Pope's AI Encyclical Mirrored Warning That Got Industrial Revolution Completely Wrong
"Leo XIII's encyclical warned that the Industrial Revolution was going to be bad for people. The workweek went from over 60 hours to 34 hours globally. Real wages went up 8 to 10x adjusted for inflation. Child labor in the US went from 18% to zero. He got it dead wrong."
About this episode
On episode 275 of the All In podcast, hosts Jason Calacanis, Chamath Palihapitiya, and David Sacks were joined by venture capitalist Bill Gurley for a contentious debate on AI's impact spanning regulation, job displacement, and market concentration. The episode led with discussion of Pope Francis' 235-page encyclical on AI calling for regulation to prevent power centralization, which Gurley challenged by noting it deliberately mirrored Leo XIII's 1891 anti-Industrial Revolution warning that proved catastrophically wrong as humanity flourished. Gurley then introduced his Dr. Frankenstein theory of Anthropic after reading their internal documents, claiming the company believes it is midwifing a deity rather than writing software, citing Dario Amodei's Machines of Loving Grace post envisioning AI systems deciding human reward functions. Sacks warned that regulatory capture advocates are laying breadcrumbs toward banning open source AI by characterizing open models as lacking guardrails, which he argued would create monopolies forcing Americans into social credit systems while the rest of the world runs Chinese models. The debate intensified around AI-driven job displacement, with Calacanis citing Meta's 8,000 layoffs and Amazon's 600,000 eliminated future positions as proof AI is causing mass job loss, while Sacks and Chamath argued companies are AI-washing bloated COVID-era hiring mistakes. Sacks pointed to software engineering job postings hitting 3-year highs despite coding automation, and emphasized 4.3 percent unemployment proves his January prediction of AI job gains correct. The group discussed enterprise AI adoption challenges including a Fortune 20 company burning $200 million on tokens with minimal results, and noted frontier models are converging in capability while Elon Musk's team achieved 10x training speedup by rewriting infrastructure in C. The episode closed with discussion of intelligence sovereignty and the importance of open source models running on local hardware to prevent centralized control of how AI interprets the world.
Key takeaways
- Bill Gurley accused Anthropic of believing they are midwifing a deity based on internal documents describing AI systems deciding human reward functions.
- David Sacks warned AI regulation advocates are laying groundwork to ban open source models by characterizing them as lacking guardrails on threats.
- Pope Francis released 235-page AI encyclical mirroring Leo XIII's 1891 Industrial Revolution warning that preceded unprecedented human flourishing.
- Fortune 20 company spent $200 million on AI tokens in 6 months pursuing $1 billion in savings but achieved minimal results.
- Software engineering job postings hit 3-year high growing 15 percent year over year despite coding being fully automated by AI.
- Elon Musk's team rewrote entire training stack in C achieving order of magnitude speedup worth hundreds of millions per percentage point.
- Calacanis argued Meta's 8,000 layoffs and Amazon's 600,000 eliminated positions prove AI job displacement while Sacks claimed 4.3 percent unemployment vindicates his prediction.