Technology

God this is beyond ridiculous.

They used ChatGPT to sort over 1,400 National Endowment for the Humanities grants into spreadsheets. Their exact prompt to the AI was "Does the following relate at all to DEI? Respond factually in less than 120 characters. Begin with 'Yes.' or 'No.' followed by a brief explanation."

They didn't define what they meant by DEI. They just asked ChatGPT to figure it out. And then cancelled grants based on whatever it said. Probably didn't even read past the yes or no.

The government's defence was essentially ChatGPT did it, not us. When asked if he regretted people losing income, he said he didn't, because reducing the deficit was more important. When asked if they actually reduced the deficit, he admitted they did not.

They caused real harm, achieved nothing they claimed to be pursuing, and expressed zero regret about it.

Just really horrible all around.

Those fucking assholes.

Restoring grants to projects that don't exist anymore and the criminals are punished by being called dumb. This is a great system and it really does work.

You know what I hate? That we don't mock the fact that DOGE was named after a fucking dog meme and a meme coin.

That is literally the origin of the name of that de facto now apparently defunct "governmental institution".

NOW CONVICT THOSE RESPONSIBLE

Yeah but the damage happened in real time. There would need to be standing to sue for the damages cause and accumulated since.

Anecdotal but it's a real story about getting doge'd. I was working on some climate research as part of a climate smart commodities grant (search that program for details), a big one which took years to gain the trust of hundreds of indigenous farmers and non profits to participate in. We were already struggling with USDA bullshit and red tape trying to get payments to farmers who had already spent money and been approved for certain payments, before doge came along.

And because of doge, the entire 50mil program got cut. Maybe not hundreds but dozens of farmers and non profits went out of business because they chose to trust the federal government to be reliable on their word. Not to mention an almost 40 person research team, practically a whole department.

and the trust that took years to build to get these historically disenfranchised groups to the table, to get them to trust the process and work with the government.. that's NEVER coming back.

This line from the article about DOGE still trying to shift responsibility to ChatGPT.

McMahon also pushes back on the government’s argument that “there is no real constitutional problem here because any viewpoint-based classification was ChatGPT’s doing, rather than the Government’s:”

"There is no distinction to be drawn here between the Government and ChatGPT. ChatGPT was the Government’s chosen instrument for purposes of this project, and DOGE’s use of AI to identify DEI-related material neither excuses presumptively unconstitutional conduct nor gives the Government carte blanche to engage in it. ...There is not a scintilla of evidence that Fox or Cavanaugh, having obtained a “DEI” rationale from ChatGPT, undertook any meaningful review of whether that rationale made sense."

After being deployed from DOGE to NEH, Justin Fox used search terms, which he labeled as “Detection Codes,” to identify grants that he dubbed the “Craziest Grants” and “Other Bad Grants.” The search terms included, among other terms, “BIPOC (Black, Indigenous, People of Color),” “Minorities,” “Native,” “Tribal,” “Indigenous,” “Immigrant,” “LGBTQ,” “Homosexual,” and “Gay.” When Fox was asked whether he “r[a]n this list of words through every grant description” he received from NEH, he confirmed, “yes.”

I'm a bit at a lost for words here. That is way beyond what I was expecting.

Probublica did a piece on something similar, they just used llm's to flag words that they considered woke, and in the process cancelled a bunch of grants that had nothing at all to do with what they were targeting. They couldn't be bothered to actually check the work of the llm.

Shocking that these fascist clowns aren't good at governing even for their own ill ends.

Honestly I don't think this is the article but idk it's similar, this is just in regards to veterans affairs. https://www.propublica.org/article/inside-ai-tool-doge-veterans-affairs-contracts-sahil-lavingia

Wait, doge used gpt not Grok?

Did the judge use AI to decide that they violated 1th and 5th amendments?