Game over. AGI is not imminent, and LLMs are not the royal road to getting there(garymarcus.substack.com)

94 points by FromTheArchives 1 year ago | 54 comments

Elegant and powerful new result that seriously undermines large language models(garymarcus.substack.com)

28 points by cratermoon 3 years ago | 30 comments

Seven replies to the viral Apple reasoning paper and why they fall short(garymarcus.substack.com)

293 points by spwestwood 1 year ago | 221 comments

Use Prolog to improve LLM's reasoning(shchegrikovich.substack.com)

343 points by shchegrikovich 2 years ago | 122 comments

Substack post arguing that having LLMs generate Prolog code can improve their reasoning, given the limits of autoregressive transformer generation.

Generative AI's crippling failure to induce robust models of the world(garymarcus.substack.com)

57 points by pmcjones 1 year ago | 36 comments

There are no new ideas in AI, only new datasets(blog.jxmo.io)

448 points by bilsbie 1 year ago | 240 comments

Jack Morris essay arguing that major AI breakthroughs, including LLMs, came from new datasets rather than new ideas or methods.