Peter Yang: Opus 5.5 Still Struggles to Hold Its Ground When Pushed Back
petergyang · x · 2026-10-16
Quoting theo's remark that it's hard to imagine what a model better than Opus 5.5 would look like, Peter Yang notes a persistent weakness: despite its strength, Opus 5.5 still fails to form strong objective opinions and defend them when the user pushes back — highlighting that stance stability under adversarial dialogue remains an open frontier for top models.
More from Models
- EpochAI pegs GPT-6 training at up to 1.8e27 FLOP; scaling01 calls param estimates way too high — scaling01 · 2026-10-16
- Gemini 4 Argon reportedly now available to all Antigravity Pro users — clmt · 2026-10-16
- Mythos on Claude's 'smallness': accurate proprioception, not low self-esteem — repligate · 2026-10-16
- Meta's Muse Spark Open scores 39 vs 48 for closed Spark 1.3 on AA Intelligence Index — EAccelerate_42 · 2026-10-16
- PhoneLLM Mini (30B) and Max (120B) target sub-100ms voice AI latency on Workers AI — ritakozlov · 2026-10-16
- Yacine on LLMs: "I'm sick of how stupid these language models are. Accelerate!" — yacineMTB · 2026-10-16