COO Successif, Trustee Effective Ventures UK, and member of the Talos Network board
Co-Director EA Germany 2023⁄24, entrepreneur for over 25 years, member of the EA Munich organizer team since 2020.
COO Successif, Trustee Effective Ventures UK, and member of the Talos Network board
Co-Director EA Germany 2023⁄24, entrepreneur for over 25 years, member of the EA Munich organizer team since 2020.
As an employer, I would not want to rely on an employee taking a below-market salary. Otherwise, I might be incentivised to keep someone on the team even if they are underperforming, undermining the work of other team members. I would want to hire for talent first and leave salary discussion for the last step, to avoid bias.
That being said, there might be good cases for early-stage orgs that might otherwise not be able to hire, or for positions that might open because of the low salary requirements. At Successif, we recommend that job-seekers use informational interviews with potential employers to explore these kinds of new roles.
My short Claude prompt was only intended as a conversation starter, so I’m happy this worked. I’m not considering investing, but if potential investors would like to carry this on and share here, this might be useful.
I agree and would encourage potential investors to take into consideration base rates of startups reaching €1M+ on profits yearly when comparing this to other forms of investments. I spent 5 min prompting Claude to come up with a BOTEC based on this post, which I haven’t checked but could be an entry point to additional research.
The board members of EA Germany are elected by the members (over 100) of the organization. The board is responsible for hiring the director.
What you write aligns with the challenges we see from our advisees, and based on your profile, you may be a good fit for our career advising program if you are open to working on AI Risk reduction.
For people we have helped in our program, we typically see transition timelines of 6-18 months, but just today I talked with someone for whom it took two years. My colleague Moneer wrote about his experiences in getting into a position, which included taking 170 actions (like applications, 1-1s, projects). This can seem like a lot, but it comes down to 3 per week on average over a year.
You don’t have to have an academic background to succeed (I never went to university myself), but regardless of your qualifications, be prepared for it to take time to find a position. In our advising, we emphasize the importance of building networks, conducting informational interviews, and getting more information on how to position oneself as a promising candidate.
That being said, you might be able to have a higher counterfactual impact if you find a position that is not in one of the well-known orgs—I would keep that in mind.
Talent pipelines. There’s no fellowship to train people to go directly into advocacy for AIS, compared to over 10 such efforts aimed at research. We’re training AIS researchers by the 100s, and leaving advocates to figure it out for themselves.
I’m not sure if this is different from what you meant, but we ran the first iteration of our AI Safety Advocacy Fellowship this year with promising results.
It seems 1- 1s are done remotely, which means this could also be done by an international organization. I assume this would allow it to be more cost-effective as you would only need one organization, director, CRM, and training for several people who could do 1-1s for several regions.
Thank you for asking this question! I have the feeling that for some national groups, we might be upholding them based on path dependence, not because they have intentionally selected the right target group. I wrote a recent comment about this, based on my experience at EA Germany.
I’m most excited about national organisations that can reach specific, narrow target groups, as many of the scalable programs would seem more efficient to do on a larger scale.
That being said, a larger organization operating at the continental or international level could still hire contractors to experiment with smaller interventions. This would mean having only one organization with one director, potentially increasing cost-effectiveness while allowing for experimentation with different target groups and markets.
A director of EA Europe could hire a team to organize a conference in the North of the UK, for example, while having an Italian-speaking contractor doing 1- 1s for Italy and organizing group calls for European CBs. As a UK CB, you take away these possibilities as you artificially narrow the focus without much reason.
I would be excited for EA UK to think more broadly in scope, connect with other European national groups, and expand the parameters within which the director would be allowed to operate.
So exciting to see that you are doing the first EAGx in South America (if I’m not mistaken)!
I am sorry that you have to deel with kind of frustration. At Successif we have done double blind work tests for the first round of the hiring interviews for advisor applications and had good results.
On the other hand we recently hired for an operation associate and didn’t do this. I’ve become less excited about written work tests given LLM capabilities and I think work tests can cover only parts of what I’m looking for. I’d rather have people describe what they did before in their application so I can talk with them about the details in the interview.
In this case we hired someone with several years of experience doing the kinds of things we were looking for in a another organization without having been involved in EA. I don’t know how representative that is, but I do nudge organizations to hire for experience before alignment when asked.
I can’t comment on your situation as there is not enough information but if people ask me about operations roles I typically recommend to upskill outside the impact space in organizations which can provide mentorship and good operational procedures—something which is sadly sometimes missing in our space.
In terms of counterfactuals I would suspect that we are sometimes seeing negative values by only hiring within the community and often very small ones for many of the competitive operation positions. For our recent hiring round we received 70 applications in four days before closing the applications. We had many brilliant and purpose driven people apply who were affected by USAID cuts.
So what could you do in addition to upskilling? Volunteer roles can be a great way to get into roles which will never be publicly announced. Similarly networking, doing small scale side projects, being around people in organizations and helping them (for example helping the local or national community) can help to stand out.
Sometimes it can also be the most impactful thing to have a solid career, donate a be the person to help others in the community. Imagine a strong hub of people in Rome* doing impactful work together because you were there to consistently support them and fill the gaps in the local space. This could be awesome.
*I don’t know anything about EA in Rome so only speculating
I can see democratic models providing value, but the practical implementation is tricky. I can only speak from my experience in EA Germany, where member engagement in national-level strategy and participation in the national community seemed much lower than what I experience on the international level (in this forum, for example) or even at the city level at times.
I would be more excited about either local structures (cities or small regions with fewer than 10 million people) or larger structures (sub-continents, professional groups, etc.) where people truly form a community in the sense that they see each other in person, or where there is a large enough body to allow for meaningful participation in democratic processes.
Thank you for offering this! Are you collaborating with Tlön?
My guess would be that the main pros of having democratic deliberation doesn’t come from when the going is normal but rather as a resillience mechanism?
Perhaps, but I can also imagine that a hand-selected nonprofit board may be able to spot risks and react to them better than a board voted in an assembly. The coordination function of an assembly in trying to fill specific board roles seems lower than if a smaller group of existing board members can discuss it.
Picture it. The year is 2035 (9 years after the RSI near-miss event triggered the first Great Revolt). You ride your bitchin’ electric scooter to the EA-adjacent community center where you and your friends co-work on a local voter awareness campaign, startup idea, or just a fun painting or whatever. An intentional community.
We run something similar in Munich, where we have a coworking space that also hosts EA-adjacent events (including crafting events), located in the middle of the city, allowing people to bike there. So, very sympathetic to the idea of having local groups doing this.
I agree with the benefits of local community structures. However, I don’t believe that national EA groups can offer as much as informal local groups. I help manage both formal and informal networks of EA (adjacent) individuals in Munich, and there, I see these points much more clearly. Running a coworking space, hosting in-person events, convening private meetings, and having one-on-ones seem like activities that would fit your list.
Thank you for writing this up!
Some critics worry that democracy might impede nimble decision-making or divert energy from high-impact goals. Yet EA Norway’s record — attracting steady funding and successfully supporting members with their careers and donations — suggests otherwise.
I’m somewhat confused about what led you to this conclusion. I was the co-director of EA Germany for two years, an organization that is similarly structured. When I compare it to the memberless nonprofits where I’m a board member, the overhead for organizing a general assembly has been greater, yet it hasn’t resulted in significant decision-making input from the members.
Having fee-paying members suits an organization that benefits its members. At EA Germany, the target group for the interventions wasn’t the members, but rather people in earlier stages of the talent pipeline. If I want to contribute to talent pipeline development, I would prefer to donate to the charity I consider most cost-effective. It is unlikely that this would be the national EA group, given the numerous players in this space. Therefore, I would personally hesitate to join a national group that requires fees, unless tax reasons or special insider knowledge lead me to believe this is the best use of my donations.
Overall, I worry that national membership groups in EA lead people to make decisions that are not solely motivated by EA principles. My theory is that the main activities currently undertaken by national EA organizations could be carried out more cost-effectively by fewer players with a broader geographic reach. I fear that membership organizations are not the best structures to critically evaluate their existence and shut down if they believe members’ time and money could be better spent elsewhere.
First, ageism. Lots of young people are simply biased against older people—assuming that we’re closed-minded, incapable of learning, ornery, hard to collaborate with, etc. I’ve encountered this often in EA.
I’m not sure what age group you’re referring to, but as someone who just turned 50, I can’t relate. I did have to upskill not only on subject matter expertise (as mentioned in the post) but also on ways that people of the age group and the community are communicating, but this didn’t seem much different than switching fields. The field emphasizes open-minded truth-seeking, and my experience has shown that people are receptive to my ideas if I am open to theirs.
Second, political bias. In my experience, ‘signaling value-alignment’ in EA organizations and AI safety groups isn’t just a matter of showing familiarity with EA and AI concepts, people, strategies, etc. It’s also a matter of signaling left-leaning political values, atheism, globalism, etc—values which have no intrinsic or logical connection to EA or AI safety, but which are simply the water in which younger Millennials and Gen Z swim.
The EA community as a whole is indeed more left-leaning, but I feel that this is less the case in AI safety nonprofits than in other nonprofit fields. It took me some time to realize that my discomfort about being the only person with different views in the room didn’t mean that I was unwelcome. At least I was with people who were more engaged in EA or who were working in this field.
At the same time, organizations that are not aware of their own biases sometimes end up hiring people who are very similar to their founders or are unable to integrate more experienced professionals. This is something to be aware of.
Apart from taking part in programs such as those of Successif and HIP (that have limited slots), I would like to see experienced professionals new to AI Safety team up with young professionals who are more embedded in the community but lack the experience to fundraise for ambitious projects by themselves.
Talking for Successif, we have ramped up our capacity in the last months and are currently admitting a high rate of applicants to our program. I am biased here, but I think our advisors can help individuals think more specifically about how much time to spend on learning what concepts, whether to volunteer or work on projects and when to double down on applying. We’re only focused on helping mid-career and senior professionals get into AI risk, and our advisors usually have multiple calls and email exchanges with advisees over several months to always discuss the best next steps.
I broadly agree with the post, but I know from my own experience that it can be hard to decide when to prioritize upskilling, networking, projects, or applications. Some people in our program struggle with imposter syndrome, which can lead to spending too much time learning concepts when this is not their current bottleneck.
Yesterday, I did 7 one-hour walks with Munich EA community members. Here’s what I learned and why I would recommend it to similarly extroverted community members:
Created an info document and 7 one-hour Calendly slots and promoted them via our WhatsApp group
One hour worked well as a default timeframe − 2 conversations could have been shorter while others could have gone longer
Scheduling more than an hour with someone unfamiliar can feel intimidating, so I’ll keep the 1-hour format
Walked approximately 35km throughout the day and painfully learned that street shoes aren’t suitable—got blisters that could have been prevented with proper hiking boots
Directly invited two women to ensure diversity, resulting in 3⁄7 non-male participants
Noticed that people from timeslots 1 and 3 spontaneously met for their own 1-1 while I was busy with timeslot 2
Will actively encourage more member-initiated connections next time to create a network effect
My prepared document helped skip introductions and jump straight into meaningful discussion
Tried balancing listening vs. talking, succeeding in some conversations while others turned into them asking me more questions
Expanded beyond my usual focus on career advice, offering a broader menu of discussion topics
This approach reached people who initially weren’t interested in career discussions
One participant was genuinely surprised their background might be impactful in ways they hadn’t considered
Another wasn’t initially interested in careers but ended up engaging with the topic after natural conversation flow
2 of 7 people shared personal issues where I focused on empathetic listening and sharing relevant parts of my own experience
The remaining 5 discussions centered primarily on EA concepts and career-related topics
Received positive feedback suggesting participants gained either new information, decision-making help, or expanded horizons
Sent follow-up resource links to most participants after our walks
Scheduled specific follow-up meetings with two participants who wanted to continue our discussions
Deepened community connections, which I view as an investment in motivation for both others and myself
Thoroughly enjoyed the 1-1 interactions on a sunny day, exactly as I had hoped
Found this volunteering format personally rewarding and well-aligned with my strengths
Recommend to other community members who enjoy walking conversations—it’s an accessible way to contribute
I agree and argued in a similar direction in a comment last year.