AGI Preparedness Workshop (1st)

​To sign up, register your interest here. The 2nd session will be decided based on your response to the above form.

Will MacAskill, who is coming to Cambridge on the 30th of October, is known for his work on cause prioritisation and longtermism and as being a foundational figure in the EA movement. More recently he has been working on AGI preparedness: major AGI-related challenges that are not directly related to misalignment. In a recent Forum Post MacAskill points out that these challenges are currently neglected by both the EA and AI safety communities, and that university groups in particular should consider prioritising them more.

During this two-part workshop we will read and discuss existing research on non-misalignment AGI challenges (drawing heavily from research by Forethought). Topics will include:

• ​The “intelligence explosion”

• ​Making the long-term future go well

• ​Gradual Disempowerment

• ​Extreme power concentration

​We will then spend some time brainstorming if, and how, we should promote working on these topics at Meridian. The workshops are scheduled to be before MacAskill’s visit so we can make the most out of the Q&A when he visits.