ChatGPT 3 turned 3 today. It still hasn't come close to meeting expectationsgarymarcus.substack.com·10 pts·garymarcus·2
Why do all LLMs, even the most recent ones, hallucinate so often?garymarcus.substack.com·1 pts·garymarcus·0
Why the future of AI is neurosymbolic. (A rare optimistic post from Gary Marcus)garymarcus.substack.com·36 pts·garymarcus·36
The biggest risk of large language models is that they may bury us in liesscientificamerican.com·44 pts·garymarcus·39
Alt Intelligence vs. AI (Gary Marcus Critique of DeepMind’s de Freitas)garymarcus.substack.com·2 pts·garymarcus·0