Frontier AI is probably already more accurate than most individual humans across a broad range of cognitive work. How is this not AGI? Are we just constantly moving the goal post?
热度趋势
百分比基于当前可用热度信号，而非评论数或独立用户人数。
OpenAI 相关模型动态已经出现，适合跟踪能力变化、生态影响和后续可用性。
最近发布的GPT-5.6 Sol等前沿模型展示了令人惊叹的能力，引发了关于这些进展是否已达到通用人工智能（AGI）的讨论。用户反映，像Codex这样的模型在处理复杂任务时，其准确性和效率已超越大多数个体人类，大大减少了返工需求。这种提升的性能使得用户能够完成以前无法想象的项目，表明通过人工智能，智能正日益成为一种商品。
Recent releases of frontier models like GPT-5.6 Sol have demonstrated insane capabilities. A lot of the tasks I am handing over to agents like Codex nowadays are extremely complex. We are talking about tasks that would take a human much more time and dedication. Before GPT-5.6 sol, I thought the main advantage of AI was that it can type faster than you. It can code faster than you. It doesn’t matter if you need to go back and fix the code because that takes less time than writing it yourself. Now Codex writes the code in a way where I don’t have to go back and fix it most of the time. I have been able to build and maintain projects that I couldn’t even dream of building without AI. The little rectangle we keep in our pocket is now a window to greater intelligence. I feel like we can finally talk about Intelligence like it’s a commodity.