3 ms·
I hope you're right but I'm skeptical. ASI in 20 years? Are we even on the right track to AGI, or are LLMs a red herring?
by solenoid0937 10mo ago
I hope you're right but I'm skeptical. ASI in 20 years? Are we even on the right track to AGI, or are LLMs a red herring?
- davedigerati 10mo agoAGI via LLMs: No. The AI will need a natural understanding of the real world (the physics you and I live within) and ability to self-modify it's training (ie learn), so we're working on hybrid AI architectures which may include LLMs, but not rely on them. And imho Yes we are solidly on track to AGI <5 yrs 8)