MIT Technology Review on technology flops and Sam Altman’s AI hype

MIT Technology Review on technology flops and Sam Altman’s AI hype

MIT Technology Review’s December 18, 2025, edition of The Download introduces an annual selection of technology flops and a look at Sam Altman’s influence on how people imagine AI. The newsletter suggests that some technological misadventures offer a lesson, but when technology depends on power, the simpler lesson may be to avoid it. In James O’Donnell’s account, Altman has been an influential voice behind striking ideas about AI, even when those ideas were rarely provable at the time he voiced them. O’Donnell says Altman’s persuasive case links a road toward AI’s great or terrifying possibilities with the need for OpenAI to raise large sums to steer it. The newsletter says it examined almost everything Altman had said about the technology to trace how his voice shaped understanding of AI’s capabilities. A separate item on AI and materials research describes a possible climate-tech use: discovering materials for batteries, semiconductors and magnets. It also says the field still needs to show that it can produce materials that are both novel and useful. The edition does not name the flops in its opening, or establish whether AI can deliver useful new materials.

Publication: 05c95305-250c-cf3c-4dfa-d412f9de6a4b

attribution — pass: This is one newsletter issue that comments on its pieces in its own words, not a bare listing. The reading attributes the Altman item to James O’Donnell and treats the flops and materials items as newsletter content, which matches the bylined blurbs. One sentence says “the newsletter” examined Altman’s statements, while the blurb’s “we” belongs to O’Donnell’s story. Claim 3 credits O’Donnell correctly, so this is not material.

we read almost everything he’s ever said about the technology

summary — pass: A general reader can follow the summary: the flops list, O’Donnell’s argument about Altman’s influence and fundraising framing, and the materials-research caveat. Altman is tied to OpenAI through the fundraising point. The source’s sharper framing, that Altman is “the ultimate hype man” whose words “set the agenda,” is somewhat softened, but the title and the “rarely provable” point keep the argument’s force. The question follows directly from the source’s stated open issue.

In this sense, he is the ultimate hype man.

date — pass: The originalPublished instant, 2025-12-18T13:10:00Z, is December 18, 2025 in UTC, which matches the reading and the URL path. The reading discloses that this is an archived copy and gives the retrieval date. The archive capture time on the same day is not a substantive edit. The reading makes no hindsight claims and does not draw on the later related-story sidebar.

support — pass: Each claim is supported by a verbatim excerpt from the source: the advice that it would have been better to stay away from power-dependent technology, Altman’s claims being rarely provable, the review of his statements, and AI’s potential for materials research alongside the need for proof. The reading presents the materials benefit only as possible. It flags as unknown whether AI has produced materials that are both novel and useful, and adds no invented stakes.

But the field still needs to prove it can make materials that are actually novel and useful.

media — pass: The reading selects no media. No verified media candidates were supplied, so this choice is appropriate.