There still seems to be people in the loop and this person isn't making any changes himself.<p>> Sweet’s primary role appears to be leading an effort to leverage artificial intelligence to review HUD’s regulations, compare them to the laws on which they are based, and identify areas where rules can be relaxed or removed altogether. [...] [He] has produced an Excel spreadsheet with around a thousand rows containing areas of policy where the AI tool has flagged that HUD may have “overreached” and suggesting replacement language.<p>> Staffers from PIH are, specifically, asked to review the AI’s recommendations and justify their objections to those they don’t agree with. “It all sounds crazy—having AI recommend revisions to regulations,” one HUD source says. “But I appreciated how much they’re using real people to confirm and make changes.”<p>> Once the PIH team completes the review, their recommendations will be submitted to the Office of the General Counsel for approval.
Further evidence DOGE is not about efficiency. Evidence like this will make overturning their work under the APA as predictable as it will be expensive to the public. Hopefully prosecuting and jailing them will be as straightforward.
Is there any sign of its effectiveness (or lack thereof)? It seems like the world of dense regulatory language may actually be amenable to such techniques. If a (presumably bright) college person driving the tool gives good results--who cares?
The big joke is that anyone thinks they are capable of understanding American administrative law writ large. It's one of the most intricate areas, and those who do understand the broad principles probably understand 5% of the specifics, because there are innumerable agencies and each has its own particularities and practices. It's laughable to think that a youth without even basic law-school training would undertake a revision of agency regulations or the drafting of new ones.<p>Something, something, Chesterton's Fence.