I've said this before in not so many words but it's worth repeating:
Everytime I ask a serious question of AI, I find errors in the reply.
At this stage in the game, I believe the constant stream of misinformation to be purposeful.
I tend to think that the populations that enjoy AI the most do little critical thinking or honest fact checking.
I could see the case for Western AI (Claude, GPT, Mistral, etc.), especially since the datasets they're trained on are based around Common Crawl... a disgusting dataset designed with unethical garbage in mind.
DeepSeek takes a different approach with their reasoning model, R1, as it's dataset is not only open, but the local model is FLOSS (MIT license). Sure, it may have some restrictions due to Chinese laws, but otherwise, do way better than Western models.
Yeah. There's probably some decent open-source AI out there. Ty.
imput & structure}AIia69r*/*ya but scuse 4 offtopIC/ICsum new thotz bro, on that music stuff and have an idea for V4V p2p/****U*& mE th@ might facilitatE a trip/o;.;Ore gon port land outlier/sleep pattern disrupted so - NA worries Micah
