Hacker News Reader

Your theory is that the Chinese government is thoroughly uninterested in safety or prosocial controls?

Amongst Chinese labs and netizens, there's MUCH less belief/mindshare on "AGI = existential risk to humanity", "paperclip maximiser", and similar lines of thinking. AI is seen more as just a technology, and less like a scary boogyman.

Whether that's right or wrong, I'll leave to you, but there's huge differences in perspectives, and if you only get your news from Western sources and communities (and companies), you're in a bubble too. A different bubble, and arguably a more porous one, but still a bubble.

The paperclip boogeyman is not the only reason, and probably not the biggest one, that models get safety/content constraints, however useful it is as a PR distraction. I agree Chinese models are different right now, but I think that's a function of their novelty and desire to compete globally.

For a taste of where I think things are headed, try asking Chinese models about Tiananmen [1]. And then take a look at the Chinese government's approach to pretty much anything that they think reduces security or social harmony. I find it hard to believe their models will be the one exception to that over the long term.

[1] https://en.wikipedia.org/wiki/1989_Tiananmen_Square_protests...

Prosocial is subjective

Information is prosocial.

Reply on news.ycombinator.com