MIT has voiced concerns about a well-known paper discussing the impact of AI on research and innovation, suggesting that due to certain uncertainties, it should be removed from public dialogue. The author of the paper, "Artificial Intelligence, Scientific Discovery, and Product Innovation," holds a Ph.D. student from MIT's economics department. It suggests that while the integration of an AI tool into a sizeable anonymous materials science lab resulted in the exploration of more materials and patent filings, it did dampen researchers’ job satisfaction.

Renowned MIT academics Daron Acemoglu (a recent Nobel Prize winner) and David Autor were among the admirers of the paper, with Autor sharing his fascination with the Wall Street Journal. In MIT's recent statement, Acemoglu and Autor recognized the paper as widely acknowledged and conversed within the AI and science literature domain, notwithstanding its absence from any refereed journal.

However, now Acemoglu and Autor express that they do not hold any trust in the authenticity, dependability or validity of the data or in the truthfulness of the research.

The WSJ said that a computer scientist with expertise in material science contacted Acemoglu and Autor to voice concerns about the research. These doubts got escalated to MIT, initiating an internal scrutiny.

The author of the controversial study appears to have left MIT, but MIT has not revealed the review's findings, citing student privacy rules. While the announcement does not specify the author, both the draft of the paper and the initial news coverage indicate him as Aidan Toner-Rodgers. (TechCrunch has tried reaching Toner-Rodgers to get his opinion on this.)

MIT expressed its desire for the paper to be retracted from The Quarterly Journal of Economics, where it was pending publication, and the preprint site arXiv. Generally, it's the authors who lodge arXiv withdrawal requests, but MIT states that Toner-Rodgers has not yet made any such request.