Physicists Surrendering to AI Overlords

Elite physicists are surrendering human ingenuity to AI overlords at major conferences, risking the soul of American scientific innovation in an era demanding real American leadership.

Story Highlights

Physicists at high-profile gatherings like Flatiron Institute and IAS demos show AI handling entire research workflows, from literature to simulations.
David Kipping warns of “abdicating agency” to AI out of fear, devaluing human struggle central to discovery.
Steve Hsu’s AI-assisted paper in PLB sparks “AI slop” debates, with errors infiltrating peer-reviewed physics.
Journals turn to AI for reviews amid overwhelming volume, eroding trust in scientific rigor.

AI Takes the Wheel in Theoretical Physics

Flatiron Institute researchers demonstrated AI managing end-to-end cosmology research in mid-2025, starting with literature reviews and ending in simulations. Institute for Advanced Study astrophysicists followed suit, letting AI direct research paths. This shift prioritizes speed over deep understanding amid publish-or-perish pressures at major physics collaborations. Conservative values emphasize individual merit and human effort; AI dominance raises alarms about outsourcing American genius to machines controlled by Big Tech elites.

Critics Sound Alarm on Agency Loss

Columbia astrophysicist David Kipping highlighted physicists surrendering to AI at IAS-linked events in his late 2025 video, calling it fear-driven abdication. Peter Woit, a Columbia mathematician, critiqued the influx of “AI slop”—plausible but flawed ideas polluting journals. Kipping argues this erodes the beauty of human understanding, a cornerstone of traditional scientific pursuit. In 2026, as America fights costly wars abroad, we cannot afford to cede intellectual sovereignty at home.

Controversial Papers and Publisher Responses

Steve Hsu published a December 2025 PLB paper using undisclosed large language models for core directions, fueling debates on transparency. Journals like Nature report over 50% of researchers using AI in peer reviews against guidelines. arXiv implemented stricter endorsement policies on December 10, 2025, to combat slop. Publishers face volume overload, deploying AI referees and sparking calls for LLM co-authorship. This chaos foreshadows broader scientific malpractice, undermining merit-based progress conservatives champion.

UChicago announced human-AI hybrids in February 2026 for tough problems, blending optimism with caution. Diverse views range from acceleration enthusiasts to pessimists fearing internet-like misinformation floods.

How AI shook the world's largest meeting of physicists | New Scientist https://t.co/iOIKaMI5Qj

— Andro (@AndroOxinu) March 24, 2026

Implications for Science and Society

Short-term, AI overwhelms peer review, worsening signal-to-noise ratios and dividing physicists into AI users versus traditionalists. Long-term, it could dismantle insane publish-or-perish incentives, spurring reform. Economic burdens include high article fees funding AI tools; socially, it devalues human struggle in discovery. Theoretical physics suffers most, contrasting empirical fields. As MAGA supporters question endless wars and government overreach, this AI creep in elite institutions alerts us to threats against American innovation and self-reliance.