
Rocky Duan: from OpenAI to Covariant and building foundation models for robotic automation
The Robot Brains Podcast
00:00
The Power of Reinforcement Learning in Machine Learning
We don't really need to prescribe how exactly the computation happened in terms of code. Design the architecture that enables on those priors. And then the windows we can build powerful systems turns out to work quite well against across very, very different disciplines. The most popular paradigm used in the industry is supervised learning and reinforcement learning.
Transcript
Play full episode