[dorkwood]

You're right, there's a very real danger that bad actors within OpenAI could hand their research to China. But that's not an inevitability. We've managed to block certain countries from developing nuclear weapons technology. We can do it with this too.

[cooper_ganglia]

I wasn't even thinking about an OpenAI defector releasing secrets. I think it's a matter of research and compute power, both of which China is increasing in these sectors so they don't trail behind the USA.

If we stop developing our tech, China will continue advancing even without leaks from OpenAI, and eventually they will develop better models, even if it takes another decade. Banning something only to cower in fear of ourselves while we watch someone else use it to destroy us seems like poor planning.