​Anthropic CEO Dario Amodei recently called on AI labs to slow how quickly their models improve. OpenAI's Sam Altman agreed within hours. But with how fast AI is developing, the stakes of one lab slowing down (or not) are higher than ever.

​This session's resolution:

​Leading AI labs should unilaterally pace the frontier.

​For the debate, "pace" will mean slowing how fast AI capabilities improve, without stopping development.

​If you want:

• ​to prepare for that upcoming interview

• ​to hold your own in your next coffee chat

• ​an excuse to learn more while you connect with others

​Then this is your opportunity.

​How this works

​Each debate club event is centered on an AI safety topic.
We expect participants to do some research and prepare, at most, a few talking points for both sides.

​At the event, you'll be paired with another participant and randomly assigned a position, in favor or in opposition.
Each side makes an opening statement, then a rebuttal answering the other's arguments. Finally, partners exchange feedback.

​We'll send out a short reading list before the event.

​Who is this for?

​People at all levels of AI safety fluency are welcome and there is no debate experience required.

​In the spirit of ACX meetups, "Please feel free to come even if you feel awkward about it…"

​What's the event format?

• ​Arrive and mingle: 15 minutes

• ​Welcome and format overview: 10 minutes

• ​One-on-one debates:

  • ​Prep: 3 minutes

  • ​Pro opening: 4 minutes

  • ​Con opening: 4 minutes

  • ​Pro rebuttal: 4 minutes

  • ​Con rebuttal: 4 minutes

  • ​Peer feedback: 8 minutes

• ​Either a second debate round or open group discussion: 20 minutes

• ​Group debrief: 15 minutes