Gary Marcus meets Pope Leo XIV: five AI principles and a critique of LLMs
Gary Marcus · rss · 2026-10-31
Gary Marcus recounts meeting Pope Leo XIV at the Vatican, calling him the clearest thinker on AI among world leaders and praising the encyclical Magnifica Humanitas. The Pope's five points: AI's future depends on human choices; machines are tools, not persons (rejecting legal personhood); decisions over life and force must stay under human oversight, with support for banning lethal autonomous weapons at November's Geneva talks; AI must serve the common good; and the urgent need for international institutions with shared standards and accountability.
Marcus also summarized his own remarks at a private session (after Tegmark, Russell, and Pisano): agreeing the current path is untenable, and arguing the field's fundamental mistake is equating prediction—LLMs' one superb component—with all of intelligence. As a cognitive scientist he stresses the mind is "not one thing but many," and claims current models still reason poorly and fail to follow instructions, relying on correlation over causation with superficial semantics.
More from AGI Musings
- AI-generated imagery is already everywhere in the physical world, and it's hard to unsee — signulll · 2026-11-01
- LLM-aided content is prettier but painfully samey — podcast host longs for human voice — TOEwithCurt · 2026-11-01
- In 1956 AT&T was forced to open 7,820 Bell Labs patents — innovation rose 32% — krishnan · 2026-11-01
- Matthew Yglesias: many 'normal technology' AI apps are worsening human welfare — wfithian · 2026-11-01
- Cognitive sovereignty in practice: it's about who writes the first draft — DrKavner · 2026-11-01
- Reddit debate: AI replaces honest jobs while scam jobs stay safe — patmull · 2026-10-31