Learning from Nature
Andreas discusses the differences between human and AI representations, highlighting the challenges AI faces with unknown situations. He emphasizes the need for machines to learn from biological brains, suggesting that understanding inductive biases could improve generalization beyond training data. The conversation also touches on the inefficiencies of current data collection methods in AI, advocating for a more intelligent approach to learning from fewer examples.In this clip
From this podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Engineering a Less Artificial Intelligence with Andreas Tolias - #379
Related Questions