We can build it. The question is whether we should — and how fast. Leading AI researchers have called for a moratorium on training systems more powerful than GPT-4, warning of existential risk. Others argue that pausing is impossible to enforce and would simply hand the lead to less cautious actors.
Artificial Intelligence · Real arguments from SuperDebate members below
We don't allow pharmaceutical companies to sell untested drugs. We don't let airlines fly untested planes. Yet we release increasingly powerful AI systems with no safety testing framework, no regulatory approval, and no mechanism to recall them if they cause harm.
Posted by jconnor
The race between AI labs to build the most powerful model incentivizes cutting corners on safety. A coordinated pause would break this destructive dynamic and give researchers time to develop alignment techniques before systems become too powerful to control.
Posted by michael
No con argument has been posted yet. Be the first to push back.
Debate this exact resolution against the AI coach and get scored feedback.
Practice with the coachWe don't allow pharmaceutical companies to sell untested drugs. We don't let airlines fly untested planes. Yet we release increasingly powerful AI systems with no safety testing framework, no regulatory approval, and no mechanism to recall them if they cause harm. (Argued by jconnor on SuperDebate.)
Yes. This topic has been debated live on SuperDebate once, and the replay is public. You can watch the debate and vote on the winner.
On SuperDebate. Post a written argument on the topic page, rehearse the resolution against the AI debate coach, or take it to a live debate night at a club near you. Joining is free.
