general
critique
bearish
There is a fundamental conflict in how the term AGI is being used: one group claims AGI will be superintelligent enough to kill humanity, while another claims we've already achieved AGI because LLMs beat average people at some tasks
Gary Marcus08 Aug 2026