The Scale ReportOpenAI Faces Backlash Over Aggressive Mathematics Research
Mathematicians describe a cutthroat culture at OpenAI as the firm chases prestigious proofs while sidelining established academic researchers.
Section
Research
The latest research coverage from The Scale Report desk.
The Scale Report
The Scale ReportOpenAI Faces Backlash From Top Mathematicians Over Proof Claims
Twenty-five Fields Medal winners accuse AI labs of prioritizing speed over academic integrity, citation, and verification.
The Scale ReportOpenAI Accused of Scooping Academic Team on Historic Math Problem
An NYU professor claims the AI giant weaponised leaked progress and tens of millions in compute to race ahead on Navier-Stokes.
The Scale ReportHarvard and MIT Researchers Unveil 8.3 Billion Synthetic AI Personas
The MatrAIx simulation system models consumer behavior across 1,290 traits to stress-test software before launch.
The Scale ReportAnthropic Research Shows Automated Systems Outpacing Human Engineers on AI Alignment
A new study finds automated agents can patch model vulnerabilities faster and at a fraction of human labor costs.
The Scale ReportBeijing Medical Resident Uses AI to Produce Candidate Proof for Decades-Old Math Conjecture
A 16-hour autonomous session with OpenAI's system yielded a potential solution to Crouzeix’s Conjecture, a matrix theory problem unsolved since 2004.
The Scale ReportThe Data Efficiency Gap: Why Toddlers Still Trounce Trillion-Token AI Models
As web data reserves dwindle, AI researchers and cognitive scientists are turning to child development to figure out how to learn more from less.
The Scale ReportMIT System Uses AI and Muscle Stimulation to Directly Actuate Human Hands
The Media Lab’s 'Human Operator' project translates visual data and spoken prompts into electrical pulses that move a user's fingers and wrist.
The Scale ReportMIT Media Lab Uses AI and Muscle Stimulation to Directly Guide Human Hands
The experimental 'Human Operator' setup pairs Anthropic's Claude with electrical pulses to physically steer motor tasks.
The Scale ReportBeijing Medical Resident Uses Autonomous AI Run to Tackle 22-Year-Old Math Conjecture
A 16-hour session with GPT-5.6 Sol produced a candidate proof for Crouzeix’s Conjecture, an open problem in numerical linear algebra.