LLM
2 articles
Karpathy: Spent 4 Hours Polishing an Argument with an LLM, Then Asked It to Argue Back and Got Demolished
Andrej Karpathy spent four hours polishing an argument with an LLM, felt invincible, then asked the same LLM to argue the opposite — and got completely dismantled. LLM sycophancy is a real trap, but flipping it around is genuine alpha.
Disaggregated Planning: Why the Strongest Reasoning Model Isn't Always the Best Coder
SemiAnalysis argues that agentic coding might follow the same pattern as disaggregated prefill—splitting planning from execution. If the spec is good enough, cheaper models could handle execution, cutting costs significantly.