I get what you are saying, but its not just data egress, its how the model was trained. I’m thinking of models trained to inject security weaknesses, for example.

I see no reason US based AI companies would not carry the same risk especially with the well documented cases of groups like the NSA openly asking US technology companies to provide back doors.

For sure, that’s why I dont trust any of them. Its interesting because unlike open source, open weight models don’t tell you anything about what was included or excluded.

Yes, rather we trust the capitalists who want to literally remove technology from our control and force us to use their cloud and their AI models that are built on good old American ingenuity and honesty, trust me bro you’re better off looking for weaknesses in our models by spending on tokens to see what happens .

Not everything China does is a psyop.

And spied on to extract personal data we can sell.

Trade secrets, too.

this article doesnt address local models. this is about kimmi k3. i had to look because its very hard to restrict local ran software

kimi k3 is going to release on huggingface tomorrow. it will be open weight, assuming nothing gets between now and then. they just wanted a short exclusivity period.

“Trust us! We’re a bunch of deranged genocidal pedophiles who want to put all the poors in death camps! We’re the good guys!”