RSVPlease (click on the buttons above (no account or email is needed) to let us know you'll be there)

The event will take place at the Haus des Engagements, on the second floor of Große Bergstraße 3-5 (Neues Amt Altona). We'll update this post with details on how to get in shortly before the event.

About the event:

Imagine: A frontier AI lab is preparing to release a more capable model. The release matters commercially. Safety test results are concerning, but ambiguous. Investors want growth. Competitors may be close. Product teams want a launch date. Policymakers want answers. And the news is looking for drama.

Nobody has to be evil for safety to lose.

The question is: when pressure to launch is high, what would actually make you delay deployment?

This session is a gamified exploration of AI safety incentives and alignment. We will use an AI lab scenario and you will be dealt cards defining the setup. What would you do and what can we learn about the strategies that change the outcomes toward safety?

This is hosted by Effective Altruism Hamburg, a group focused on using evidence and reason to figure out how to do the most good. New faces are always welcome!