Is Neural Networks the Way to General Intelligence?

The efficiency aspect is really key to us. So the ability to learn within your lifetime and learn very quickly and acquire skills very efficiently is actually like a really core thing that I think a lot of our current systems are really missing. And so we use whatever tools work. We're very willing to look out for other baselines and try things out and see what works. The Avalon reinforcement learning environment, a benchmark for generalization and RL. It has a shared reward function across all tasks. In order to get an agent to do something in the environment, really the environment is changed. Its reward is the energy at the end of a level. You can gain energy by eating food and lose energy

Transcript

Play full episode

The AI-powered Podcast Player

Save insights by tapping your headphones, chat with episodes, discover the best highlights - and more!

Get the app