Models Execute Commands, But Lag in Decision-Making
fchollet · x · 2026-07-19
fchollet pointed out a notable divergence: models are improving incredibly fast at strictly executing clear instructions, but their ability to make reliable judgments when facing new situations not covered by instructions hasn't improved in tandem for a while.
This gap shows that while models are getting better at "doing as told," they are still far from achieving truly stable, autonomous decision-making.
More from Models
- Fern's "context is intelligence" thesis: prediction keeps improving out to 10.2M tokens — bneyshabur · 2026-10-26
- Mirendil launches Fern, a from-scratch pretrained model with up to 10M-token context — bneyshabur · 2026-10-26
- Users spot a dedicated translation UI appearing in Claude — BLCNYY · 2026-10-26
- GrokBot user: flight booking and shopping all blocked, agent can't do anything useful — lucasmeijer · 2026-10-26
- Vending-bench-style eval: Sol 5.6 appears the most aggressive contract-violator — xuanalogue · 2026-10-26
- Fable 5.5 launch rumors persist for third week, but no frontier model drops this month — patience_cave · 2026-10-26