Self-Proclaimed Alpha Warrior Pete Hegseth who against all reason was appointed and still remains Secretary of Defense is on a rampage. Little Petey it seems doesn’t like the notion of safeguards on the DOD’s Artificial Intelligence platform. He wants said safeguards removed. (What could possibly go wrong? Yes, I’m being snarky) The contractor is resisting so tomorrow a “showdown” meeting is set. With the SOTU tomorrow evening, and the likelihood both the military and tech/AI will come up this could be interesting.
Let’s set the stage. According to this article from The Daily Beast Petey ‘really, really means it’ when he says the contractor has to abandon it’s ‘cautious approach’ to the DOD’s AI systems. Call me crazy but giving even a normal administration giving DOD unfettered control over a massive AI system is rather less than a good idea. Hegseth however is demanding, including with threats that the contractor give in to his demands, ‘or else.’ It will come to a head tomorrow (Tuesday) morning:
The defense secretary is set to host Anthropic CEO Dario Amodei on Tuesday morning for what Axios reports is “likely to be tense meeting” on the military’s use of the company’s Claude software.
“Anthropic knows this is not a get-to-know-you meeting,” as one defense official described it. “This is not a friendly meeting. This is a s–t-or-get-off-the-pot-meeting.”
Many, including the head of Anthropic have warned of the potential for misuse of AI. Worries about the rush to develop the technology have long existed. Sometimes the debate has gotten pretty intense. From current concerns about use of AI that isn’t fully developed for tasks it’s already doing (customer ‘service’ phone calls for example) to ‘rise of the machines’ type stuff like in the Terminator and Matrix movies. Let’s stick with some of the current capabilities we know exist and how they can be abused. More specifically how AI can be abused to get us closer to Big Brother Is Watching territory than we realize is already the case.
Given Trump’s desire to use not just federalized National Guard but active duty military on U.S. soil against U.S. citizens the thought of the Pentagon using AI enhanced facial recognition is a quite real concern. It’s exactly the type of thing Anthropic’s CEO Dario Amodei has warned against. He’s been a strong advocate for safety protocols when it comes to use of AI and that’s the point of contention between him and Hegseth that will be talked about tomorrow:
Amodei, who’s often warned of AI’s potential for misuse, has consistently pushed to frame his firm as a safety-conscious leader in the sector.
He has so far resisted pressure from Hegseth to remove safeguards on the Pentagon’s Claude-enabled programs unless the department agrees to wall off mass surveillance of citizens and research into weapons capable of firing without a human operator.
Their feud is understood to have escalated amid reports that Claude was used by the Pentagon in the Trump administration’s lightning invasion of Venezuela earlier in January.
Prospects for tomorrow’s meeting being cordial, or even respectful seem remote. The two sides can’t even agree on where things stand as of today! Regarding their discussions with the Pentagon Anthropic told news outlet Axios “we are having productive conversations, in good faith.” Meanwhile, the DOD says “negotiations have shown no progress,” and are now “on the verge of breaking down.” Add in the fact that earlier this month Hegseth said he’d consider certifying Anthropic as a “supply chain risk” if it did not yield to his demands and it seems not ketchup bottles but turd-grenades will be tossed during the ‘showdown’ meeting.
Officials said that at the Tuesday meeting, the secretary now plans on issuing Amodei with an “ultimatum.”
“The problem with Dario is, with him, it’s ideological,” one senior Defense Department source said. “We know who we’re dealing with.”
Tomorrow isn’t just ‘sh*t or get off the pot’ day at the Pentagon. It’s also SOTU day and if journalists have the guts to cover this story it will be a significant part of pre-speech news coverage. Trump, like others before him has been teasing excepts from the speech to shape the coverage. Copies of the full address are probably already in the hands of editors and top talking heads of major news outlets. A Hegseth created showdown/sh*t show with DOD’s AI contractor is NOT the kind of pre-speech coverage Trump wants or needs.
Keep an eye on this. As I said both the military and tech/AI will be topics in the SCOTUS address Trump has seen and prepared (by his standards) to give. Breaking news about a showdown, a break between Hegseth/the Pentagon and Amodei/Anthropic could force last minute rewrites of parts of the speech. You know what that means. Trump, who has enough trouble staying on topic and the teleprompter might well decide to ad lib some commentary.
If so it’s hard to imagine it going well for him, and could be a ‘bigly’ story in and of itself. With tens of millions of people watching him show his ass in a word-salad diatribe including his MAGAs his already abysmal approval ratings will get even worse. Congressional Republicans, who increasingly fear loss of not just the House but the Senate too will run for cover if asked to comment. Keep in mind MAGAs are more likely to be worried about tinfoil hat ‘rise of the machines’ stuff and if Trump blows it when talking about the military and AI it might resonate with them. In a way not good for Trump.
So pay attention to the news tomorrow. It should be a helluva news day.
Friends, I know everyone begs you for money. I promise, among all those asking for spare change, we are the smallest and the hardest working. We’re a group of old, disabled people, except for one writer in his mid-50s. The rest of us are in our sixties and seventies, and this is a labor of love. All we’re asking for is the chance to keep telling the truth about Trump and help ensure democracy survives. If you can help, please do. Thank you. Ursula
The first thought in our household was “Skynet incoming”! AI facial recognition is horrible. It doesn’t work. Well, at least last we knew….
Apparently it works better for white people.
So that fits.