OpenAI: AI Now Asks The Right Research Questions
OpenAI's Alex Lupsasca reveals how AI, once seen as a calculator, now matches human experts in identifying critical follow-up questions for cutting-edge research. Implications for builders.
🔬How GPT‑5 derived new results in theoretical physics and quantum gravity — Alex Lupsasca, OpenAI
With Brandon, RJ Honicky, Alex Lupsasca · Sunday, May 10, 2026
Alex Lupsasca from OpenAI discusses how GPT models have achieved superhuman capabilities in theoretical physics, detailing two groundbreaking papers on gluon and graviton scattering amplitudes where AI solved problems that stumped human experts. He explores the transformative impact of AI on research methodologies and the future of scientific publishing, while also addressing questions of AI's creativity and its role in scientific discovery.
OpenAI's Lupsasca: Ditch Papers, Your Next Report Is An LLM Chatbot
Alex Lupsasca from OpenAI argues traditional papers are dead. Future research lives as interactive LLM-powered documents, forcing AI to verify its own work.
GPT Solves Year-Long Gluon Puzzle with Simple Linear Formula
OpenAI's GPT models overturned a physics mystery, finding a simple linear formula for gluon interactions previously thought impossible. What founders can learn.
GPT-5 Reproduces Physics Paper in 30 Minutes: What Superhuman AI Means for You
OpenAI's Alex Lupsasca details how GPT-5 replicated his complex physics paper in 30 minutes. Forget basic tasks—AI now beats experts on hard problems. Here's what to do.
AI Gives Founders 'Superpowers': How to Cut 'Confused Time'
OpenAI's Alex Lupsasca reveals how AI slashes research confusion and lets physicists test 10 ideas at once. Founders, apply this AI superpower.