Bovine 27

Jonas Hallgren's viral post 'Bovine 27' presents a satirical dialogue between Socrates, Plato, and Aristotle debating an absurd prediction: that bovine biomass will exceed human biomass by a factor of seventeen within three generations. Using a creative methodology, Hallgren generated character dialogues with Claude AI, then edited approximately 90% of the content to craft a coherent philosophical critique. The piece was posted to LessWrong on March 13, 2026, and quickly gained traction within AI and rationality communities.

The core satire targets common pitfalls in AI forecasting and existential risk assessment. The philosophers cite four 'independent' lines of evidence—mathematical projection, empirical observation, Platonic Forms, and divine revelation—only for Socrates to reveal they're actually circular references. This mirrors real concerns about how AI predictions can create false consensus through methodological echo chambers. Hallgren describes the piece as 'good faith satire' while expressing support for prediction exercises like 'AI 2027,' making it both critique and contribution to forecasting discourse.

The post's viral success stems from its clever framing of contemporary AI safety debates through classical philosophy. By having ancient philosophers debate bovine overpopulation using modern forecasting terminology, Hallgren highlights how even rigorous-seeming methodologies can produce questionable conclusions when assumptions aren't properly examined. The piece has sparked discussions about epistemic humility, the difference between inside and outside views in prediction, and how to maintain intellectual rigor while making forecasts about transformative technologies.