Google's Jeff Dean on the Coming Transformations in AI
- Source
- youtube.com
- Author
- Sequoia Capital
- Date
Context
Jeff Dean, chief scientist at Google's parent Alphabet, says in this Sequoia Capital conversation that many fields have computationally expensive simulators, such as for weather, fluid dynamics or quantum chemistry. A scientist can use a simulator's outputs as training data for a neural network that approximates it.
He says the approximation can be around 300,000 times faster, which he argues changes the workflow: instead of running a simulator for a year on compute you lack, you could screen 10 million molecules over lunch. He offers this as an example of AI influencing science broadly. It is his general account in a talk, not a result for a named project, and it does not say how accurate the approximations are.
videoBringing Continual Learning to Every Company: Trajectory’s Ronak Malde
videoCrusoe CEO Chase Lochmiller: Turning Stranded Energy Into Intelligence
videoDavid Silver and Lasse Espeholt of Ineffable Intelligence: A Superlearner for the Age of Experience
videoCracking the Code on Offensive Security With AI ft XBOW CEO and GitHub Copilot Creator Oege de Moor
Checking sign-in…
Loading comments…


