Flutterby™! : DOGE dipshits and LLMs

Next unread comment / Catchup all unread comments User Account Info | Logout | XML/Pilot/etc versions | Long version (with comments) | Weblog archives | Site Map | | Browse Topics

DOGE dipshits and LLMs

2025-06-06 17:33:33.546949+02 by Dan Lyke 0 comments

The thing is... It seems pretty clear that people who think the AI are smarter than they are are probably right.

Pro Publica: Inside the AI Prompts DOGE Used to “Munch” Contracts Related to Veterans’ Health

This was error-prone and not necessary, as accurate contract information can already be found in publicly available databases like USASpending. In some cases, this led to the AI system being given an outdated version of a contract, which led to it reporting a misleadingly large contract amount. In other cases, the model mistakenly pulled an irrelevant number from the page instead of the contract value.

Ars Technica: DOGE used flawed AI tool to “munch” Veterans Affairs contracts

Peter @peter@thepit.social summarized as:

this guy 100% set up a RAG pipeline on a local GPU running Llama 3.2 or whatever. there's a whole class of dudes who have had their brains melted by this stuff, it's basically a religious belief, they are **convinced** they can do magic with this stuff and have utter contempt for any kind of real knowledge or expertise.

[ related topics: Religion Health Sports Artificial Intelligence Clowns Databases ]

comments in descending chronological order (reverse):