AI Ran Out of Internet. Now It’s Learning by Playing Games Again.
Alex Duffy / Playtesting
Length3m
About this audiobook
Welcome to the first edition ofPlaytesting, our new column in whichAlex Duffyexplores how games can help make AI smarter and more beneficial for people. In addition to being a contributing writer, Alex is the cofounder ofGood Start Labs, a company dedicated to using games to improve AIthat was incubatedat Every.—Kate LeeWas this newsletter forwarded to you?Sign upto get it in your inbox.Earlier this year a version of Gemini won agold medalat the International Mathematical Olympiad (IMO). Some models can diagnose certain medical conditions as well as humanphysicians. And AI is helping us forecast weather faster and more accuratelywithout the need for supercomputers.But in other areas, performance is just plain bad. AIs used in legal research, for example,have fabricatedhundreds of facts and even whole cases that don’t exist, leaving attorneys who’ve trusted models facing fines and other liabilities.It doesn’t have to be this way. It’s all down to the data. This era of generative AI was trained on publicly available information scraped from the internet—a biased dataset rich in some domains of knowledge, and wanting in others. And now that they’ve hoovered that up, it will take years to generate more high-quality knowledge to ingest and create more reliable outputs for every user. AI models are, as a result, incredibly good at tasks where they have lots of high-quality examples, and weak at those they don’t. This phenomenon is often described using the metaphor of ajagged frontier.AI could answer routine medical questions, triage symptoms, or explain test results while doctors focus on complex problems and novel treatments. But we can’t deploy it in critical areas such as healthcare, legal research, or financial advising if it’s not reliable. If we want to realize the promise of AI, this jagged frontier needs to be filled in. Games can help make that happen.In a game, we can create any scenario—a negotiation, a crisis, a moral dilemma, a portfolio to manage—and watch exactly how the AI responds.I’ve learned so much from games.Runescapetaught me how to type, how markets work, and how to not get scammed. Redstone inMinecrafttaught me about circuits long before my electrical and computer engineering degree.League of Legendstaught me collaboration under pressure and awareness, and almost everyone I’ve asked has similar stories about games.We can iterate with games until the model does what you need. Maybe you want a model to lie less, get better at using many different tools, or be funnier. Thesesynthetic playgroundsare how this generation of AI grows up and works better for people.Games teach us what AI can and cannot do, so it can learn to do more things for us that fit our preferences. After years playing with AI and watching AI play, I’ve learned why...Become apaid subscriber to Everyto unlock this piece and learn about:The data crisis forcing AI labs to rethink how models learnWhat happened to professionalGoplayers after AI crushed the world's best in 2016Why Alex's startup is using aCards Against Humanity-style game to solve AI's alignment problemSubscribeClick hereto read the full postWant the full text of all articles in RSS?Become a subscriber, orlearn more.
