LOGO

MIT Disavows Doctoral Paper on AI Productivity - Controversy

May 17, 2025
Topics:AIMITTC
MIT Disavows Doctoral Paper on AI Productivity - Controversy

MIT Requests Withdrawal of AI Research Paper

The Massachusetts Institute of Technology (MIT) has requested the withdrawal of a prominent research paper. This action stems from significant concerns regarding the paper’s “integrity” and the validity of its findings.

Paper's Core Claims

The paper, titled “Artificial Intelligence, Scientific Discovery, and Product Innovation,” was authored by a doctoral candidate within MIT’s economics department. Its central argument posited a correlation between the implementation of an AI tool in a materials science laboratory—the specific lab remains unnamed—and an increase in both material discoveries and patent applications.

However, this apparent progress was reportedly accompanied by a decrease in job satisfaction among the researchers involved.

Initial Praise and Subsequent Concerns

Last year, the paper garnered praise from esteemed MIT economists Daron Acemoglu, a recent Nobel laureate, and David Autor. Autor notably expressed being “floored” by the research in an interview with the Wall Street Journal.

In a statement released alongside MIT’s announcement, Acemoglu and Autor acknowledged the paper’s prior circulation within the academic community. They stated it had been “already known and discussed extensively in the literature on AI and science,” despite lacking publication in a peer-reviewed journal.

Loss of Confidence in Research Validity

Subsequently, both economists expressed a complete lack of confidence in the research’s foundation. They now question the “provenance, reliability or validity of the data” and the overall “veracity” of the study.

Investigation Triggered by External Concerns

The reassessment was initiated in January when a computer scientist specializing in materials science contacted Acemoglu and Autor. This individual raised concerns about the paper’s methodology and data. These concerns were then brought to the attention of MIT, prompting an internal review.

Review Findings and Author's Status

MIT cites student privacy regulations as preventing the disclosure of the review’s specific findings. However, the university confirmed that the paper’s author is “no longer at MIT.”

While MIT’s announcement does not explicitly name the author, both the preprint version of the paper and initial news reports identify Aidan Toner-Rodgers as the individual responsible. TechCrunch has attempted to contact Toner-Rodgers for a response.

Withdrawal Requests and Current Status

MIT has formally requested the paper’s withdrawal from both The Quarterly Journal of Economics, where it was under consideration for publication, and the preprint server arXiv.

It is standard practice for authors to submit withdrawal requests to arXiv. However, MIT reports that “to date, the author has not done so.”

#MIT#AI#artificial intelligence#productivity#research#disavow