OpenAI and the Military Drama

First off, the biggest thing everyone's talking about is how OpenAI just officially signed a deal with the U.S. "Department of War" (yeah, they're calling it that again, apparently) to put their AI models on their classified network. Seriously, like, for military stuff. This is a massive deal. People in the comments are freaking out, wondering if this means AI is going full autonomous warfare. It's at Sam Altman's Twitter.

And linked to that, there was a whole thing about Anthropic, another big AI company, getting called a "supply chain risk" by the Secretary of War. Anthropic put out a statement basically saying, "whoa, we're a public benefit corporation, we have moral redlines, we don't want our AI used for mass surveillance or autonomous weapons." But then people were debating if being a "public benefit corporation" actually changes much when investors want their money back. Total drama.

Naturally, because of all this, a lot of people were looking up "How to delete your OpenAI account". The link to their help page was super popular. Comments were speculating if it was because OpenAI's CEO, Brockman, donated to a pro-Trump PAC, or the Pentagon talks. People are definitely getting a bit spooked by the military connections, and some are talking about just running local AI models instead.

Oh, and speaking of OpenAI drama, there was a post called "The whole thing was a scam" by Gary Marcus, a known AI critic. He's basically saying all this OpenAI stuff, especially the military deals and the ethical compromises, proves his point that it was all a bit shady from the start. People in the comments were agreeing, saying the "anti-Altman rhetoric" might have been right all along. Check it out: Gary Marcus's Substack.

Big Geopolitical News

Okay, switching gears, but also super intense: the U.S. and Israel launched a major attack on Iran. This is HUGE. The comments section on CNN's report was absolutely blowing up, with over 2400 comments. People are worried about nuclear escalation, supply chain collapse, and some were even talking about how Iran's top diplomat said negotiations were actually going well before this. Crazy timing, huh?

And then, related to that, there's news that Iran's Ayatollah Ali Khamenei was killed in one of those Israeli strikes. This is massive, after 36 years of his rule. The comments on NPR's story were full of historical context, like how the US overthrew a secular government in Iran back in '53, which led to all this. People are debating if this will actually bring democracy or just more chaos.

Other Interesting Tech Stuff

On a slightly lighter note, but still about big tech, Google Gemini had to "Address Antigravity Bans". Apparently, some people got banned for using "antigravity" features, whatever those are, and Google had to reinstate access and change their TOS retroactively. One comment mentioned how they ban people for things that weren't even prohibited when they started! Here's the discussion on GitHub.

And hey, if you're into local AI, Qwen3.5 122B and 35B models are supposedly offering "Sonnet 4.5 performance on local computers". That's pretty wild if you can get that kind of power on your own machine! People in the comments were sharing their setups, like using an AMD AI max+ with 128GB RAM to run a 120B model. Sounds like the local LLM scene is still heating up. Check out the VentureBeat article.

Okay, gotta run, but thought you'd want to hear about all that! Talk soon!