AI performance already overkill? Suggests daily use of Kimi-K3, occasional Fable or GPT5.6; OpenAI/Anthropic should pivot to cybersecurity

A user argues that AI performance has surpassed the "good enough" threshold. For daily tasks, Kimi-K3 suffices 80% of the time; only when needed, Fable or GPT5.6. He suggests OpenAI and Anthropic should offer cybersecurity services instead of nerfing models, as they could use unrestricted models internally.

• Anthropic
• OpenAI
• 网络安全
• Fable
• AI性能
• GPT5.6
• Kimi K3