Paul Bakaus(@pbakaus):what’s interesting about this is that it doesn’t really work often, especially not for frontend development. there is a hard ceiling for what a model like gpt 5.6 can accomplish, and when it realizes that it cannot move the needle on a harsh judging agent, it will simply give up.

2026.08.18 16:38

what’s interesting about this is that it doesn’t really work often, especially not for frontend development. there is a hard ceiling for what a model like gpt 5.6 can accomplish, and when it realizes that it cannot move the needle on a harsh judging agent, it will simply give up.

Jacky Kwok@jackyk02

2026.08.17 18:37

Scaling self-verification with DeepSeek V4 Flash beats Claude Fable 5 on Terminal-Bench 2.1, while being 11x cheaper 💰 As open-source models become more capable, they can now generate large numbers of high-quality candidate solutions and verify their own outputs at very low cost. For example, we find that sampling just 5 solutions with DeepSeek V4 Flash and ranking them using the same model with LLM-as-a-Verifier can lead to a significant boost in accuracy (79% → 88%), outperforming closed frontier models on Terminal-Bench. Try it out today: More on verification scaling in my previous post.