Hacker News new | past | comments | ask | show | jobs | submit
antirez | on: Root shell on Samsung TV
"Why you bother about local AI? LOL with the $ you spend for <insert hardware here> you can buy 5 years of 20x account". Yeah. 5 years where you are told what you can do.
gessha | parent | next
It’s quite something when western companies act more like authoritarian governments than actual authoritarian governments. I just wish RAM was less expensive ;(
mrheosuper | root | parent
>companies act more like authoritarian governments than actual authoritarian governments.

Yeah definitely Cyberpunk moment here.

pixl97 | parent | next
My guess is we're within a year of the first governments banning local LLM AI at any scale. Declare GPU's as munitions that must be registered and ban any card with more than X GB of RAM and more power than a 3060 or something like that.

That could be done with an executive order in the US, and it probably wouldn't take much to get it to pass congress.

Now, don't worry, businesses can still buy a lot more GPU compute than that, but they have to register and track their usage just like any other company making/buying weapons.

trollbridge | root | parent
There is zero chance this will happen.
mrheosuper | root | parent | next
By banning the Chinese from accessing the latest and fastest GPU, this is somewhat happening now.
handoflixue | root | parent
You only need to ban compute on a scale that can actually compete with frontier models, which is currently something like a $500,000 investment.

"Must declare any purchase of 10 or more GPUs" does not strike me as particularly implausible.

There's already plenty of jurisdictions where possession of certain tools is considered evidence of criminal intent.

I don't think any of this is good behavior on the part of our legal system, but I do not doubt the legal system's ability to regulate this if it so chooses - and Anthropic has already gotten slapped down by the Trump administration once before.

trollbridge | parent | next
I use local abliterated models + paid for subs like OAI or A\ and avoid touching anything that would hit their TOS + anonymously paid for Chinese pay as you go for things in between those two.

Works very well.

neoCrimeLabs | parent | next
5 years, assuming the terms don't change.
ThrowawayTestr | parent
5 years where you can get arrested if the AI tattles to the feds.
pixl97 | root | parent
Just wait till your GPU legally requires an online connection to run an LLM.