Gary Marcus Discusses the Anti-LLM Bubble Theory

GaryMarcus · x · 2026-07-20

Gary Marcus reposted a highly pessimistic comment regarding current Large Language Models (LLMs) and asked netizens for the best counterargument.\n\nThe quoted tweet argues that the AI industry's hype is mostly lies, and that LLMs are merely useful search tools for processing large datasets (a commodity) rather than revolutionary magic. The tweet even asserts that leading AI companies like OpenAI and Anthropic will face difficulties, and the bubble across the entire industry has already burst.

Original post →

More from AGI Musings

AGI Musings channel →