This one's got a lot to mock.
> "we can't discuss our sooper seekret deception operations"
The ghost of kyuv, the snek island thing, and a bunch of other retarded bullshit that only mentally handicapped redditors and middle-aged childless women were willing to repeat. I think the guy knows he could very well list them, and is just embarrassed. He thinks he has dignity, and wants to keep it.
> "crowdsourcing possible war crimes"
At the start of the war, these retards seemed to think they could apply the same approach to fighting Russia that they did to doxing a construction worker who did not show sufficient respect to an obese black transsexual. It appears they still think this is viable. Maybe because it's all they've got.
https://i.poastcdn.org/7b7e5f1f0e00896ca491537b3ec575808b7e4e44a68a258ae932b53b88892952.png
Also, this idiot is the latest moron to jump on the "what if we fine-tuned an LLM on <insert dataset with no value in LLM fine-tuning here>?" bandwagon. Nothing new, none of these people know how to code, and none of them are smart enough to learn.
The irony is that there's plenty that can be done with machine learning in warfare, especially if you're as morally blasé as the modern U.S. government. They're just too stupid to put talent into autonomous drone swarms instead of chasing some marketing drivel.
https://i.poastcdn.org/5a40e820ad5cb1e69177120905cc057ff6f533c693d06bcd3df6ce2ea69a6e05.png
The remainder of the article is a full page and a half of meaningless drivel meant to sound inspiring. It's like a gayer, less competent version of the USSR.
The author is the modern army, incapsulated. Zero technical knowledge but offering suggestions on how to use multi-billion parameter generative models. Throwbacks to century-old wars because that's the last time the U.S. could win one. Utter refusal to acknowledge that the demographic situation means that everything remotely effective is now off the table.