas from numerous disciplines.</p><p><strong>Keep learning. Keep thinking. Keep improving.</strong></p><p>I’ll catch you in the next article. Cheers!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e31882eb83a9" width="1" height="1" alt=""><hr><p><a href="https://medium.com/illumination/3-counter-intuitive-ideas-to-challenge-your-thinking-e31882eb83a9">3 Counter-Intuitive Ideas To Challenge Your Thinking</a> was originally published in <a href="https://medium.com/illumination">ILLUMINATION</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[3 Ways To Turn Knowledge Into Action]]></title>
            <link>https://medium.com/illumination/3-ways-to-turn-knowledge-into-action-3978282c7493?source=rss-9e666ce9a7ab------2</link>
            <guid isPermaLink="false">https://medium.com/p/3978282c7493</guid>
            <category><![CDATA[self-help]]></category>
            <category><![CDATA[software]]></category>
            <category><![CDATA[crystalizemylearning]]></category>
            <category><![CDATA[self-improvement]]></category>
            <category><![CDATA[productivity]]></category>
            <dc:creator><![CDATA[Lye Jia Jun]]></dc:creator>
            <pubDate>Tue, 30 Aug 2022 08:14:02 GMT</pubDate>
            <atom:updated>2022-08-30T08:14:02.948Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*SrTpSo1In49KdEiL.jpg" /><figcaption><a href="https://www.pexels.com/photo/person-holding-and-reading-book-during-daytime-115001/">Image by Negative Space via Pexels</a></figcaption></figure><p>Do you read a lot of self-help books? Do you watch self-improvement videos frequently to learn how you can better yourself?</p><p>If you do, one difficulty you’d undoubtedly face on this journey is <strong>‘actionability’ </strong>— that is, <em>taking action and applying new knowledge you’ve learned to better yourself.</em></p><p>You see…</p><blockquote>Knowledge is not power.</blockquote><blockquote>Knowledge is POTENTIAL power.</blockquote><blockquote>If you don’t use your knowledge in some way, shape, or form, it is innately <strong>worthless</strong>.</blockquote><p>Yes. Knowledge is <strong>worthless </strong>if you don’t tap on them.</p><p>So what’s stopping us from taking action on our knowledge? Well… there are a couple of reasons. Let’s explore.</p><h3>Problem #1 — We THINK we HAVE TO make really big changes</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*LtrtDYPCiFAQvqqWnGEu2w.jpeg" /><figcaption><a href="https://pixabay.com/photos/to-go-biking-cycle-heaven-5360648/">Image by Muscat_Coach via Pixabay</a></figcaption></figure><p>We often think that <strong>big changes are necessary for us to grow</strong>. However, big changes <strong>require a lot of work</strong>.</p><p>If you put yourself in a position where you can easily get<strong> overwhelmed by the large amount of work</strong>, you’d very likely <strong>give up</strong>.</p><p><strong>Example</strong>: You want to start meditating to clear your mind. You set a goal to <strong>meditate every day for 30 minutes.</strong> However, the <strong>large time commitment quickly overwhelms you</strong> and you <strong>give up on meditation altogether.</strong></p><h3>Solution #1 — Focus on making REALLY small sustainable changes</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/302/1*ykr_X_Mbqb9C_Ayd4mf4ig.jpeg" /><figcaption><a href="https://pixabay.com/zh/photos/quadricycle-child-karting-go-kart-1208708/">Image by kristenliam31 via Pixabay</a></figcaption></figure><p>The solution to this problem requires a <strong>mindset shift.</strong></p><ul><li><strong>Understand and internalize that small changes are enough</strong>. Oftentimes the change you need is <strong>SMALLER than you think.</strong></li><li>Starting small makes it <strong>mentally easy</strong> for you to get started. You can <strong>ALWAYS ramp up your efforts</strong> as time passes.</li></ul><p><strong>These little bits of effort will add up in the long run.</strong></p><blockquote>“Little by little, a little becomes a lot” — A Famous Tanzanian Proverb</blockquote><p><strong>Consider This: </strong>Start by meditating for <strong>3 minutes twice a week</strong>. You can meditate on the weekends when you have more free time. <strong>Once you start to notice the benefits, you can slowly ramp up</strong> your meditation frequency and duration. It’s <strong>much easier </strong>for you to pick up the <strong>habit of meditation</strong> this way.</p><h3>Problem #2 — We don’t have a clear plan</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*avbgULevuz2zLZSNXcsQXA.jpeg" /><figcaption>Credit: Unsplash Images</figcaption></figure><p>Here’s a great example.</p><ul><li>Bob and Alice are both trying to become healthier.</li><li>Bob says that he will <strong>“try to become healthier this year”.</strong></li><li>Alice says that <strong>she will run 3 times a week on Monday, Wednesday, and Friday, from 8 am to 9 am.</strong></li></ul><p>Who do you think is <strong>more likely to accomplish their goal?</strong></p><p>They say that <strong><em>“ambiguity is the mother of procrastination”</em></strong> and that couldn’t be closer to the truth.</p><h3>Solution #2 — Make a plan and be specific</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*Tx8dIiI2X286lZ7TPlJmQQ.jpeg" /><figcaption>Credit: Unsplash Images</figcaption></figure><p>To eliminate ambiguity, you <strong>need a plan</strong>. In general, <strong>the more specific and the more clear you are about your plan the better.</strong></p><ul><li><strong>What’s the first task</strong> you will tackle when you wake up?</li><li><strong>How long will you spend</strong> on your afternoon workout?</li><li><strong>Where will you do</strong> your evening meditation?</li></ul><p>The goal is to<strong> turn your desired outcomes into a series of very clear behaviors.</strong></p><h3>Problem #3 — We don’t remember to do it</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*8wKgSIGjKXWbH1_dXQdBng.jpeg" /><figcaption><a href="https://unsplash.com/photos/so1L3jsdD3Y">Image by Robina Weermeijer via Unsplash</a></figcaption></figure><p>There are <strong>many golden nuggets of wisdom</strong> in self-help books and yet we<strong> struggle to remember</strong> most of them.</p><p>Even if we somehow remember the lessons vaguely, we <strong>fail to act on them at the right times.</strong></p><ul><li>A lesson about leadership may <strong>only be relevant when you are in a leadership position.</strong> However, you get so <strong>overwhelmed </strong>while leading that you <strong>fail to use</strong> the leadership tips you learned in a book.</li><li>Your breathing exercises will <strong>only come in handy when you’re in high-pressure situations. </strong>However, when you’re stressed, you <strong>fail to remember</strong> the breathing exercises that are supposed to calm you down.</li></ul><h3>Solution #3 — Take notes and set triggers to remind you</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/432/1*BdxLtA5zs18YttS9FenNcg.jpeg" /><figcaption><a href="https://unsplash.com/photos/5bYxXawHOQg">Image by David Travis via Unsplash</a></figcaption></figure><p>The first thing you need to do is realize the <strong>fallibility of human memory.</strong></p><ul><li>To remember the lessons from self-help books, <strong>you NEED to take notes.</strong></li><li>Taking notes not only helps you to <strong>remember key ideas</strong> from books you’ve read, but it also <strong>saves you time</strong> from flipping through the book to search for ideas you vaguely remember.</li></ul><p>After keeping track of these valuable lessons, you’d want to<strong> set some triggers to prompt yourself to take action on them.</strong></p><ul><li><strong>A trigger is any stimulus that influences behaviors (similar to a cue)</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/360/1*8ZWESyet8E1bnjiXX3yl-A.jpeg" /><figcaption>Credit: Unsplash Images</figcaption></figure><ul><li><strong>Time-based trigger:</strong> It’s 7 am — my alarm rang so I will get up to head for work</li><li><strong>Location-based trigger:</strong> I passed by the park so I will slow down, try to be more present, and take deep breaths to relieve my pressure</li><li><strong>Emotional State Trigger:</strong> I feel stressed so I will sit down and meditate</li></ul><p><strong>Ultimately, writing notes allows you to track your knowledge, and setting triggers</strong> <strong>reminds you to take action on them.</strong></p><p>That’s all! These are the <strong>3 problems making it difficult for you to take action on your knowledge</strong>, and <strong>3 ACTIONABLE solutions you can implement right now.</strong></p><p>You see, the self-improvement journey is one that will be full of ups and downs.</p><h4>the truth is…</h4><blockquote><strong>actionability will always be a struggle</strong></blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/0*kh3pKv_JaCjEwz1R" /><figcaption><a href="https://unsplash.com/photos/12yvdCiLaVE">Image by christopher lemercier via Unsplash</a></figcaption></figure><ul><li>Struggling to <strong>take action on new knowledge I’ve learned</strong> is a challenge that is very <strong>close to my heart.</strong></li><li>This is also what <strong>motivates me to search for solutions to this problem</strong>, though I haven’t had any luck so far.</li></ul><p>Ultimately, with my good friend, I decided to build an app to solve this very issue.</p><p>The app we built — <a href="https://journeyos.com/"><strong><em>JourneyOS</em></strong></a><strong><em> </em></strong>— is <em>an all-in-one </em><strong><em>note-taking</em></strong><em> and </em><strong><em>task management app</em></strong><em> that aims to </em><strong><em>bridge the gap between knowledge and action.</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xRoP_9xdUCHQ2CKu.png" /></figure><p><strong><em>JourneyOS</em></strong> is something I’m experimenting with, and it is <strong>currently in beta. </strong>We are focusing on <strong>getting feedback from a small pool of users.</strong></p><p>If you clicked on this article, I believe <strong>you too struggled with translating knowledge into action</strong>, or at least, <strong>are interested in using your newfound knowledge to become a better person.</strong></p><p><strong>If my message resonates with you and if you believe in our vision</strong>, I warmly welcome you to <a href="https://journeyos.com/"><strong>join <em>JourneyOS’s</em> beta waitlist here</strong></a><strong>.</strong> I’d sincerely appreciate any feedback too!</p><p><strong>To wrap up this article, I’d like to leave you with a quote:</strong></p><blockquote><strong>Knowing is not enough; we must apply. Willing is not enough; we must do.</strong></blockquote><p>I hope you found some value in this article.</p><p>Take care and I’ll catch you in the next one. Cheers!</p><h4>Summary</h4><ul><li><strong>Problem #1 — We THINK we HAVE TO make really big changes</strong></li><li><strong>Solution #1 — Focus on making REALLY small sustainable changes</strong></li><li><strong>Problem #2 — We don’t have a clear plan</strong></li><li><strong>Solution #2 — Make a plan and be specific</strong></li><li><strong>Problem #3 — We don’t remember to do it</strong></li><li><strong>Solution #3 — Take notes and set triggers to remind you</strong></li></ul><p><a href="https://journeyos.com/"><strong><em>Click here</em></strong></a><strong><em> to join the waitlist for JourneyOS </em></strong>— an<em> all-in-one </em><strong><em>note-taking</em></strong><em> and </em><strong><em>task management app</em></strong><em> that aims to </em><strong><em>bridge the gap between knowledge and action.</em></strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3978282c7493" width="1" height="1" alt=""><hr><p><a href="https://medium.com/illumination/3-ways-to-turn-knowledge-into-action-3978282c7493">3 Ways To Turn Knowledge Into Action</a> was originally published in <a href="https://medium.com/illumination">ILLUMINATION</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>