<?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 Steven Willmott on Medium]]></title>
        <description><![CDATA[Stories by Steven Willmott on Medium]]></description>
        <link>https://medium.com/@njyx?source=rss-2408aa3ac35a------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*MZ072AHN7bxHgYTV.jpg</url>
            <title>Stories by Steven Willmott on Medium</title>
            <link>https://medium.com/@njyx?source=rss-2408aa3ac35a------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Tue, 17 Nov 2026 16:24:12 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@njyx/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[The Generative AI Flywheel]]></title>
            <link>https://njyx.medium.com/the-generative-ai-flywheel-4dc85566367?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/4dc85566367</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[openai]]></category>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Sun, 15 Jan 2023 03:55:01 GMT</pubDate>
            <atom:updated>2023-01-15T04:55:37.390Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d6uF7j-CD1sIpt-sPMU7dg.png" /></figure><p>Viet Le’s post earlier this week on <a href="https://vietle.substack.com/p/defensible-machine-learning">what it takes to build a defensible machine-learning based company</a> is a great read. There’s a lot in the post, but I’ll pick out three key points that will be most important to anyone looking to build something useful and valuable with generative AI:</p><ol><li>The core of what is being delivered has to be a value driving product. In other words, it’s the value delivered by the service that’s important, not whether or not it uses AI.</li><li>The products that will succeed in the current cycle are those which have an interface with the user (and, by inference, a relationship with the user). This is in part, because many companies will have access to the same foundational models and in part because this is the best place to understand user queries and their satisfaction with responses.</li><li>It’s critical to be able to capture data and models (or, said another way: you need a “<a href="https://twitter.com/karpathy/status/1599852921541128194?lang=en">data engine</a>”). This means that as your service does its work, it is critical to capture what works, what does not and continue training models.</li></ol><p>Almost all good product development and growth to scale is the result of a well-thought-out instance of a <a href="https://en.wikipedia.org/wiki/Good_to_Great">Jim Collins flywheel</a>:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/653/0*LJMxLF5RaeNmYQzE" /></figure><p>The key is that the value necessarily drives usage, which necessarily drives improvement, and with each loop around the flywheel, momentum and usage grows. In a bakery, this would be innovating on the formula for a particular cake. The better it gets, the more people talk about it and come to buy. The more people buy, the more can be invested in experimenting with the recipe and improving the ingredients.</p><p>In a standard tech product, the cycle revolves around solving a particular use case well, attracting the first customers, learning from their needs, and using the revenue from their usage to improve and expand their usage. If the product has a network effect, such as a messaging network, the product effectiveness improves just with an increase of users, with no product change even needed.</p><p>In a generativeAI context, what does it take to deliver value over the long term? The flywheel looks a little like this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/698/0*lKpZzud6mU9XLZEK" /></figure><p>The key point here is that there must be some genuine utility in the service that an audience is willing to ultimately pay for. Then flywheel improvement may well come from product feature improvements. However, more critically, it comes from capturing user input data and training the models which provide the service. The faster this flywheel spins, the better data and models will become. This benefits the users and will likely make those companies that are fastest at tapping into user demand (assuming they are able to build the right data engine) runaway leaders in their categories.</p><p>It’s important to note the difference between this user-driven model building and the model training which is being done by the likes of OpenAI and others at the foundational level. Foundational models are hugely expensive to train (though <a href="https://the-decoder.com/training-cost-for-stable-diffusion-was-just-600000-and-that-is-a-good-sign-for-ai-progress/">Stable Diffusion initial training costs were potentially only around $600,000</a>, which is much less than what one would expect). Even at Stable Diffusion’s potentially low price point, given the skills needs, it is very unlikely it makes sense for most companies to build their own foundational model.</p><p>What many of the foundational models and services (OpenAI, Midjourney, Dale-E, Stable diffusion, etc.) are now enabling is the ability for many other companies to make that first step in value on the flywheel and begin serving a need. The next steps are up to you though:</p><ul><li>Deliver specific value.</li><li>Reach people in order to use it.</li><li>Learn from that usage to improve your own models.</li><li>Build a data / learning engine which drives the flywheel as fast as possible.</li></ul><p>In order to develop something really valuable, the long term objective has to be to reach more and more users and grow/learn the best models.</p><p>As a side note: <a href="https://vietle.substack.com/p/defensible-machine-learning">Viet Le posits</a> that Google Search will lose much in the new era of ChatGPT since it is also going to be strong in such AI solutions. I think the t <a href="https://area67.org/2023/01/07/chatgpt-how-would-you-replace-google-search/">hreat is bigger that it may it appear</a> and it’s more to do with the risk that interfaces (and hence relationships with users) may now fragment more.</p><p>Images generated by Midjourney AI:</p><ul><li>/image a tiny robot doing a benchpress with a city backdrop –aspect 3:2</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/80/0*GbZId1FQx_ertqct" /></figure><p>Senior Director at RedHat Inc. Ex. CEO of 3scale Inc. Concerned citizen of the world. <a href="https://area67.org/author/njyx/">View all posts by njyx</a></p><p><strong>Published</strong></p><p><em>Originally published at </em><a href="https://area67.org/2023/01/14/the-generative-ai-flywheel/"><em>https://area67.org</em></a><em> on January 15, 2023.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4dc85566367" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ChatGPT: How would you replace Google Search?]]></title>
            <link>https://njyx.medium.com/chatgpt-how-would-you-replace-google-search-5cd83740584e?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/5cd83740584e</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[chatgpt]]></category>
            <category><![CDATA[google]]></category>
            <category><![CDATA[search]]></category>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Sat, 07 Jan 2023 18:59:00 GMT</pubDate>
            <atom:updated>2023-01-09T01:25:59.821Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qRcEtyGN76eX_EKHacs7sg.png" /></figure><p>Could Google’s grip on search be under threat by an AI chat system such as <a href="https://openai.com/blog/chatgpt/">ChatGPT</a> (or even more immediately ChatGPT itself)? The idea first popped up in early December ( <a href="https://www.nytimes.com/2022/12/21/technology/ai-chatgpt-google-search.html">here is an NYT article for example</a>, and it also popped up in some form in <a href="https://avc.com/2023/01/what-will-happen-in-2023/">Fred Wilson’s 2023 tech predictions</a>). This week got even spicier with the launch of a several quick-to-market “search engines” that include ChatGPT: <a href="https://www.perplexity.ai/">Perplexity.AI</a> being the first one I saw.</p><p>There is also a Chrome Extension to put ChatGPT results alongside any search engine you use (I’m not linking to it since it’s an unknown developer, and you should do diligence before installing something which sends your search queries anywhere! You can find it in the chrome extensions store)</p><p>Lastly, there have been reports that Microsoft is considering adding ChatGPT results to Bing (reported <a href="https://searchengineland.com/microsoft-to-add-chatgpt-features-to-bing-search-390936">here</a> and <a href="https://www.theinformation.com/articles/microsoft-and-openai-working-on-chatgpt-powered-bing-in-challenge-to-google?rc=xe76d1">here</a>), or even Word. However, they have not been substantiated and this could just be an extrapolation of the hypotheticals people have been spinning.</p><p>So what gives? Is this really a threat to Google? To search in General? Would it work?</p><p>As I already said in my <a href="https://area67.org/2023/01/01/ai-and-creativity-predictions-for-2023/">2023 Generative AI predictions piece</a>: I don’t think this is an immediate threat to search engine dominance, but it does pose future questions. So let’s dig in.</p><h3>What ChatGPT thinks</h3><p>Before we start to pick the topic apart, the obvious place to uncover any master plan is with ChatGPT itself — how would it take over? If you ask Chat GPT the question in the obvious form, you get a very general answer.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/840/0*n36v2NVQmqMqjFC4" /></figure><p>All these things are true, but we’re not building a search engine. We’re trying to replace one.</p><p>If you as a more specific question, you get this:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/816/0*qFG7bCq9JIIGkyil" /></figure><p>This is a solid answer and hints at some important points. The first being that it would be a requirement to connect to (and continuously train on) an extremely strong search index (which Google has and few others do). The response also misses the point a little in that it responds as though what we want is a list of results. It’s unclear that’s the case, often we may just want “the answer”.</p><h3>Different types of queries</h3><p>To really understand how ChatGPT might replace search it helps to look at how it actually works and (more importantly) what types of search use-cases it might replace. A search engine to date can best be thought of as a giant database of all the web links in the world, each with abundant data stored that has been pulled from the link (descriptions, headers text, as well as an overlay of powerful meta and ranking data that defines how relevant a given link is to a given query.</p><p>ChatGPT on the other hand can be thought of as a powerful transformer that has within it a large encoded set of knowledge items (often gleaned from web pages). On top of this, it has its own “learned” mapping that can craft answers it believes relevant to a query based on encoded knowledge.</p><p>In many ways these things are similar. However, the critical difference is that the search engine is returning the most relevant pieces of knowledge, whereas ChatGPT is returning a synthesis of those most relevant pieces of knowledge.</p><p>Another key difference is that a search engine typically ingests and ranks new web pages as soon as they appear, and systems such as ChatGPT require quite some time to encode new knowledge and update their models.</p><p>These differences are particularly relevant when thinking about the types of queries one might want to make of a search engine or an AI system. Here’s a rough breakdown. Queries could be:</p><ul><li><strong>Generative</strong>: A creative output which a human user can use as inspiration or a start point but (crucial) can validate themselves. In other words something like a creative story, or even a block of code to do something.</li><li><strong>Factual</strong>: A query about a world fact. Anything from a query about physics such as “what is the boiling point of Gold?” (2700–2800 Degrees C if you’re interested) or which ferry companies operate between Spain and Morocco?</li><li><strong>Opinion</strong>: A question about what is “best” product or solution for a given problem. An example here would be “What are the best adidas shoes for trail running?”</li><li><strong>Give me a link</strong>: A query to find a link to a service (known or unknown) that you want to use. An example here would be “Spotify” or “United Airlines.” In this case, the intent is often to get a link to web site for the service so you can interact with that provider.</li><li><strong>Current affairs</strong>: Questions about things happening in the world now, such as “Will Kevin McCarthy become the next US House of Representatives speaker?” (which as of today January 7th, 2023 is an unresolved question). [Update: it’s seems the answer turns out to be “<a href="https://www.theguardian.com/us-news/2023/jan/07/kevin-mccarthy-republican-house-speaker-congress">yes</a>”.]</li></ul><p>One could come up with other classifications or types, but these are quite illustrative for now. Each of these queries is quite different in intent and, in fact, needs different types of capabilities to answer it well.</p><h3>How do Google and Chat GPT compare on these types queries?</h3><p>For <strong>factual</strong> / <strong>opinion</strong> / <strong>give me a link</strong>, and <strong>current affairs</strong> queries, Google is the gold standard we know. Other search engines such as Bing do also deliver much the same result but Google’s dominance from the late 90s onwards and its continual expansion since give it an 84% market share in terms of search queries ( <a href="https://www.statista.com/statistics/216573/worldwide-market-share-of-search-engines/#:~:text=Global%20desktop%20market%20share%20of%20search%20engines%202015%2D2022&amp;text=As%20of%20December%202022%2C%20online,market%20share%20was%202.55%20percent.">Statista</a>). As we’ll see below though: it’s often shocking to see just how heavily monetized search queries on Google’s now are. The days o Google returning a clean list of options have long gone.</p><p>Notably, for the “generative” query, Google simply isn’t set up to provide a useful answer. Here’s the result for “Describe a walk in Central Park”:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/813/0*nplEGQ4lZ19LkYS6" /></figure><p>This provides some useful information if you’re actually considering making that walk yourself, but it won’t help you produce a fresh fictional description of such an event. In this category ChatGPT clearly gives something meaningfully different:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/814/0*MhFiAOEdgc-KM2Mz" /></figure><p>Modifying that query could give you many interesting variations. Here is “Describe an Ice Cream Van driving around Central Park”:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/784/0*fJJG43AYBM7lZj8x" /></figure><p>But ChatGPT does know its history. For example “Describe Abraham Lincoln talking a walk in Central Park”:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/801/0*mIJfaHx1aTZ5vpnV" /></figure><p>These generative queries are clearly ChatGPTs forte and a large part of the “wow” effect of using this technology. They are queries which have not been possible before at this level of depth and sophistication.</p><p>How about the other queries then?</p><p>This is where things get more interesting since the other types of queries ARE what Google and other search engines are used for today.</p><p>There are really two questions here: 1) can ChatGPT produce meaningful responses? and 2) could ChatGPT do so consistently in a way that could be trusted?</p><p>In answer to question #1, ChatGPT does surprisingly well at the next three query types: <strong>factual</strong>, <strong>opinion</strong> and <strong>give me a link</strong>.</p><p>The <strong>factual</strong> query: “What is the melting point of Gold?” and “Which ferry companies operate between Spain and Morocco?”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/815/0*fL5_ZCA5IAbNqWYh" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/846/0*GHz1rar_kdEio4lV" /></figure><p>The <strong>opinion</strong> query: “What are the best Adidas shoes for trail running?”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/830/0*_--1Zqi1Vasyc9ue" /></figure><p>In each case, these results are on-point, concise, and useful. Google’s search result for “What is the boiling point of Gold?” also includes a factual summary at the top. It mentions a different number, but they are close.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*VbnPmUGbSyI0VRy1" /></figure><p>In comparison, Google’s response to the running shoes query is enough to make anyone cry:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/945/0*uQglDXwtIqR0LhVY" /></figure><p>A large percentage of the, above the fold, screen is concerned not with “which is best” but where to buy and for how much. There are more articles below comparing products, but there are also heavily interlaced with links which purely focus on the purchase. Unsurprisingly Google monetizes these types of opinion queries heavily.</p><p>The last type of query ( <strong>current affairs</strong>/ <strong>news</strong>) is simply out of scope for ChatGPT:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/829/0*PbRhfomTc3hIrQEp" /></figure><p>I guess Google (and CNN) can breathe a sigh of relief here. This illustrates that, for now at least, the AI models that power something like ChatGPT are slower to update than a search index. This will likely be true for quite some time, but you can bet that with the right heuristics and filters, updates which “patch” a model will start to cut down this time lag.</p><p>Reading all this, it’s clear why there is interest and excitement about ChatGPT changing our experience of the search for answers.</p><p>Before jumping to this conclusion though, it’s important to think about that second question “could such results be produced consistently in a way that could be trusted?”. This is really where things start to get murkier. On the surface, both systems are providing answers to questions. However, underneath, they are doing this in different ways:</p><ul><li>Goggle surfaces the links for a human to potentially check and determine which “answer” is most useful. Often this means one has a view on the credibility of a source for an answer.</li><li>ChatGPT returns an answer that it determines to be relevant. However, it is much harder to know if this answer is truly correct or just a “best guess”.</li></ul><p>For certain classes of problems, this may not matter greatly. It certainly doesn’t matter a great deal for generative queries, since a human is likely on hand to validate or score whether an answer is good. It also likely doesn’t matter to much for very common facts (e.g. the boiling point of water) which ChatGPT likely will always get right. Low risk situations (e.g. estimate the distance to the moon — probably low risk unless you are Nasa) trivia queries are probably also fine. It gets more problematic however for queries with high risk such as: “What supplies should I take on a hike in death valley?”. For the record, the ChatGPT response for the death valley query is:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/822/0*cBygPerG5iVv8sRf" /></figure><p>Pretty good, but you might want to check some official sources before setting out!</p><p>The accuracy of results also begins to be more critical in “opinion” and “link” queries. For the former, it’s important to see where a particular opinion set comes from. How neutral is it really? What bias does it have? A trusted running website may be a better bet and it’s biases will be clear. For Link queries, ChatGPT clearly isn’t optimized for this. A link returned may be right for most queries, but a selection of options might be more appropriate so I can quickly choose the most relevant.</p><p>This concern with accuracy isn’t just a hypothetical or something which can be “addressed” later. The way generative AI models work is that they encode knowledge and then navigate it as they process queries, integrating it in a synthesis. It will be extremely hard to ensure any such model is truly “accurate” for 100% of the forms of any given query. OpenAI’s Sam Altman said this well himself:</p><p>ChatGPT is incredibly limited, but good enough at some things to create a misleading impression of greatness.</p><p>it’s a mistake to be relying on it for anything important right now. it’s a preview of progress; we have lots of work to do on robustness and truthfulness.</p><p>- Sam Altman (@sama) <a href="https://twitter.com/sama/status/1601731295792414720?ref_src=twsrc%5Etfw">December 11, 2022</a></p><p>That work on robustness on trust will involve a lot of work on ensuring certain query types are safe. For many common queries, this may not be so much different to the pre-compiled answers Google provides. However, for the vast majority of long-tail queries, it is likely going to be better to think of ChatGPT as a very smart human who might be wrong about things and could certainly be biased. There will be many cases where getting a list of the “authoritative” written sources on a topic will be more valuable.</p><p>ChatGPT does well on many query types, and the form factor of its results can be incredibly appealing. So … will it replace Google?</p><h3>Replacing Google Search</h3><p>The central threat to Google’s business model that’s been talked about is that if a chat system can answer a question instantly, and without the need to click on any links, no links need to be shown and, thus, there is no room for ads to be shown.</p><p>There is clear appeal to this, and for some query types, ChatGPT responses are definitely appealing. In the short term though, it seems highly unlikely ChatGPT will outright replace Google searches.</p><ul><li>Google is not only embedded in many devices; it is the default from most user search behavior.</li><li>ChatGPT results, while appealing, often lack the next step: “take me to the site to do what I need to do”</li><li>Accuracy is a concern for many query types.</li><li>A deep search index and knowledge base is still needed underneath to train models. Very few organizations have this.</li><li>Certain types of queries (e.g. search) remain weak or impossible.</li><li>Right now, ChatGPT is free. Anyone that would like to use it at “Google Scale” would have to determine how to monetize sufficiently to defray the server costs. This may (ironically) mean bringing back ads.</li></ul><p>Search company Algolia also has a <a href="https://www.algolia.com/blog/ai/why-chatgpt-wont-replace-search-engines-any-time-soon/">useful post on why certain query types are difficult</a>. This long road. However, is not impossible, and ChatGPT technology certainly should give Google pause for thought. In particular:</p><ul><li>It is another step in the shift to people wanting different formats for search results. In many cases, it really is better to just “get an answer.” Similar to how queries to Amazon’s Alexa skip ad content to give just one answer, user expectations are changing.</li><li>Other large search engine providers certainly could jump on the technology and use it to make market share inroads. Either from pure hype (“now with AI”) or by providing genuine utility.</li><li>The new use-cases, such as generative queries, could become important enough that they become a gateway for people using ChatGPT for some proportion of the queries they would normally send to Google.</li></ul><p>I would expect Google to also begin “adopting” similar features for its search to get ahead of possible hype from other competitors. Perhaps this will be as simple as replacing it’s current canned responses to common queries, or a fully fledged ChatGPT clone.</p><p>None of this is likely to happen fast in terms of adoption by the general public. Search habits are deeply embedded and Google also has a powerful advantage in being the incumbent search engine on IOS today, and in controlling Android. However, there will no doubt be a lot more “ChatGPT for search” stories for a while.</p><p>The technological achievement OpenAI has made is truly remarkable. The fact that we’re even talking about disrupting search is indeed impressive.</p><p><em>Originally published at </em><a href="https://area67.org/2023/01/07/chatgpt-how-would-you-replace-google-search/"><em>https://area67.org</em></a><em> on January 7, 2023.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5cd83740584e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI and Creativity : Predictions for 2023]]></title>
            <link>https://njyx.medium.com/ai-and-creativity-predictions-for-2023-1c67425fcbb1?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/1c67425fcbb1</guid>
            <category><![CDATA[midjourney]]></category>
            <category><![CDATA[predictions]]></category>
            <category><![CDATA[generative-ai]]></category>
            <category><![CDATA[2023]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Sun, 01 Jan 2023 18:38:38 GMT</pubDate>
            <atom:updated>2023-01-01T18:48:30.010Z</atom:updated>
            <content:encoded><![CDATA[<h3>AI and Creativity : Predictions for 2023</h3><p>The New Year is upon us, and so is the time to think about what’s coming next in the world. 2022 has been an incredible year for breakout technology (from <a href="https://arstechnica.com/science/2022/07/deepmind-research-cracks-structure-of-almost-every-known-protein/">gene folding</a> to bringing the fail whale back to Twitter!).</p><p>For AI, in particular, it marks a year where generative AI has really hit mainstream consciousness. <a href="https://www.midjourney.com/">Midjourney</a> reached the Discord server limit and onboarded more than 1million users. <a href="https://www.nytimes.com/2022/12/07/style/lensa-ai-selfies.html">Lensa</a> turned people into new versions of themselves (sometimes not entirely appropriately). Most dramatically of all <a href="https://openai.com/blog/chatgpt/">ChatGPT</a> took the world by storm in December with its highly believably generated text responses to almost any question.</p><p>Using AI for creativity is the core of what we’re doing at <a href="http://timewarp.com/">Timewarp</a>, and it’s been an interest of mine for many years, so seeing all the end-of-year AI predictions flying around, I figured I’d give my own take on what 2023 is going to bring us.</p><p>Let’s get started!</p><h3>#1. Generative AI will continue to generate headlines</h3><p>There is something particularly fascinating about a machine producing a fully formed text or creative work. The results pull mental strings because the pieces themselves are often powerful. Using Lensa to produce variants of our own faces is shocking because we see ourselves staring back. Perhaps this is what Kings and Queens from previous centuries felt when seeing the results of long sittings for a portrait.</p><p>The explosion in the usage of these tools will continue to make them better and most likely more surprising. As millions of transformations are generated by users and subsequently scored, the algorithms learn and improve. By their nature, these generative processes also explore the space of possibilities in such a way that they’ll often generate things a single user doesn’t expect (see a discussion of this <a href="https://area67.org/2022/12/11/bounded-rationality-what-generative-ai-is-really-doing-for-you/">here</a>).</p><p>In addition to the shock headlines, it’s likely that we’ll see technical improvements. This will likely mean some popular audio, video and text content at the very least, will start to contain AI-generated content. Each small piece raising a “XYZ using AI generated content” headlines.</p><p>All this will mean that there’s unlikely to be a let up in the news articles. I’d expect generative AI related news to be 3–5x in 2023 than it was in 2024.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*rdPQ6v92Y6CiQ24k" /></figure><p><em>(Generative AI as a term has only gained traction in the last few months, but the main well known tools have each generated massive headlines. </em><a href="https://trends.google.com/trends/explore?q=%22generative%20AI%22,midjourney,Stable%20diffusion,DALL-E&amp;hl=en-US"><em>https://trends.google.com/trends/explore?q=%22generative%20AI%22,midjourney,Stable%20diffusion,DALL-E&amp;hl=en-US</em></a><em>)</em></p><h3>#2. People will generally understand that the AI is “winging it”. Discrimination of when that’s fine may get lost.</h3><p>Generative AI works by learning relationships between huge numbers of images, texts, prompts, and other relevant media and then using input prompts to explore the space of possible analogous results. This is the equivalent of saying, “what might look like a competent answer to this question based on all the questions and answers I have seen?”. In other words, superficial similarities are being remixed and combined to produce something plausible, but there is no deep structural understanding of the question being asked.</p><p>The resulting effect is very similar to what one might find on a Sci-Fi movie set. From the right angle, the result looks very believable and even spectacular. Behind the scenes though, everything is held together with string, green screens, and duct tape.</p><p>The underlying errors this can create are absolutely fine in some domains (art, music, story-writing, etc.), especially if a human creator is able to generate many variants and fuse the best (using their creative judgement and knowledge of the world). Such errors would be disastrous in other fields (medicine, structural engineering, code), to name but a few. In these instances, the resulting artifacts are potentially so complex that they can’t be validated but look “plausible”.</p><p>The amusing results of Art generators will no doubt amuse us. The idea that we might be blindly using ChatGPT instead of search is altogether more worrying (see below!).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*n0cAVTqkSCWToTZR" /></figure><p><em>(Midjourney: Prompt — “Paris as a movie set with strings and wires holding everything together”)</em></p><h3>#3. Text to Video, Text to 3D.</h3><p>Throughout 2022 we’ve been wowed by prompt-to-image and prompt-to-text applications in particular. We’re now also seeing fledgling new applications in:</p><p>These are much harder problems, and there is also less data available. However, as people play with these tools, they will improve rapidly. This problem is significantly harder than image generation, so in 2023 it’s likely that breakthroughs will be limited to certain types of video and objects. For example, there is already a plethora of “talking head” type video generators (see <a href="https://www.synthesia.io/features/text-to-video?utm_term=text%20to%20video%20ai&amp;utm_campaign=Basic+Search&amp;utm_source=adwords&amp;utm_medium=ppc&amp;hsa_acc=5132031546&amp;hsa_cam=12887270356&amp;hsa_grp=124284584804&amp;hsa_ad=536499098295&amp;hsa_src=g&amp;hsa_tgt=kwd-1056648944536&amp;hsa_kw=text%20to%20video%20ai&amp;hsa_mt=p&amp;hsa_net=adwords&amp;hsa_ver=3&amp;gclid=CjwKCAiA-8SdBhBGEiwAWdgtcJKQ0trGLHL_vIOauAal8rNxYP6QUyuMj1jlLp2knMLZFMI78A25YBoCoAcQAvD_BwE">Synthesia</a>, for example).</p><h3>#4. Prompt Engineering and Prompt Management</h3><p>“What was your prompt for that?” is probably the most common comment in Reddit and Facebook forums for generative art. In other words, “How on earth did you get that?”. Unfortunately, it’s not always apparent to a creator using one of the systems to know exactly why they got a certain result. Running the query again may well generate something very different, particularly if the underlying system is actively learning.</p><p>Prompt engineering is thus already emerging as a clear requirement for using generative AI at scale. One way to think of Art, Programming, or anything else in the far future is “Design the prompt”, that will get you what you need.</p><p>A few places to get started with prompt engineering:</p><p>The challenge, however, is that there is such a huge gap between the abstract nature of the prompt (and the information it contains) to the specificity and detail of the result that a deeper and deeper understanding of the model may be required to produce useful results.</p><p>In addition, we’ll likely need to get much better at “Prompt Management”, which I would term as tracking the prompts used, their results, and using the resulting log to understand the prompts better. We’ll need software solutions to help us do this at scale, especially when teams of people are collaborating and using AI.</p><h3>#5. Search space completion will be feasible in some domains.</h3><p>AlphaFold’s exploration of <a href="https://arstechnica.com/science/2022/07/deepmind-research-cracks-structure-of-almost-every-known-protein/">almost all possible protein combinations</a> is an example of a type of breakthrough that could be tremendous in many fields. By generating vast numbers of combinations and “testing” them, one can eliminate non-viable combinations. While in many fields the number of combinations is so vast there is no way to “map out” the viable combinations. When it does happen though, it’s extremely powerful and essentially represents finding rules to collapse the search space from impossibly large to tractable.</p><p>It remains to be seen how complete AlphaFolds projections are, but if we can find more hidden structure in other places, we’ll see stunning scientific breakthroughs. Adjacent fields to AlphaFold’s current domain will be the first candidates.</p><h3>#6. ChatGPT won’t replace search (yet)</h3><p>One of the most intriguing ideas stemming from ChatGPTs success in November was the notion that it represents an existential threat to Google’s search business. The argument (articulated <a href="https://www.nytimes.com/2022/12/21/technology/ai-chatgpt-google-search.html">here</a>, <a href="https://seekingalpha.com/article/4562719-chatgpt-threat-to-google-search-business">here</a>, and elsewhere) goes that “If ChatGPT simply returns the answer to a query in one-shot, there’s no need for a user to click on a link and hence no opportunity for a search engine to monetize those clicks.”</p><p>In reality, this is already happening (Google and other engines already provide ‘short form’ responses to many popular queries), and there’s no doubt something like ChatGPT could provide a search engine rival a leg-up.</p><p>However, as mentioned in #2, ChatGPT and its like are essentially returning what looks to be “a plausible synthesis of relevant content”. This can often be way off the mark in terms of accuracy. The systems do not have an underlying model of reality that would enable them to tell an obviously false answer from a correct one.</p><p>A further issue is that when it comes to search, knowing which sources are reliable is required information when knowing what to build a model on.</p><p>So it’s unlikely that 2023 will see ChatGPT replacing Google search for common queries.</p><p>In the mid-term, however, a bigger challenge for Google is likely to be that the interface users expect for search queries will change. Why not voice only? Why not a synthesis of the top results? Google will still have a data advantage, but its business model may need to change. That is a massive shift.</p><h3>#7. AIA Act, Algorithmic Accountability should shape thinking.</h3><p>The <a href="https://artificialintelligenceact.eu/">European AIA Act</a> and the <a href="https://www.techrepublic.com/article/know-algorithmic-accountability-act/">American Algorithmic Accountability Act</a> have been creeping along their respective legislative tracks. While it’s still unclear how they will impact the use of AI (and in the case of the US Act if it will ever pass), both acts show a desire by governments to shape how automated decision-making systems are used. It is genuinely positive that legislation will start to put some ground rules around AI use.</p><p>Neither of these Acts will directly affect generative AI since generative AI is typically not being used to make automated decisions impacting individuals.</p><p>Having said this, the legislation does touch upon issues of user data, so any of the companies (here’s looking at you Lensa) that actively bring in User data to process will certainly need to comply with data management provisions.</p><p>Most players in the space will probably ignore these acts in 2023. However, those with long-term ambitions should be thinking about how their results affect users, how data is managed, and whether or not legislation like this could create situations where their models could be deemed to be discriminatory.</p><h3>#8. Copyright and Plagiarism debates will continue to rage</h3><p>One of the hottest topics in generative AI has been the strong sentiment from Artists artists that many of the systems unfairly reuse existing Art to create new work in a way that violates copyright.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/0*EkPkRD-2pHMdXeqf" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/512/0*nbgetTjlXWeKd9J_" /></figure><p><em>(Midjourney: Prompts— “Basquiat in Barcelona” and “Barcelona in the style of Basquiat”)</em></p><p>Many of the image generation systems are trained using huge databases of online images, many of which have unclear copyright attribution. As such, one can argue that some “essence” of these images leaks out in the results the AI returns for any given prompt.</p><p>Even more upsetting can be prompts that ask for results “in the style of” another artist, such as the ones above. Or this one:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*GMqhtieuLn7AiNYl" /></figure><p><em>(Midjourney: Prompt — “The mona lisa in the style of Edward Hopper”)</em></p><p>In the later example, it’s unclear who should be more upset: Leonardo da Vinci or Edward Hopper.</p><p>The counterargument, however, goes that Human Artists have been thriving on these types of interpretations since the beginning of Art. Many famous paintings are inspired by other paintings.</p><p>It’s unlikely these arguments will get resolved quickly since it will be hard to draw a boundary of what exactly is a derivative work, how similar things can be, and what controls are in place to prevent out-and-out plagiarism.</p><p>In 2023 and for quite a while beyond, it seems unlikely there will be new legislation to cover these types of images. Instead, true legal disputes, should they arise, will likely be handled in the same way (to the same standard) as they would as if a human had produced the image. This is, ultimately, the only really true measure since AI is a tool in the hands of a human operator.</p><p>Much of the angst related to the copyright topic is likely tied up with the related feeling that praise (or monetary benefit) for such an image is “unearned”. This feeling of unfairness is. a natural result when one considers human artists might need months of work to produce something similar. The image above for the Hopperized Mona Lisa options took just 15 seconds for the system to render.</p><h3>#9 AI for code generation.</h3><p>Whilst image and text generation have been the most visible aspects of generative AI in 2022, the application of similar techniques to code generation has (surprisingly) started to get somewhere.</p><p><a href="https://github.com/features/copilot/">Github’s co-pilot</a> was the first to get wider acclaim. The system relies on <a href="https://openai.com/blog/openai-codex/">OpenAI’s Codex</a> and helps a programmer formulate lines of code in real-time from natural language prompts. The results are impressive, though not perfect (and there are also some <a href="https://www.infoq.com/news/2022/11/lawsuit-github-copilot/#:~:text=GitHub%20previewed%20Copilot%2C%20an%20OpenAI,code%20and%20comments%20on%20GitHub.">open source violation claims </a>in flight). Some of the other players include <a href="https://blog.salesforceairesearch.com/codet5/?utm_source=thenewstack&amp;utm_medium=website&amp;utm_campaign=platform">T5</a>, <a href="https://www.cogram.com/">Cogram</a>, and <a href="https://github.com/VHellendoorn/Code-LMs">Code-LMS</a>.</p><p>None of these systems yet replace a programmer entirely, but the increasing quality means that 2023 may, at the very least, see robust, usable products for some niche programming tasks. These could include building data integrations, specific types of websites or data mapping functions. The more specialized the domain with a large but tractable set of combinations, the more likely these tools are to succeed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*h_ScDPi81v1ofaO2" /></figure><p><em>(Midjourney: Prompt — A robot building a robot on an assembly line”)</em></p><h3>#10 AI in Context will bring the biggest wins</h3><p>Building general AI is very hard. Even in the context of a specific function like “image generation,” it is hard to serve a wide range of query types equally well. It’s likely that the most successful systems in terms of generating valuable outputs will be those which constrain the domain of operation tightly. Be that in code generation for specific application types or text generation in given domains.</p><p>In 2022 we were wowed by what’s possible, but we see the cracks. In 2023 we’ll likely see how high the bar can go when AI is constrained to be specific.</p><p>I expect to see at least a few “domain” or “style” specific AI generation services popping up and becoming favorites. This is something we’re also aiming to do at Timewarp for certain types of gaming generation (so we’ll be making part of this prediction true!).</p><h3>Conclusion</h3><p>Predictions are tough, but it’s an interesting game to play. Making them focuses the mind on what you think is going to be important. What are your AI and Creativity predictions for 2023?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*PYP_7gBE2msruwW3" /><figcaption><em>Midjourney: Prompt — “A Robot Sunbathing”</em></figcaption></figure><p><em>Originally published at </em><a href="https://area67.org/2023/01/01/ai-and-creativity-predictions-for-2023/"><em>https://area67.org</em></a><em> on January 1, 2023.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1c67425fcbb1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Art of Team Communication]]></title>
            <link>https://njyx.medium.com/the-art-of-team-communication-bc8cbe43bb7e?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/bc8cbe43bb7e</guid>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Sat, 17 Dec 2022 22:08:36 GMT</pubDate>
            <atom:updated>2022-12-18T07:56:48.976Z</atom:updated>
            <content:encoded><![CDATA[<p>I first met Don Rossmoore whilst working together in 2016. It was clear right away that he had an uncanny knack of understanding what people were really saying when they spoke. His understanding of why certain communications weren’t working was also far deeper than I’d heard before.</p><p>As humans, our most natural reaction to a challenging communication situation is to think, “If only I can express myself better so I can get the other person to understand.” However, this is almost never the real problem. Instead, it’s very likely that the real problem is that at least one of the speakers (and very likely you) is being unconsciously defensive about something. This leads to the following:</p><ol><li>Not listening to or understanding the worldview of the other person.</li><li>Being manipulative or evasive in what’s being said rather than openly discussing the core truth of what’s at stake.</li><li>Bringing points across in a way that implies you are not open to compromise or criticism.</li></ol><p>This defensiveness is extremely natural and arises from our deep instinct as humans to feel safe in a group, to feel right, to feel accepted. However, it can wreak havoc on group communications.</p><p>Don’s techniques for discussion and interaction were a huge help in unblocking impasses and enabling our team members to understand each better. Everyone was deeply motivated for the whole team to succeed, and getting communication right not only helped that happen but made those key discussions much warmer and more fulfilling.</p><p>Given the uniqueness and value of Don’s communication support techniques, quite a few people on the team encouraged him to turn them into book form so others could also try them out. It’s been a pleasure to work with him on that book and see that it’s now finally here. Welcome to <a href="https://www.amazon.com/Art-Team-Communication-communication-breaks-ebook/dp/B0BP1X8FJW/">The Art of Team Communication</a> (*), which was published on Amazon last week. I highly recommend it to anyone working in work teams, either as a manager or as a participant.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/313/0*6txn9DdrisCE-h06" /></figure><p>I worked with Don to help bring the book together, and it was a pleasure to do so. It was enlightening to talk through why our teams behave the way they do and what drives our behaviors.</p><p>We rarely listen as well as we think we do. We’re often preparing what to say next rather than taking in what we’re being told. This book has great practices on how to avoid these problems, but also a deeper dive into why they occur in the first place.</p><p>Not every technique may work for your team, but I’m betting there’ll be something that will make your communication more effective and joyous!</p><p>(*) $7.99 on kindle but discounted to $0.99 during some of the weeks post-launch in Dec’22.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/80/0*A7huL4oH7nnoKUSF" /></figure><p><strong>Published</strong></p><p><em>Originally published at </em><a href="https://area67.org/2022/12/17/the-art-of-team-communication/"><em>https://area67.org</em></a><em> on December 17, 2022.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bc8cbe43bb7e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Bounded Rationality: What generative AI is really doing for you]]></title>
            <link>https://njyx.medium.com/bounded-rationality-what-generative-ai-is-really-doing-for-you-c53335c4ea65?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/c53335c4ea65</guid>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Sun, 11 Dec 2022 09:50:36 GMT</pubDate>
            <atom:updated>2022-12-11T10:01:34.291Z</atom:updated>
            <content:encoded><![CDATA[<p>One of the most important principles in AI and, arguably, in daily life is the principle of Bounded Rationality. The term was coined by Herb Simon when discussing models of perfectly rational humans in economics in the late 50s. He defined it as follows:</p><blockquote><em>Broadly stated, the task is to replace the global rationality of economic man with the kind of rational behavior that is compatible with the access to information and the computational capacities that are actually possessed by organisms, including man, in the kinds of environments in which such organisms exist.</em></blockquote><blockquote><a href="https://plato.stanford.edu/entries/bounded-rationality/"><em>Herb Simon, 1957</em></a></blockquote><p>If we drill into the “computational capacities” part of the statement, though, there is a simple concept that often escapes people:</p><blockquote><em>You do not know all the consequences of the things you know.</em></blockquote><p>We generally assume we know what things can/cannot happen or that make sense in a given situation. On closer examination, this turns out to almost never be true.</p><p>Let’s imagine you know all the facts about something, and you also know most of the rules that would tell you how those facts can be combined. An example would be that “A is a liquid” and “B is a sponge” + that “Sponges normally absorb liquid”. Even with a relatively small set of facts and a small set of “rules” it very quickly becomes impossible to imagine all possible consequences and combinations of the facts.</p><p>The problem we face in any work, knowledge, or creative task is “How to explore the space of possibilities?”. As humans, we absolutely forced to use smart guesses, direct our attention and leave out many combinations. We simply don’t think of them because we don’t have time. Our biases also mean that there are many combinations we will not even imagine.</p><p>This is what the current wave of generative AI is doing:</p><ol><li>Pulling in new facts (sources) relevant to a particular question. Like relevant images, you didn’t know or facts/text that you had never read.</li><li>Combining things in ways that your brain had not come up with yet. This is really applying rules to combine things that you had forgotten about or didn’t know.</li></ol><p>In other words, generative AI is allowing us to jump around the space of possibilities in a different way to just sitting in a chair and thinking things through. This is an amazing superpower. Some of the results can be inspiring and enable us to create things we would not have thought of.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*r9JuQqPpEYzYYE0o" /></figure><p>Of course, each generation of technology will hit its own limits. AI is also bounded in its rationality. We’ve only just figured out the most basic ways for generative AI to stitch together facts. In many cases, results can still be highly non-sensical.</p><p>So think of generative AI as an accelerated and tunable exploration device rather than a second brain.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*jTC_FNlcQkpeDCei" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/80/0*P78ZZn-Xgka31_lC" /></figure><p>Senior Director at RedHat Inc. Ex. CEO of 3scale Inc. Concerned citizen of the world. <a href="https://area67.org/author/njyx/">View all posts by njyx</a></p><p><strong>Published</strong></p><p><em>Originally published at </em><a href="https://area67.org/2022/12/11/bounded-rationality-what-generative-ai-is-really-doing-for-you/"><em>https://area67.org</em></a><em> on December 11, 2022.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c53335c4ea65" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The wrong question to ask about AI]]></title>
            <link>https://njyx.medium.com/the-wrong-question-to-ask-about-ai-d833f8e1be88?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/d833f8e1be88</guid>
            <category><![CDATA[work]]></category>
            <category><![CDATA[society]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Fri, 09 Dec 2022 22:47:45 GMT</pubDate>
            <atom:updated>2022-12-09T22:50:28.363Z</atom:updated>
            <content:encoded><![CDATA[<p>The explosion of generative AI (this week’s <a href="https://openai.com/blog/chatgpt/">ChatGPT</a> was the latest phenomenon) is bringing an explosion of amused joy at what these tools can do and … angst about whether they might replace people in some new area. <a href="https://twitter.com/MichaelTrazzi">Michaël Trazzi</a> ‘s widely copied fun meme is just one example of the later:</p><h3>Michaël Trazzi on Twitter: &quot;alignment pilled pic.twitter.com/Dk0iRmOau3 / Twitter&quot;</h3><p>alignment pilled pic.twitter.com/Dk0iRmOau3</p><p>There will certainly be a time when AI is able to do many of the things that human’s currently do for work. But when that will happen (or even if) is fundamentally the <em>wrong question to ask</em>. What is really much more important to ask is</p><blockquote><em>What can I do WITH AI?</em></blockquote><p>Human + AI will do so much before acts on its own that the former question is hardly relevant to today’s discourse. What’s more, what we do with AI and how we treat it will probably determine how AI treats us when it finally can do everything we can.</p><p>You might want to think hard about those Midjourney and ChatGPT prompts 🙂</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/80/0*lrcmEbf2E7NmECcU" /></figure><p>Senior Director at RedHat Inc. Ex. CEO of 3scale Inc. Concerned citizen of the world. <a href="https://area67.org/author/njyx/">View all posts by njyx</a></p><p><strong>Published</strong></p><p><em>Originally published at </em><a href="https://area67.org/2022/12/09/the-wrong-question-to-ask-about-ai/"><em>https://area67.org</em></a><em> on December 9, 2022.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d833f8e1be88" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Why are you Making Games? (With AI?)]]></title>
            <link>https://njyx.medium.com/why-are-you-making-games-with-ai-db4f955b4ddf?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/db4f955b4ddf</guid>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[games]]></category>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Wed, 26 Oct 2022 12:58:59 GMT</pubDate>
            <atom:updated>2022-10-26T15:09:27.535Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lttpftmV1N4zl02vOymY3Q.jpeg" /></figure><p>In the past few months we’ve been able to talk to more people about what we’re doing at <a href="http://timewarp.com/">Timewarp</a> (Yay!). (Spoiler — we’re working on the on our first game Mafia Pigs and <a href="https://store.steampowered.com/app/2106110/Mafia_Pigs/">looking for Alpha testers right here!</a>)</p><p>As we’ve explained what we’re doing, the same (very natural!) question has come up many times.</p><blockquote><em>Why did you leave a field you worked hard in for many years (Web APIs) and jump into something totally new — Games?</em></blockquote><p>Even more confounding, why do it with AI? isn’t AI going to put us all out of a job?</p><p>These are great questions! The “making games question” is also very understandable since almost everyone who writes code probably dreamed of making games at some time or another.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*SvAPjfV2-6PEWVB0" /></figure><p>Like many others, the first lines of code I ever wrote were for games. First this was copied line by line from PC magazines in late 80s, then writing a never to be finished text adventure in GW Basic. I later developed a computer Go playing engine called Gobi that got pretty good at solving certain parts of Go play but that’s as far as I got in terms of building an actual game.</p><p>Despite this being an obvious motivation, the pure joy of game creation isn’t my reason for making games today. The reason is rather different and rooted in the belief that today’s games are the seeds of our online interactive future and that we need better ways to facilitate their creation. Even more specifically myself and the rest of the team at Timewarp has two strong beliefs:</p><ol><li>That as online worlds become more interactive and “game like” it will be critical that anyone can be creator in such worlds — even if they do not have deep technical skills,</li><li>That AI (Artificial Intelligence) can play a major role both in making these online worlds more vibrant and in helping creators bring them to life.</li></ol><p>So while Timewarp does develop games, what’s really going underneath is that we’re building a platform to enable others to create and extend those games. A platform that makes it much easier to manipulate the behavior and logic behind a game than anything available today. AI helps us to do this by acting as a support for human creators and as technology that allows the worlds being created be rich and dynamic.</p><p>The hope is that it becomes as easy to create an online world that participants can explore, interact with and play or work in as it is to create a professional YouTube video or a Website today.</p><p>Great examples of games that allow user creation today include Minecraft, Roblox, Gary’s Mod, The Sims and the ever continuing Warcraft III modding community. Each of these modifiable games has unleashed a wave of creativity. There is still much more to do in terms of making game creation (and creation in games) available in many game worlds!</p><p>As well as this main driver for “why” games. I’ll go a little deeper into some of the reasons.</p><h3>Games drive technology</h3><p>Technology gets better if it is used to try to solve problems at the limits of what it can do. The programmer who can’t wait a day for an algorithm to run tries to optimize the algorithm, the hardware designer pushes the performance further and further as applications need more power. Games have always been one of the biggest drivers in technology. This is true both in terms of raw power required to run the leading games and in terms of imagining the possibilities of the format.</p><p>The interactive worlds of games such as <a href="https://www.cyberpunk.net/us/en/">Cyberpunk 2077</a> (yes, it had problems on release but the created world is incredible) and the amazingly atmospheric world of <a href="https://www.youtube.com/watch?v=fxHTk_aQo7o">Ghost of Tsushima</a> are just two extremely impressive examples. These worlds pushed the boundary of the possible in terms of technology (in Cyberpunk’s case possibly too hard on the PS4!) but also showed just how immersive things could be.</p><p>The same is true in AI. Computer Go seemed a nearly intractable problem in the mid 1990s when I worked on it. The solutions were way below human level play and required complex, careful optimization. When Google’s<a href="https://en.wikipedia.org/wiki/AlphaGo_versus_Lee_Sedol"> Alpha Go AI beat one of the worlds top human players Lee Sedol in 2016</a> it was a milestone achievement. A truly incredible feat of engineering.</p><p>What made Alpha Go even more remarkable was the way in which it learned how to play at this level. I mastered the game not through painstakingly re-created human knowledge but playing endless internal tournaments to train the algorithm. Since this landmark Alpha go has gone on to play and win at many other games (including <a href="https://www.youtube.com/watch?v=cUTMhmVh1qs">Starcraft II</a>) as well as finding applications in other areas such as <a href="https://www.quantamagazine.org/in-new-math-proofs-artificial-intelligence-plays-to-win-20220307/">mathematical proof discovery</a>.</p><h3>Interactive online worlds are key to our digital future</h3><p>It seems certain that online interactive spaces will be a huge part of the future.</p><p>Video, images, documents and streaming are all well-developed today, but the creation of worlds you can step into and interact with opens up huge possibilities (even if those worlds are 2D, 2.5D or text rather than VR) — it is the interactivity and responsiveness to actions that it is important.</p><p>Games have already shown the way by creating some of the most amazing experiences to date, but it seems obvious that the same technology can produce valuable experiences for business such as models of buildings that can be manipulated by architects or biological simulations which can be controlled by scientists.</p><p>Some games like AmongUs and Animal Crossing have brought millions of people together in what is essentially a social setting supported by a game mechanic.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*hIJywDvgCTg_IE9O" /></figure><p>In this context we think it’s critically important that almost anyone can be a creator of such online worlds and within such worlds. It should not just be in the power of software engineers to craft such experiences. The more people that can create, the more diversity and richness will come into being and the more the social and economic benefits of creation will be shared around.</p><p>The more technology can support broad swathes of people to create the impactful it will be.</p><h3>But what about online subverting the real?</h3><p>Do we really want all that screen time?</p><p>Many of the science fiction stories about future online worlds are deeply dystopian (here’s looking at you Neuromancer, Snowcrash, Ready Player One…). More on how I think we avoid that in a future post! However, one very valid question is: If we’re creating these more interactive worlds will we not be increasing screen time to unhealthy levels? Worse is it part of disconnecting people from the real world?</p><p>Surely we should be going hiking on the weekend? Not spending it in a virtual world?</p><p>I’m 100% a fan of getting out into nature, spending time with people in the real world and believe hugely in the benefits of this.</p><p>However, for me, much like the rise of the Internet, the Web and other communications technologies, the online world is an incredible engine for learning, productivity and social experiences. When used in the right way, online interactive spaces will make possible work and play that simply isn’t available today. They will also create opportunities for those who aren’t fortunate enough to live somewhere where work (or even nature) is abundant. It would be hard to imagine a world without email today or a world in which you could not just reach out to a leading individual in your field of choice via LinkedIn or Twitter. These technologies have made new things possible. Including the potential for projects that develop solutions to pressing world problems.</p><p>So we do need to take care that these new technologies are applied in ways which allow people to keep a balance between online and the real world.</p><p>I’d also argue that in today’s age much of the screen time is in the form of passive viewing rather than active creating. While we will always need both, a shift to more active participation, choice making and creation during online experiences will hopefully be healthy in empowering people to get more out of their time spent in front of a screen.</p><h3>The team and our first game</h3><p>The last BIG reason for working on games and creator teams is the awesome people I get to work with. The mix of engineering, game talent and artistry is a fantastic. Having fun at work is definitely a big plus!</p><p>We’re now close to chipping the Alpha version of our first game Mafia Pigs. We have a long way to go but it’s super exciting to be able to share something.</p><p>We’d love feedback on the Alpha test so please consider signing up on <a href="https://store.steampowered.com/app/2106110/Mafia_Pigs/">Steam</a> (wishlist and signup for the test!) so we can send you the build when it is available! Even if you don’t consider yourself a gamer we’d love feedback. Hopefully the experience we’re creating will be fun and engaging no matter how game-savy you feel yourself to be!</p><p>I’ll be posting more about the industry and our progress over the coming months. Glad to be back! 🙂</p><h3>Images in this post</h3><ul><li>Lego photo by <a href="https://unsplash.com/@helloimnik?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Hello I’m Nik</a> on <a href="https://unsplash.com/s/photos/lego-play?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a>.</li><li>Animal crossing image by <a href="https://unsplash.com/@stereophototyp?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Sara Kurfeß</a> on <a href="https://unsplash.com/s/photos/animal-crossing?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a></li></ul><p><em>Originally published at </em><a href="https://area67.org/2022/10/26/why-are-you-making-games-with-ai/"><em>https://area67.org</em></a><em> on October 26, 2022.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=db4f955b4ddf" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Favorite Books of 2021]]></title>
            <link>https://njyx.medium.com/favorite-books-of-2021-65f3f95e1fc7?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/65f3f95e1fc7</guid>
            <category><![CDATA[books]]></category>
            <category><![CDATA[books-recommendation]]></category>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Tue, 04 Jan 2022 03:00:12 GMT</pubDate>
            <atom:updated>2022-01-04T03:07:37.126Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/637/1*wDPJQfhFCPsefjRZTJuIgQ.png" /></figure><p>The Pandemic has certainly helped on the reading front! In 2020 I felt like I read so much my brain could hardly hold an original thought, so in 2021 I toned it down and set an upper limit on books + most of those were fiction.</p><p>The books read in 2021 books contained some gems though, here are five of the best. A warning upfront that I tend to read a lot of sci-fi, economics, and business books. Here they are in any case:</p><ul><li><a href="https://www.goodreads.com/book/show/23209924-the-water-knife"><strong>The Water Knife</strong></a> (<em>Paolo Bacigalupi</em>): this book is near-term Science Fiction and focuses on the near-collapsed society of the Western US after water dries up. It’s hard to read at times, violent and gut-wrenching, but the characters are vivid and the story is (sadly) not that far from a reality that could really come to pass. Well worth the read for its take on how humans make the best and worst of a crisis at home. It’ll have you digging wells in your back garden (or moving to Switzerland).</li><li><a href="https://www.goodreads.com/book/show/23533039-ancillary-mercy"><strong>Ancillary Mercy</strong></a><strong> </strong>(<em>Ann Leckie</em>): this is the third and final book in Ann Leckie’s Imperial Radch trilogy and I read both books before this in 2020 (which I’d highly recommend doing — start with <a href="https://www.goodreads.com/book/show/17333324-ancillary-justice">Ancillary Justice</a>). The world Leckie creates is really immersive and different. The story is told mostly through the eyes of a body-possessed human (the “Ancilliary” of the title) who starts life as literally as a puppet of a warship. The stories are low on combat (though there is some) but really high on character EQ, creative consequences of choices made in society, and tension. The series is pretty different to most high-concept Sci-Fi, definitely worth a look!</li><li><a href="https://www.goodreads.com/book/show/8727466-why-we-get-fat"><strong>Why We Get Fat and What to Do About It</strong></a> (<em>Gary Taubes</em>): there are a bazillion pseudo-scientific books on nutrients, metabolism, weight loss, and so on. In a few ways this book suffers from some of the problems they all do: some hyperbolic repetition, a fair bit of science, but still the feeling of some picking and choosing etc. I somewhat hesitate to recommend this but … it is quite refreshing and (to me) at least does contain a central core argument that almost all such books ignore: the fact that while eating too much (of the wrong stuff) does make us gain weight, the real driver to understand is why we end up having the appetite to do so in the first place. There are some pretty compelling passages that talk about how some of the vicious cycle between food and the need to eat occurs. The second half of the book is a really interesting deep dive into how the metabolism really works. Your mileage may vary on this one!</li><li><a href="https://www.goodreads.com/book/show/29496453-the-book-of-joy"><strong>The Book of Joy: Lasting Happiness in a Changing World</strong></a> (Audiobook edition) (<em>The Dalai Lama, Archbishop Desmond Tutu, and Douglas Carlton Abrams</em>). I listened to this as an Audiobook and I think, in this case, it’s a must: the voices of the presenters add a tremendous amount to the material. I’m not sure a paper version would be anywhere near as powerful. This is another type of book I hesitate to recommend. It’s a big topic, and it’s hard to know what joy means to people, but there are some really inspiring passages and insights in the book. It is not religious, despite the two main protagonists being members of two of the world’s faiths (and the facilitator being Jewish). It is a long series of conversations between the Dalai Lama and Desmond Tutu (who recently passed away) talking about their experiences and philosophies. Much will not be new to you if you have read some modern philosophy, greek philosophy, and Buddhist philosophy. However, there are many moments of clear explanation, examples, and insight that really do stand out. It’s also striking how similar positions from different faiths can be.</li><li><a href="https://www.goodreads.com/book/show/13530973-antifragile"><strong>Antifragile: Things that gain from Disorder</strong></a> (Nassim Nicolas Taleb): This was a re-read, something I do with this book every couple of years. Taleb is pretty combative and deliberately offensive to his critics at times (I guess he’d say it represents his “skin in the game”) and the book is funnier for it. The real core of it though is the idea that systems that benefit from randomness exist, can be harnessed, and indeed are pretty fundamental to our existence. Furthermore, modern societies tend to spend their time trying to reduce randomness and as a consequence, inadvertently making things far more fragile than they need be. The book covers a wide range of important consequences of this point of view. It is really worth getting to grips with this concept (and indeed most of the rest of Taleb’s work).</li></ul><p>There were so many great books in 2021 I have to do at least a few honorable mentions: I’m re-reading the <a href="https://www.goodreads.com/series/49118-culture"><strong>Culture novels</strong></a> by Iain M. Banks, <a href="https://www.goodreads.com/book/show/50396269-a-crown-of-talons"><strong>A Crown of Talons</strong></a> by Katherine and Elizabeth Corr, <a href="https://www.goodreads.com/book/show/55849312-qalea-drop"><strong>Qalea Drop</strong></a> by Joel Shepherd, plus a lot more.</p><p>The good-reads list of all the books I tracked in 2021 is online <a href="https://www.goodreads.com/user/year_in_books/2021/75107226">here</a>.</p><p>Happy Reading!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/80/0*7KpFYjvYDFFnhyIC" /></figure><p>Senior Director at RedHat Inc. Ex. CEO of 3scale Inc. Concerned citizen of the world. <a href="https://area67.org/author/njyx/">View all posts by njyx</a></p><p><strong>Published</strong></p><p><em>Originally published at </em><a href="https://area67.org/2022/01/03/favorite-books-of-2021/"><em>https://area67.org</em></a><em> on January 4, 2022.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=65f3f95e1fc7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[For 2022: Reduce the Bad and Schedule the Good]]></title>
            <link>https://njyx.medium.com/for-2022-reduce-the-bad-and-schedule-the-good-c0a58a0082e6?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/c0a58a0082e6</guid>
            <category><![CDATA[2022]]></category>
            <category><![CDATA[goals]]></category>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Sat, 01 Jan 2022 03:41:00 GMT</pubDate>
            <atom:updated>2022-01-01T03:44:30.664Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qH18B5Go3U17V84MqLV6PA.jpeg" /></figure><p>Well hell… it’s been a long time since the last post here! Just one post in 2021 after a good stretch of regular posts before that. It’s been a super busy (and in many ways great!) year, so it’s not that I fell off a cliff. Life happened though, with lots of great work at Timewarp, family events, and other things taking time away from blogging.</p><p>2022 felt like a period of deep learning on a lot of topics and with a lot of thinking to do!</p><p>As the year ends, I feel a lot more reflective and I’m missing the process of blogging. So hopefully there’ll be more again in 2022.</p><p>Given it is the end of the year, there is a process that others might find useful. I stopped making New Year’s resolutions a considerable time ago, but I did very much double down on thinking about the year just gone and the one to come. It’s the mix of gratitude for what has been, and hope/excitement for what is to come that I really love. I typically set a few goals, but less as a thing that “must be achieved” and more as a direction to head in.</p><p>The reflection I do isn’t too dissimilar to what Tim Ferris recommends in this podcast here:</p><p>Time Ferris’ Past Year Review</p><p>You might not love everything Tim Ferris has done, but his blog often includes some really smart insights. This one, in particular, is worth a listen: 6 minutes.</p><p>The core of the method is simple:</p><ul><li>Go back through the year month by month and try to identity the times/events that were really happy (hopefully quite a few) and those that were truly bad (hopefully less of those). What drove that happiness? and where did those bad times come from?</li><li>Now look at 2022 and start to schedule things throughout the year that match those happy events: how can you do more? When should you allocate time for them? How can you make more of them this year?</li><li>On the other side of the equation: make conscious decisions to avoid those events that made you unhappy last year, try to avoid doing too much more of that.</li></ul><p>This sounds pretty simple, but of course, it is harder than it looks. Trying to schedule things far in advance can feel daunting and many of those negative experiences weren’t exactly things you asked for: it would have been hard to predict when they would happen. Still, it’s really worth a shot.</p><h3>Scheduling the Good</h3><p>It is truly important to schedule those happy times first. Having these items in the calendar sooner rather than later makes sure they happen. It makes sure the friends and family you want to spend time with are more likely to be there too. Or, if it’s just you, you at least get the best deal on travel and you know when to request PTO for.</p><p>Most importantly though, these events act as anchors for your year: times you’ll look forward to and be able to recharge.</p><p>It might feel selfish to pin down the “me-time” first but all in all, it’ll make you more effective and likable throughout the year. Everyone will thank you for it!</p><h3>Eliminating the Bad</h3><p>Eliminating the bad is harder. How can you really spot when they will happen? This isn’t obvious to begin with, but if you really dig there are often patterns. What was truly terrible in 2021?:</p><ul><li>Health events?</li><li>Financial loses?</li><li>Admin and paperwork nightmares?</li><li>Family fallings out?</li><li>…</li></ul><p>None of these are ever really predictable, but there can be signs or patterns that show up beforehand. Or, for other items, things would have been much simpler if they were handled earlier rather than later.</p><p>In some cases, eliminating the bad might be simply avoiding going to an event you know you’ll dislike or deciding to break ties with a person who routinely causes you grief. In others, it may be about doing something to pre-empt negative events:</p><ul><li>What health check-ups have you been skipping? Maybe it’s time to get them in the calendar early. (I’ve been super guilty of this over the years: it’s hard to schedule a check-up you don’t really want to hear the news for, but it’s better to get ahead of a problem.)</li><li>Is your financial situation out of hand? Maybe it’s time to sit down and really quantify the problem? Make a long-term gradual plan to get it under control and build an emergency fund in case problems arise?</li><li>Are there important renewals coming up that could catch you out? Passports? Driver’s licenses? When do taxes have to be filled? Can you already plan the dates when you will do the tax preparation, why not make it well ahead of the deadline?</li><li>Are there family challenges under the surface that everyone is avoiding? What needs to be said? What is the best way to say it and when?</li></ul><p>None of these steps are easy, but taking your list of tough times from 2021 and maybe the previous few years can really help focus the mind. What don’t I want to happen again?</p><p>Being a little proactive and allocating some time to solving those problems early can really help.</p><h3>Happy 2022!</h3><p>I’m definitely not qualified to give life coaching advice, but going through this process each year over the past few years has helped a lot.</p><p>I used to dislike New Year because it seemed to hold so much misplaced optimism: why will anything be different in the next year? It seemed delusional.</p><p>But, by taking the time to reflect, we can indeed make the next year different from the last. We have the power to change our planning and how we react to events. That won’t change the luck we receive during the year, but it might mean we have a better shot at making more good things come true.</p><p>Good luck in 2022!</p><p>Image by: Photo by <a href="https://unsplash.com/@kalvisuals?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">KAL VISUALS</a> on <a href="https://unsplash.com/?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/80/0*R--RNC7BnJHt_8QH" /></figure><p>Senior Director at RedHat Inc. Ex. CEO of 3scale Inc. Concerned citizen of the world. <a href="https://area67.org/author/njyx/">View all posts by njyx</a></p><p><strong>Published</strong></p><p><em>Originally published at </em><a href="https://area67.org/2021/12/31/for-2022-reduce-the-bad-and-schedule-the-good/"><em>https://area67.org</em></a><em> on January 1, 2022.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c0a58a0082e6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Fragility of Institutions]]></title>
            <link>https://njyx.medium.com/the-fragility-of-institutions-4f2ff3fd6ec8?source=rss-2408aa3ac35a------2</link>
            <guid isPermaLink="false">https://medium.com/p/4f2ff3fd6ec8</guid>
            <category><![CDATA[usa]]></category>
            <category><![CDATA[politics]]></category>
            <category><![CDATA[society]]></category>
            <dc:creator><![CDATA[Steven Willmott]]></dc:creator>
            <pubDate>Mon, 18 Jan 2021 05:54:06 GMT</pubDate>
            <atom:updated>2021-01-18T06:13:35.354Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SHbjPOLbjKMqti9MLToOgQ.jpeg" /><figcaption>Palacio de Cristal, Madrid</figcaption></figure><p>The last couple of weeks have been scary in the United States. Continued rejection of November’s election result, a high stakes election in Georgia with plenty of divisive rhetoric, votes in congress to deny the election result, and then the invasion of the US Capitol building in Washington by a large group of President Trump’s supporters.</p><p>This week’s presidential inauguration looks set to take place under heavy guard and the impeachment vote in the House of Representatives that just passed means there will be an ongoing fight about the threat of violence and how the events of the 6th of January arose.</p><p>What all of this brings home to me is how fragile human institutions can be, yet how willing people are to attack them for personal gain. It looks like the breaching of the Capitol will not overthrow democratic processes in the United States but they highlight just how many people are willing to risk rocking a foundation without consideration for what they are doing.</p><p>Words and actions matter. Adherence to the constitution, functioning elections, certificating those elections, and upholding the law rely on nothing but words. They rely on the individuals in the process honoring those institutions and choosing to abide by them. Once this ceases to happen the institutions crumble and die. That in turn means there is less and less common ground between parties: no framework for debate.</p><p>This applies equally to Republicans and Democrats. President Trump’s actions have certainly most visibly been encouraging disregard of institutions and individuals he has no love for. However, on the Democrat side immediate calls to apply a measure such as the 25th Amendment when its applicability is highly dubious, also look like an attempt to use something inappropriate to a partisan end.</p><h3>Isn’t it just Protest?</h3><p>One of the arguments made in support of the January 6th Capitol invasion has been that it is a legitimate protest. Protests about the fact that voices were not being heard in the courts regarding the claim that the November 6th election outcome was incorrect.</p><p>Peaceful protest certainly looks different to what happened on January 6th. The potential threat to lawmakers was significant and the protestors were effectively attacking several institutions at once:</p><ul><li>The joint sessions of Congress and Senate then in motion.</li><li>The processes in the courts which had already heard many of these claims (including the supreme court).</li><li>The decisions of the states to certify their own votes.</li></ul><p>All of these were processes that the protestors were violently dismissing through their actions.</p><p>The election result hadn’t been what they’d hoped for (or what they had been told to expect). If there are genuine feelings that the result is wrong, then peaceful protest seems legitimate. The voices should be heard. This is different from breaking into the place of government and stopping the appointed process by intimidation. At the end of the day, there are processes in place for a reason. It is hard to argue that the Trump campaign has not had sufficient opportunity to try to prove the validity of their claim but has failed to so.</p><p>By disrupting the process with threats and violence, the protests seek to derail the institution in favor of their minority view.</p><h3>The Deeper Problem</h3><p>Whether one thinks what the protesters did was just protesting or something worse, it is particularly worrying when seen as the culmination of a lot of attacks and slights on the institutions of the United States. These didn’t just come from the protestors but from others supporting the “Stop the Steal” message of President Trump.</p><p>These statements attack the legitimacy of states, courts, or elected representatives and of the Supreme court. They are cheap politics that can backfire in a major way. The certification of the vote was being carried out by elected officials who presumably could hear peaceful protests and make decisions on the basis of what they thought was right.</p><p>The Supreme Court is the ultimate arbiter in the United States. The court can only function however if it is seen as legitimate by the population and by elected officials. Attacking its rulings and threatening to go against them (even in veiled terms) undermines its authority and incites others to cease believing in authority.</p><p>The composition of the Supreme Court has also seen a great deal of political maneuvering over the past 6–7 years. Now to the point that the last 3 appointments have had a strong conservative lean. For Democrats, it is also hard to see past the comparison of Republican delays of their Supreme Court nominee at the end of the Obama administration and the rushing through of Amy Coney Barret’s nomination in October</p><p>This type of overwhelming influence on an institution can bring significant political benefits, but take too far it can break the legitimacy of an institution.</p><p>With actions like these, the risk is that trust in processes that are designed to help government function and make decisions gets eroded on both sides of the political spectrum. Taken to its extreme it means that there may end up being no common ground anymore.</p><h3>The Unfairness of “Fair” Systems</h3><p>As a researcher, newly arrived in Spain, I got an interesting introduction to the Spanish process for appointing tenured professors in the University system. I saw this through the annual selection processes my then boss went through each year.</p><p>On the surface, the system is designed to be 100% fair. However, in practice, it leads to extremely unfair outcomes.</p><p>While it may have changed over the years, at the time, the process worked roughly as follows:</p><ul><li>Each year, all the public universities in Spain indicate whether they would like to open a tenured professorship that year + how many. Some do, some do not. This produces a count of potentially open positions.</li><li>The Ministry of education then opens a call for Professors with around 1.3x — 1.5x the number of openings that the count would indicate. These are <em>accreditations of fitness to take a tenure professor role</em> (not actual appointments). Winning one of these accreditations is a requirement to being considered for an open position.</li><li>The call then leads to a process (<em>Oposiciones</em>) based on a series of complex criteria that change slightly each year to select the candidates who will be given the accreditation that year. The candidates have a lot to prepare.</li><li>The selection of the successful candidates is carried out by a jury of 7+ Professors in each area of competency that is randomly selected from those eligible in the whole country. Hence each year sees a different Jury with a different composition.</li></ul><p>At first glance this seems like a logical and fair system: a random qualified jury to ensure no ongoing biases and a number of slots based on need.</p><p>In practice it leads to at least two significant pathologies:</p><ol><li>Many universities do not declare they will open a position because they are unsure preferred candidates will be accredited. Since they do not want to be stuck appointing someone they do not want, instead, they wait to see if candidates they are already in contact with are accredited that year or not. If those candidates do win accreditation only then do they open a position and appoint the candidate. The more this happens the greater the risk that those Universities that opened a position are not able to fill it since many of the accredited professors end up taking previously undeclared positions.</li><li>The randomness of the jury means that any single particular combination of a jury is extremely unlikely to occur again in the future. So if a set of colleagues from one University, from one school of thought, or some other grouping are selected for the jury in a given year, this combination is highly unlikely ever to occur again. As a result, they are highly incentivized to support candidates that they have a positive view of rather than being entirely objective. They may not always do this, but the reason for the incentive is clear: given the lucky combination of that year there is pressure to “use this opportunity”. In any other given year another alignment of interests will likely also be under pressure to take advantage of their good fortune., so if they do not act now they are simply losing out.</li></ol><p>Of course in many cases, good decisions are still made, but strikingly poor decisions are also made. One might imagine that in the long run “things even out” since everybody gets their turn. However, this is to misunderstand randomness. As humans, we often think of randomness quickly cycling through all the likely results but in reality, randomness looks a lot less random than we think it ought to.</p><p>Here for example are the first three strings of 1s and 0s I just generated on a random number generator (there are many such as <a href="https://www.random.org/sequences/?mode=advanced">this</a> and <a href="https://commentpicker.com/random-number-generator.php">this</a>):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/524/0*8t99z4S5Sw6rWhnG" /></figure><p>Showing these to a human likely would generate the response that these are not random, yet they are! The point here is that it can take a long time for conditions to “even out” and they in fact technically never do (flipping 4 heads on a coin toss does not increase the probability that the toss throw will result in a tail). A single academic career also has a time limit in human years from when a person reaches the level where they might apply for a tenured position to when they retire.</p><p>What was the point of this divergence into Spanish University procedure?</p><p>The point here is that systems that seem fair: a group is selected, then another and another… to run things can seem fair but end up with pathological results.</p><p>This also applies to politics.</p><p>The system is designed to be “fair” in how people are elected, but elected officials are also keenly aware that when elected, the particular coalition of parties in power, the political climate, and many other factors are unique to that time. This means there is a strong incentive to aim at short term beneficial outcomes for themselves and their supporters without care for what comes later.</p><h3>This gets particularly bad when…</h3><p>… one of the things that is in the power of the elected group is influence over the future outcome of elections or decisions beyond their term.</p><p>In the Spanish professorial system, the damage from unfairness is localized to each year (if it happens at all). In systems where the incumbent party can affect the process itself though, damage can be done that spans the ages. In the United States this includes:</p><ul><li>Court appointments (lower courts and the supreme courts)</li><li>Redistricting of voting districts</li><li>Voter registration requirement changes</li><li>Statehood discussions such as those for Washington DC and Puerto Rico</li><li>Electoral College rule changes</li></ul><p>These are all powerful tools to affect the institutions of government directly. There are also many indirect effects (many of which have been practiced in the past four years):</p><ul><li>Attacks on the validity of voting processes</li><li>Attacks on judges</li><li>Violation of unwritten rules of ethics</li></ul><p>Each of these weakens authority and trust. They are actions that lessen the status of institutions. They create an increasing risk that some individuals or groups will simply begin to disregard certain institutions wholesale.</p><p>Within a given margin this usually provides some tactical advantage to one group or another, but taken too far it may simply break the institution completely. In extreme cases this could lead to increasing violence, boycotts of votes, to even more extreme redistricting, and so on.</p><p>Hopefully, the new United States administration will take the view that respecting institutions and healing their integrity is more important than trying to generate some short term gain. Hopefully, the Republican opposition will recognize that the common ground they represent is something they also need.</p><h3>Changing Institutions?</h3><p>If changing institutions is dangerous are they always stuck in the past and can never change? No, but change has to happen in the right way.</p><p>At their core institutions are <em>an agreed mechanism for coordination of individuals/groups for a common purpose</em>.</p><p>In the case of US government institutions, the purpose is the government of the United States for the benefit of its citizens.</p><p>The mechanism is a thin sliver of agreed process that all parties need to respect, so changing it is hard. It can be done, but really for change to work, it would need to be done with the agreement of all parties participating in the mechanism. Without this, the back and forth between parties in power will simply create an ever greater wrenching back and forth. Changing an institution with a partisan only vote will likely mean that the other party will simply reject the change, fight against it, and seek to overturn it next time around.</p><ul><li>One would hope for example that BOTH Democrats and Republicans are invested to shore up trust in the voting system.</li><li>That they are also both invested in having a strong judicial process to resolve election vote challenges that is beyond reproach</li></ul><p>It seems less likely that Republicans will want to address the fact that the electoral college consistently deviates from the popular vote or that redistricting has been extremely strongly in their favor in some states. In these areas, Democrats will need to find a good way to seek common cause to get the kind of changes they probably want.</p><p>Changing the rules of the game needs to involve all the players. Hard to do but necessary.</p><p>In any case, institutions are just words and shared agreement. US government institutions have survived some dangerous attacks in the past weeks and months. Hopefully, both parties will prioritize strengthening them rather than putting them at further risk!</p><p>I think it’s important to add. Yes, I’m an armchair pundit on this topic and I can’t even vote in the US. On the other hand, I do live here, and if an insight happens to resonate perhaps that helps in some small way!</p><p>Photo by <a href="https://unsplash.com/@alevtakil?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Alev Takil</a> on <a href="https://unsplash.com/?utm_source=unsplash&amp;utm_medium=referral&amp;utm_content=creditCopyText">Unsplash</a>. (Palacio de Cristal, Madrid Spain)</p><p><em>Originally published at </em><a href="https://area67.org/2021/01/17/the-fragility-of-institutions/"><em>https://area67.org</em></a><em> on January 18, 2021.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=4f2ff3fd6ec8" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>