>assuming that AGI can be achieved by just scaling up current LLM models on text, video, audio, etc. Is any sane person actually trying this? I can't imagine

>assuming that AGI can be achieved by just scaling up current LLM models on text, video, audio, etc. Is any sane person actually trying this? I can't imagine an LLM ever going AGI.