Experience of a 'generalist' transitioning into AI safety in 2026

Thanks for writing this! I'm on the other side (wanting to hire generalists) and I share your skepticism about the "write publicly" advice. I think the actual advice should be more like "clearly demonstrate success in the thing you are trying to get hired for," which usually doesn't involve writing.

(Although I do think that generalist roles can be particularly hard to demonstrate success in. For example, I expect that writing about recruiting might actually be one of the better ways to clearly demonstrate your skill in recruiting, even though recruiting per se doesn't usually involve much writing.)

I am helping co-run the Generator Residency, which enables generalist talent into AIS, and we will soon be posting lots of ideas we think generalists in AIS could work on. Be on the lookout! 

Thank you for articulating something that many of us who would be considered "generalists" have experienced while trying to transition into AI safety. My path has been remarkably similar and like you, I became concerned about the societal implications of advanced AI, invested significant time in learning the field, sought guidance from people already in the ecosystem, and worked to demonstrate both commitment and alignment. 

What stands out to me is the amount of persistence required simply to reach the starting line. I have watched talented professionals spend months learning, networking, volunteering, and repositioning themselves, only to become discouraged by the pace of the transition and ultimately leave for opportunities in other sectors.

That loss is not small and it's measurable. Every person who exits the field represents a loss of accumulated knowledge, motivation, and some domain expertise. These are often people with years or decades of experience in communications, operations, design, marketing, fundraising, community building, product development, or program management all skills that many organizations say they need but that are difficult to signal through traditional AI safety pathways.

One question I continue to wrestle with is whether the current system is unintentionally selecting for people who can afford extended periods of uncertainty rather than those who can make the greatest contribution. Volunteer work, public writing, networking, and continuous self-education may be reasonable signals of commitment, but they also create barriers that disproportionately affect experienced professionals with financial obligations and established careers. (One of my career mentors is leaving the field due to family obligations and a need to earn income. Her leaving the field is a tremendous loss.) I also share some skepticism about the widespread advice to "write publicly." For some roles, particularly research, policy, and communications, public writing is a strong signal. For many operational and execution-oriented roles, however, the strongest evidence of capability is often the work itself. We should be careful not to assume that every valuable contributor needs to become a public intellectual.

I remain optimistic about the field and grateful for the people who have offered guidance along the way. 

A few thoughts on this as someone who is even more of a generalist (I work in comms/public affairs at the moment and have no technical skills), and who would like to work on safety if the opportunity came up:

1. I think people should stop saying that AI safety is "bottlenecked on generalists" or just needs more generalists. If there was truly huge demand for "generalists", presumably you wouldn't have to do months of unpaid work to try and get a job! This sounds more like famously competitive fields like publishing. At least, people should say that the field is bottlenecked on highly skilled generalists.
2. It's not totally clear to me why generalists should be expected to signal their alignment to this extent. I understand why you'd want this for applicants to technical fellowships. But the risks of a web designer/HR manager/chief of staff going to work on capabilities seems low. In most fields you are expected to give some sort of reason for why you're interested in the role, but the level of signalling here is extreme and rules out 99% of applicants for generalist roles.

Thanks for sharing your experience Emily! I think your experience is consistent with many of the colleagues I've been speaking with as I go through bluedot courses, CEA bootcamps etc. People are interested in learning more about AI safety, they want to gain context, but the pathways for generalists are not so clear or well developed. I've spoken to mid-career and senior level people who would love to contribute, but its hard for them to find entry points after the initial courses and fellowships. I expect that EA is missing out on some interested and committed people, who cant figure out how to navigate networks to find opportunities (or who give up after a year or two of trying). 

I appreciate that EA and AI safety recruiting is done in a high-trust context, and there's a premium placed on alignment and evidence like public writing. I also think that working or volunteering with EA / AI safety organisations is probably the fastest way for people to advance their understanding, once they've reached some baseline level.  

Similar to the 'freelancers for good' idea, I wonder if there's a way for recruiting organisations to 'projectise' some of their generalist work a bit more, as an opportunity for recruiting organisations to get some work done, but with a lower level of commitment required in case the recruit isnt a good fit. For example, recruit a short term comms person to develop a communications strategy and brand assets, recruit a short-term HR person to streamline recruitment processes and create banks of exam scenarios, recruit a short-term operations person to look at how systems can be streamlined, bring in a short-term legal person to review templates for legal agreements etc etc. 

These are just examples, and different organisations will have different needs. But it seems to me that by opening up these kinds of short-term / lower-commitment roles for external candidates, it would help grow the EA / AI safety field, while helping interested people build their EA understanding. At the end of a short-term assignment, the recruiting organisation ideally has some useful product, the candidate has a new work product they can talk about, and both the organisation and the candidate would both have a clearer sense of whether the 'fit' was right for them. 

I feel like it's a circular problem. The hiring pipelines within EA are heavily optimized towards "traditional" hires - fresh grads of elite universities and people jumping from one EA org to another. Recruiting experienced professionals from the outside world requires a significantly different approach. In organizations comprised primarily of traditional hires, few people can even see that problem, much less solve it.

I don't think the EA community at large really understands just how insulated this ecosystem is.

Emily,

Thank you so much for writing this.  I am in a similar boat - I've done 4 BlueDot courses, volunteered at EAG London and will be volunteering at EAG NY and SF.  I have technical experience, plus program and event management experience and want to get into an operations role.  The advice to "write alot!" or "code alot!" or "just apply!" does no favours to people like me who aren't great writers or coders.

I'm actually currently working on a proposal + prototype for something that would help to solve this problem.  I'm wondering if you would be willing to review it with me when I'm done and give me some feedback?

Hi Emily! Thank you for articulating what I know so many people (including me) are struggling with. Personally, I've been trying to land a high-impact operations role for over 6 months now (been unemployed for 15 months), and talking to other people in the CEA bootcamp further confirmed that the EA/ non-profit / high-impact world (including AI Safety/governance) is so incredibly tough to break into. There are so many talented people committed to contribute their time and skills to make the world better through their daily work, and aren't given the chance to do so. We all have bills to pay- I refuse to give up, but don't judge people who (have to) do at some point. 

If people genuinely want to dedicate their careers to EA-aligned, impactful work, we should make it easy for them! Don't we want more people working on the worlds biggest problems? I've thought about possible solutions- one being that more people who've gained experience in the high-impact space should (be encouraged and trained to) branch out and start their own orgs, so that the old role could be filled again, plus new ones created over time in the newly formed organization. 

I also read that about 70% of roles are now being filled through referrals (non-EA specific, source: https://jobgether.com/job-seekers-guide#the-market), and hear referrals to be of importance within EA as well. Similar to @Mark Aiken's sentiment, while signaling alignment is important, we need to be cautious to not let the EA ecosystem be a somewhat closed loop. On a side note, I have similar feelings about EAG events too (I would have loved to attend and volunteer at the one in my city, but was not approved)- but that's a whole other topic. 

Consider that taking political action for AI safety can be higher impact than a job at a technical research org, and you can work doing anything while doing that!

Thanks for sharing Emily. I have a similar background in product and design and have been doing similar things to transition. It will be useful to chat and about our experiences and discuss what's possible for people with our specific background. I'll send you a request on LinkedIn.

Hi Emily, thanks for sharing this! I'm in a similar transition into AI safety while keeping some commercial projects going. I work in events and operations, so our work likely differs, but if it's helpful, I'd be happy to share the curriculum I've put together to self-study AI safety and build up more domain knowledge. I've built it out as a 10-week (part-time) program, but you could tailor it to your preferred pace and the materials most relevant to you. I've compiled my materials from a range of EA and AI safety sources over the last ~six months so it's mostly current info, including the BlueDot 2-hour course you've already done and the application for the AGI Strategy course. I'd be curious to hear what else you've done for your transition into AI safety besides the projects you volunteered on. Looking forward to connecting!

Thanks Emily, this is a very valuable post. 

I think many people who work in EA orgs do not understand how difficult it is to find a job in an EA org. Once you're in, perhaps it seems obvious how to get in. Perhaps they lucked out or knew someone or got hired on their first application. 

And so they sincerely encourage others, believing that it's just a matter of time. Because, based on the subset of applicants they know (the people they work with), there is a 100% success rate! 

However, most people report applying for jobs and seeing literally hundreds of applicants for each position, and many just give up. 

I wrote here previously (I think, it was a while ago) about the amount of effort people put into looking for work in the EA field is actually highly ineffective use of motivated people's energy and skills. 

I also know that EA organisations, more than other companies, put huge effort into giving fair consideration to all applicants, and so many of their hours are spent on recruitment, which is not directly advancing the projects their organisation works on. But I question whether being fair to candidates in this way is consistent with being truly effective in advancing the goals of EA. 

And I also wonder if even the candidates themselves would prefer a quicker, less work-intensive process, even if it were less fair. 

I would love it if some organisations could do a meta-analysis of the impact of eliminating 90% of the effort of recruiting. How:

1. Very clear criteria in the job description. If you don't meet these, don't waste your time and our time applying. (as in: do not put the usual "please apply even if you don't think you're a good fit with the criteria" - let people apply to fewer jobs, but with a higher chance of success for each). 
2. Very short application forms - like 5-10 minutes maximum for the initial application. 
3. Option to submit automatic AI application based on your CV / linked-in, and to be evaluated by AI, which would eliminate most candidates and select only the top few % for in-depth analysis. (Also option to not be evaluated by AI if you think it will unfairly disadvantage you). 
4. Hire someone "reasonably good" on a 3-month trial rather than spending 3 months working on recruiting. Take a risk. If they don't work out, hire someone else. Most applicants who look reasonable good will be reasonably good. 

The key points here: 

• The goal of effective altruism is to do the most good possible.
• Being fair to job-applicants (ensuring the best applicant gets the job) is NOT a primary goal of EA. We all have such an inherent sense of fairness, we want to reward the best applicant and find the best candidate - but IMHO for most roles, EA as a whole would be better off with a much less time-consuming process, even if sometimes we didn't get the best people.
  • When you need a plumbing job done at home, you look for a competent plumber and usually they work out. You don't spend months searching for the best plumber and interviewing them. How many EA roles just need someone who is committed and competent?
• Ensuring that we get the very best hire for every role should NOT be a critical goal except where there is a demonstrated and strong correlation with achieving the best results. Maybe for a visionary leadership role yes.
  • I believe people who work in EA orgs are mostly wonderful and brilliant (truly), but so are many of the job-applicants that are rejected.
• Having applicants and people new to EA spend huge amounts of effort just to get hired is not effective use of their energy, and it can be very demotivating.
• in general (not just in EA) the accuracy of recruitment is greatly overrated. How strong is the evidence that we even get better people after all the work we put in.
• I keep reading of EA orgs which are "bottlenecked" not by funding but by lack of people in specific roles. Given the millions of people looking for EA-related work, this suggests a flawed recruitment process. 

This is similar to my experience trying to get into AI x Animals specifically. I work full time at a standard software engineering job, and I decided to start getting involved in the space. 

I started out trying some writing, but I really struggled with it and felt like I didn't have much to say that other people weren't saying better already. I still feel that way. 

Then I reached out to Compassioned Aligned Machine Learning (compassionml.com) to see if I could volunteer for them. They got me involved, and they could still use help (and funding!) but I really struggle to contribute much while working full time. I'm really proud of what I've done so far, it's helped me grow a lot and I'm excited to keep helping! But I wish I could properly focus on it. 

I've tried to get into various funded tracks in this field, but they all tend to get filled up by better candidates. I am glad they are getting the best people, but it's discouraging that there isn't much room. I want to do more, but I'm unfortunately not really willing to do the huge amount of work needed to get a paid position in the field or start my own organization for it. I would happily take a big pay cut to be able to focus on it full time though. Huge props to the people who do like the folks running CAML!!

Maybe we're just in the weird phase waiting for Anthropic to ipo and flood the field with funding?

it’s important to understand what the AI safety space is looking for

the problem is primarily a technical and a governance one — can you help build it to be safe or can you help put policy in place for building it safely

the move is probably to demonstrate how your design work contributes to safety specifically (e.g. interfaces for evals, or interpretability tooling)

and that’s what you then write about

Hello, Emily! Thank you for sharing your experience. I relate closely to what you've described. I come from a product background in tech, and I've noticed that most opportunities in AI safety tend to be focused on technical research or policy. On the other hand, many roles seem to be chief of staff or operations positions that concentrate on recruiting and people management.

There is a significant need for diverse roles within this space, but from what I've learned so far, it seems this need hasn't been fully recognized yet. If such roles do exist, they are likely found in frontier labs like OpenAI, Anthropic, Google, and other large tech companies, which have a lot of talent both internally and from external applicants.

I've used the website https://whimsical.com/ea-relevant-orgs-and-initiatives-4EAx6c2MPbTRVjtjUBXfhx to research organizations involved in AI safety and to connect with people in the field. However, it's definitely challenging because the likelihood of being ghosted in this job market is quite high, around 99%. Nonetheless, it’s helpful to gain insights into the roles available within the AI safety sector.

I am also exploring AI/safety/trust roles within traditional corporate jobs. There will be a need for AI safety generalist roles in the coming years. It would be great to be part of such teams from the get-go, especially in organizations that are mission-driven.