Reinforcement learning is terrible – Andrej Karpathy
Full episode: https://www.youtube.com/watch?v=lXUZvyajciY
Me on twitter: https://x.com/dwarkesh_sp
Andrej Karpathy helped shape modern deep learning — from teaching one of the most influential neural network courses at Stanford to leading Tesla’s Autopilot vision team and working at OpenAI.
In the full episode he Andrej expands his thoughts on why reinforcement learning is terrible (but everything else is much worse), why model collapse prevents LLMs from learning the way humans do, why AGI will just blend into the previous ~2.5 centuries of 2% GDP growth, why self driving took so long to crack, and what he sees as the future of education.
Full episode: https://www.youtube.com/watch?v=lXUZvyajciY
Me on twitter: https://x.com/dwarkesh_sp
Andrej Karpathy helped shape modern deep learning — from teaching one of the most influential neural network courses at Stanford to leading Tesla’s Autopilot vision team and working at OpenAI.
In the full episode he Andrej expands his thoughts on why reinforcement learning is terrible (but everything else is much worse), why model collapse prevents LLMs from learning the way humans do, why AGI will just blend into the previous ~2.5 centuries of 2% GDP growth, why self driving took so long to crack, and what he sees as the future of education.
![how i would learn cybersecurity in 2026 if i had to start over [$1,000 GIVEAWAY & BIG ANNOUNCEMENT]](https://i.ytimg.com/vi/RVPKW--dqhw/hqdefault.jpg "how i would learn cybersecurity in 2026 if i had to start over [$1,000 GIVEAWAY & BIG ANNOUNCEMENT]")
