Comments by "TJ Marx" (@tjmarx) on "AI extinction threat is ‘going mainstream’ says Max Tegmark" video.

You don't need AI to spread misinformation, CH4 does a fantastic job of it every day. China do not limit themselves on biological research in the same way the west does. It's why the west fund bio research in China, because it's legal to do things there that are illegal in the west. That includes things like human cloning and human gene editing. China wouldn't sign up to a framework to limit AI, or if they did they wouldn't respect it domestically. I think that's ok though because the risk from AI isn't from competent general AI being so great. The potential risk comes from incompetent generative AI being linked in with things it shouldn't be. For example, imagine if generative AI was plugged into a popular server security product, maybe even windows defender, and it made a mistake misclassifying say all pdf files as a security threat which it then automatically deleted on it's own. How much damage would that do to society globally? How many trillions in losses? What if AI was plugged in to controlling the state of a valve at a water or gas utility. What if it jammed the valve open or closed when it should have been the opposite? What would be the potential damage there. It's incompetent AI that poses the genuine risk.
2