They fine-tuned a Llama 13B LLM with military specific data, and claim it works as well as GPT-4 for those tasks.
Not sure why they wouldn’t use a more capable model like 405B though.
Something about this smells to me. Maybe a way to stimulate defense spending around AI?
You must log in or # to comment.
works as well as GPT-4
that’s a pretty vicious self-own
Looking past all the red scare/ai bullshit, it’s probably a nothingburger. Researchers funded by the Chinese equivalent of DARPA doing something they thought would be cool.
Incoming meltdown and export restrictions on transformer models?
Seems like something the US would hype up into a new red scare tool. So many incentives line up here, I could see it happening, no matter how stupid.
