Karpathy's frank breakdown of why reinforcement learning is fundamentally limited and where LLMs have cognitive blind spots gives engineers a realistic mental model for what current models can and can't do — sharper than the usual AGI-hype narrative.
Key quotes
“Reinforcement learning is terrible. It just so happens that everything that we had before it is much worse”
Andrej Karpathy
“The way I like to put it is you’re sucking supervision through a straw.”
Andrej Karpathy
“we’re not building animals. We’re building ghosts or spirits or whatever people want to call it”
Andrej Karpathy
“One easy way to see it is to go to ChatGPT and ask it, “Tell me a joke.” It only has like three jokes.”
Andrej Karpathy
“I feel like the industry is making too big of a jump and is trying to pretend like this is amazing, and it’s not. It’s slop.”
