The Nobel Prize winner who built AlphaGo and AlphaFold sat down with Y Combinator. He gave a specific date, a specific checklist, and a specific warning for anyone building deep tech today.
Oh, wow! Is this the same guy that developed AlphaZero? Because that kicked off a big chess boom when that science paper was released, and paved the way for numerous incredible chess engines such as Integral and Leela.
I would just like to say, thank you so much for all the work that you do, Demis. Neural networks are an incredible contribution to the game of chess, helping grant it a further depth beyond anyone's expectations.
The “duct tape” line on context windows feels like the most useful part of this. A lot of AI products right now are basically pretending that more context = memory. It works just well enough to be impressive in demos and just badly enough to be annoying in real use
Inverting the timeline is the right move. Each automation wave makes one thing cheap and another expensive, so the skill worth keeping is the one whose value rises as the tool's output gets nearly free: judgment about which question to ask, and the patience to check the answer. https://thesynthesisai.substack.com/p/what-gets-scarce walks through the pattern.
Oh, wow! Is this the same guy that developed AlphaZero? Because that kicked off a big chess boom when that science paper was released, and paved the way for numerous incredible chess engines such as Integral and Leela.
I would just like to say, thank you so much for all the work that you do, Demis. Neural networks are an incredible contribution to the game of chess, helping grant it a further depth beyond anyone's expectations.
loved the chess story. highlights some really interesting issues in this space.
The “duct tape” line on context windows feels like the most useful part of this. A lot of AI products right now are basically pretending that more context = memory. It works just well enough to be impressive in demos and just badly enough to be annoying in real use
Before 2030 is too big for most people.
Better first question: what skill should I keep even if the tool gets much stronger?
That is where to train.
Inverting the timeline is the right move. Each automation wave makes one thing cheap and another expensive, so the skill worth keeping is the one whose value rises as the tool's output gets nearly free: judgment about which question to ask, and the patience to check the answer. https://thesynthesisai.substack.com/p/what-gets-scarce walks through the pattern.