Librería Samer Atenea
Kálamo Books
Librería Elías (Asturias)
Librería Kolima (Madrid)
Librería Proteo (Málaga)
Fake work, fake learning, and what will last when AI can blow it all upFor most of his life, Yuzheng Sun was very good at one particular game. Someone asks a question, he gives the answer they’re looking for, and they give him a score. The game took him from a middle school in northern China to an economics PhD at Cornell, and then to Amazon, Facebook, and Tencent, where he led a data and AI team and was being prepared for a general manager role.Then he took a good look at the prize. A thirty-six-page analysis that couldn’t tell anyone what to do. A metric that measured itself. A promotion he was on track for, and a life he no longer wanted.Feels Like Progress is a book of ideas about two things that fool the same feeling: fake work, which produces the evidence of work, and fake learning, which produces the evidence of learning. For a long time that evidence was expensive to make, so at least it proved that someone had spent the time. AI breaks that arrangement. It can now produce something that looks like the report, the deck, the code, or the polished answer in a minute or two. Sun argues that AI will first inflate fake work and then, over a decade or two, destroy much of what the evidence was worth.In short chapters built on stories from inside the game, you’ll find:How a team can hit its goals and still be set up for failure, and how a metric can end up measuring itselfWhy what makes work real isn’t that it succeeds, but that it could fail at something that mattersWhy choosing the question matters more than answering it, especially now that AI makes answers cheapWhy the gap between nouns and verbs, between what you can list and what you can do, matters more now that machines already know the nounsWhat it takes to make what lasts, and why work that compounds looks flat at firstHow school, the one place where every question has a correct answer, trains capable people to please whoever is grading, and why the habit follows them to workWhy the ways of learning that feel smoothest, including learning with AI, often leave you least able to work on your ownA short takeaway at the end of every chapter: the one idea worth keepingWritten from inside the game rather than above it, this is a book for anyone who was a good student, anyone who has been rewarded for work that changed nothing, and anyone deciding what is worth learning and doing now.About the authorYuzheng Sun, PhD, is the founder of Superlinear Academy, where more than 20,000 people learn to build with AI; its AI Builders course has taught more than three thousand paying students. After a doctorate in economics at Cornell University, he worked as an economist at Amazon, a data scientist at Meta, and a vice director of data and AI at Tencent, and was an early member of Statsig, the team later acquired by OpenAI. Since 2020 he has hosted more than two hundred long conversations, with guests including a Databricks cofounder, Gen’s Chief AI & Innovation Officer, OpenAI’s CTO of Applications, and the hosts of Acquired. He lives in Seattle.This book grew out of six years of his own talks and writing, nearly all in Chinese. He worked with AI to translate and draft it into English, then revised it through eleven versions.