I am actually currently based in Zurich and work from its EA and AI Safety co-working space PEAKS, which has around 15 desks, and is quite lively. People from the community are always welcome to join!
Any suggestions? My current considerations are Amsterdam, Zurich (if I can get in), Berlin & London. I just don’t know which one of these would be easiest to get a job in/most money left over to give/most EAs.
Trying to answer your core question here: I believe all four of those cities are good candidates and have active EA communities. London certainly has the highest absolute number of AI Safety people, and could be safely considered the biggest AI Safety hub in Europe. Berlin currently mainly has an EA co-working space, but as of this Monday, people are also starting to coordinate more on AI Safety co-working. @James Herbert has already shared a lot regarding Amsterdam, and I am happy to share more about Zurich below.
With an EU passport, you can definitely get into Zurich as long as you have some sort of financial stability (could be a remote job). Both EA communities in Zurich and Amsterdam are very active. Zurich also has a very active AI Safety community. During university semesters, there are one or two events per week on AI Safety in Zurich (although for various experience levels), and many people in the co-working space are working on AI Safety.
I am very happy to share my insights on EA and AI Safety Hubs in Europe. Feel free to DM me if you would like to have a chat!
I agree. One more point to highlight in this context is that the type of retreats referenced in this post, namely Group Support-Funded (GSF) Retreats, have a median cost per participant that is very different from the cost per person as evaluated in Ollie’s post. A rough calculation of median costs based on data provided in that same post on GSF retreats presents:
For University EA group retreats: Median attendees = 15, Median funding = $2061 per retreat. This results in a median cost per attendee of $137.
For City & national EA group retreats: Median attendees = 29, Median funding = $6202 per retreat. This results in a median cost per attendee of $214.
It is important to mention that some of these retreats were partly self-funded, but it seems that the relevant financial support from the EA ecosystem is at least in this approximation much smaller than the ~$1,573 figure used in Ollie’s post. The same post on GSF retreats further mentions that “compared to EAGx costs from 2022-2023, the GSF budget per attendee is between 16% and 48% of the EAGx cost per person.”
It therefore seems likely, that the retreats analysed by Ollie are a different kind of retreat or that these retreats have already gotten much less expensive and therefore potentially more cost-effective (assuming that the quality of outcomes didn’t equally drop by a factor of ~7-11).