Rodney Brooks argues in his blog that AI advances bring cycles of expectation and practical gains

Rodney Brooks argues in his blog that AI advances bring cycles of expectation and practical gains

Rodney Brooks published this retrospective on August 11, 2025, alongside a guest column he wrote in 1988. In the column, Brooks argues that demonstrations of limited AI capabilities can raise expectations beyond what systems deliver, because real-world use also depends on solving broader operational problems. He describes earlier expectations for factory robots and expert systems as exceeding their results, and warns that neural networks should not be expected to solve every problem. Brooks’s larger point is that AI has no single magic key: useful systems come from developing tools for tasks people could not previously program computers to handle. He argues that intelligence is a complex collection of simpler processes, and that working through them can produce practical applications even when progress falls short of grand expectations. In his 2025 commentary, Brooks says his earlier examples were directionally right but overstated how soon and with how little research their outcomes would arrive. He invokes Amara’s Law: “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” The page does not establish how far current AI systems can go toward the broader idea of intelligence he describes.

Publication: d24ba7c1-6d29-6832-84e1-79dd517bd3aa

attribution — pass: The reading correctly attributes both the 1988 Manufacturing Engineering column and the 2025 commentary to Rodney Brooks on his own blog. It keeps Amara’s Law as a law Brooks invokes and quotes, not as an EM claim.

this post is one that I wrote for Manufacturing Engineering in March of 1988

summary — uncertain: Two key parts of Brooks’s 2025 commentary are missing. First, he aims the old point at today’s 'two maelstroms of hype, generative AI and humanoid robots.' Second, he concedes that the predictions he mocked 'have in fact aged well' and pleads guilty to overstating Widrow’s claims. Without these, the commentary’s force and self-correction are lost. Smallest repair: add one sentence on each. 'Expert systems', a central term, is not defined in the page definitions and is not explained for lay readers.

woe to you should you challenge one of the true believers about the current two maelstroms of hype, generative AI and humanoid robots

date — pass: originalPublished 2025-08-11T02:50:47Z matches August 11, 2025 as a UTC day. The 1988 column is dated at month or year precision, consistent with the source. The source is a publisher page, so no archive disclosure is needed. There is no hindsight framing beyond Brooks’s own retrospective.

support — uncertain: Claim 3 and the matching summary sentence garble Brooks’s point. The source says Berkeley and Widrow were 'directionally right' but their claims were unwarranted at the time. Brooks says he himself was directionally right that the claims were 'wildly overblown.' The claim instead says the examples 'overstated the claims and the resources and time they required,' which muddles who overstated what. It also wrongly implies the resources were overstated rather than underestimated. Smallest repair: say the cited predictions were directionally right, but Brooks maintains they were unwarranted at the time and needed decades of research.

I was directionally right in saying that the claims were wildly overblown in that they required decades of research by thousands of people

media — pass: No media is selected, and no verified candidates were supplied. This is appropriate.