What FHI’s Research Scholars Programme is like: views from scholars
FHI’s Research Scholars Programme (RSP) has now been running for just under two years, and we’re excited to have launched applications for a third cohort of research scholars.
In this post, I (RSP’s project manager) want to share some scholars’ responses to a series of prompts about RSP, in their own words. I’ve removed some prompts where there weren’t many responses or I didn’t think they’d be very helpful, and sometimes lightly edited the responses for clarity.
Each scholar’s experience of RSP is different, and this sample (~11 out of 19 scholars) probably isn’t representative; but I hope it will give an illustrative idea of what RSP can be like to those who are interested in the programme or considering applying.
Note that RSP is still a young programme, and we continue to make changes (that we hope are on-average improvements), so future experiences may differ.
Why did you want to do RSP in the first place?
“I wanted to figure out my career plans, work on my health, and see what kinds of research I was good at.”
“After having led the research wing of an EA organization, I wanted to transition to doing research myself, but wasn’t sure which area to focus on. I was hoping to explore different directions to make a more informed decision.”
“I had a fairly specific vision of what I wanted to work on, and thought I could teach myself the important necessary things from online resources and books. I liked that RSP seemed to provide a good environment for independent research and study.”
“I really liked the program description, and a large amount of research freedom and space to explore.”
What are you working on at the moment?
“I’m trying to clarify and evaluate claims that distributions of opportunities for altruistic impact are often heavy-tailed – i.e. roughly that the total impact from many different activities will be dominated by the few highest-impact ones.”
“I’m working on a few projects at the moment. One is on investigating potential effects from transformative narrow AI, one is on investigating atomically precise manufacturing, and one is on mapping out potential paths to transformative AI.”
“I’m working on several projects at the moment, one of which is on the effects current and advanced AI systems might have on human autonomy. In another project, I analyse the opportunities and challenges that might arise from introducing broader impact statement requirements for ML conference submissions.”
How do you spend most of your time?
“Most of my time is spent on the research projects that I am working on. In practice, this means that I spend a fair bit of time doing solo research and writing either at Oxford’s libraries or in the FHI office. I also spend time meeting with collaborators, other researchers, or people doing work that is relevant to my research. There is an active community of researchers willing to engage on new projects, so I find there is no shortage of people interested in workshopping ideas or collaborating.”
“It depends on the time of the academic year. I spend the majority of my time on research but there are times where other things, such as teaching, presenting, or consulting take up a whole bunch of my time.”
“This was actually quite varied during the program, where in some quarters I was every few weeks in the US, attending conferences and workshops and meeting many others working in the field, other times I was mainly walking in Christchurch meadows, thinking, drawing at whiteboards, and not talking to too many people”
What problems have you faced on RSP?
“[T]he freedom to largely do what you like is a strength, but the relative lack of structure can also be a challenge. Although the RSP leadership provides some support, to some extent it’s up to you to create the structure that works for you, in your work day, say.”
“I think I underestimated how challenging I’d find the lack of structure and need to ‘be my own manager’. I thought: “I have done self-directed research with barely any supervision before”, thinking e.g. of my master’s thesis in mathematics. However, in hindsight I didn’t sufficiently appreciate differences between RSP and my previous experiences that turned out to matter a lot for me psychologically: in maths I had accumulated years of experience and a community of peers that were helpful for me assessing my own work and choosing between research directions; while finding original insights was hard, it was at least possible to unambiguously decide on the validity of a proof; and often at least the high-level goals were supplied externally (e.g. for my thesis my supervisor indicated the theorem I was aiming to prove). By contrast, in RSP I was constantly facing a choice between employing any method from any discipline, often with little experience or guidance that would allow me to tell if I had, say, identified a good paper, whether I was making progress at a reasonable pace or should change my approach etc.”
“There’s relatively little oversight. This has obvious positives and negatives. Expect a fair amount of freedom, but little guidance or structure.”
“[I]t’s taken some experimentation to set up a management structure that works for me. Carving out space for your projects is also challenging when there are always interesting conversations, reading groups, docs that need feedback, etc.”
“I think my work is rather unusual for RSP. I’m not working on academic papers, but more a mix of writing code, writing blog posts, and organizing projects. My work is also a bit separate from much of the other work at FHI. That said, this is fairly unique work so I’m not sure what groups would be much closer.”
“I found it challenging at first to get a good sense of what type of research I wanted to do. I think this is normal when you enter a new field, but I can imagine that it is more stressful for some than others. While I got a fair amount of freedom regarding research, I sometimes felt there were too many other commitments during the week (talks, seminars, sessions,...) which especially in the first year had a negative effect on my productivity.”
What most surprised you about RSP?
“[I]n many cases, people are very open and honest about their motivations and mental states—this creates a culture that is very unlike anything I’ve experienced in the corporate world.”
Do you think RSP is a good place for you to be? How is it helping you if so?
“I think that some of the highest value things I’m getting out of RSP are exposure to new project ideas and new ways of thinking, and getting to add really great people to my network.”
“It’s been hugely valuable for me. A lot of that has been having space to do lots of personal growth—both in that there aren’t many constraints on how I spend my time, and in that [the programme organisers] were hugely supportive of this. My advisory board has also been really useful for guiding my personal growth, project choices, and longer-term planning.”
“Yes. There are two main ways RSP is helping me. The first is it enables me to work on my own research projects aimed at improving the future of civilization as much as I can. If I had to take a different position elsewhere, it is unlikely I’d be able to devote so much time to these. The second is the conversations within RSP and with others at FHI are helping me come up with ideas for projects and ways to improve the projects I’m working on.”
“[I]t provides the funding, flexibility, and university affiliation to enable me to very freely pursue the highest leverage opportunities I find.”
“RSP gave me the freedom to explore and get established in a new field through institutional, financial, and community support. It was the best career decision I could have made at the time.”
On RSP, what do you most miss about previous places of work or study?
“In previous roles, I have really enjoyed working towards a common goal in a closely knit team—this is something you’re not very likely to experience on RSP.”
“I also enjoyed working in larger teams. In addition, I find Oxford a bit isolating.”
“I enjoyed RSP and my colleagues, but at times would have liked to be in a more academically-oriented environment with more senior people that could have provided more subject specific guidance. As one of the only few philosophers, I sometimes felt a bit academically isolated.”
“As someone with a different specialization than the person above, I sometimes feel the same way. I’d expect this to be a pretty common experience for anyone who comes here hoping to dive deeply into a narrow subject area.”
What are you able to do on RSP that you wouldn’t be able to do otherwise?
“RSP makes it much easier to be part of conversations within FHI. I’ve found these pretty interesting and useful.”
“I’ve […] gotten to invest in lots of useful growth—people have been really supportive of me taking time to improve my physical and mental health, to practice research skills, to try out mentorship, etc.”
“In 2018, I co-directed the [CEA] Summer Research Fellowship, a summer program for research visitors. On this and other occasions I’ve mentored junior researchers, which I enjoyed.”
“I wouldn’t have been able to get so much free headspace to think about interesting issues, to have such amazing lunch conversations, and to get involved in so many exciting projects. FHI is a vibrant place, and there is always something really interesting going on.
What do you like most about RSP?
“The sense of empowerment. It really feels like a program that’s intended to help me do the things that are best for me to do. That, and the community—it’s put me in touch with a dozen of my favorite people.”
“RSP is incredibly open-ended, which allows you to focus on the most important work and projects. This means that every project that I spend time working on is one to which I am personally committed and convinced of its potential for impact.”
“Through RSP – both directly indirectly – I’ve met many great people. Somewhat ironically, I feel like RSP has helped me much more with personal growth and ‘soft skills’ than with my career or altruistic impact.”
“The freedom to take your time to figure out what you want to do and then do it.”
“Definitely the near-total freedom.”
What do you least like about RSP?
“There are few incentives to interact with the world outside of FHI. If I look at e.g. the number of talks I gave, numbers of EAGs I attended etc., they all went down compared to a previous 2-year period of having worked at another EA organization. This suggests to me that one requirement for thriving at RSP is to have a strong innate drive to seek out such interactions.”
“RSP is a very young place, and I often felt a difference in both age and/or research experience with some members in the cohort. The extremely consequentialist flavour of RSP also takes some time getting used to.”
- Shallow evaluations of longtermist organizations by 24 Jun 2021 15:31 UTC; 192 points) (
- 2021 AI Alignment Literature Review and Charity Comparison by 23 Dec 2021 14:06 UTC; 176 points) (
- 2021 AI Alignment Literature Review and Charity Comparison by 23 Dec 2021 14:06 UTC; 168 points) (LessWrong;
- More Centralisation? by 6 Mar 2023 16:01 UTC; 164 points) (
- 2020 AI Alignment Literature Review and Charity Comparison by 21 Dec 2020 15:25 UTC; 155 points) (
- 2020 AI Alignment Literature Review and Charity Comparison by 21 Dec 2020 15:27 UTC; 137 points) (LessWrong;
- EA Updates for August 2020 by 28 Aug 2020 10:29 UTC; 34 points) (
- EA & LW Forum Weekly Summary (6th − 12th March 2023) by 14 Mar 2023 3:01 UTC; 15 points) (
- EA & LW Forum Weekly Summary (6th − 12th March 2023) by 14 Mar 2023 3:01 UTC; 7 points) (LessWrong;
As a semi-outside viewer, who both works with several people in RSP, and has visited FHI back in the good-old-non-pandemic days, I highly recommend that EAs both apply to the program, especially if they aren’t sure if it’s right for them (but see who this program is for,) and talk to or work with people in the program now.
That said, I think that these comments are both accurate, and don’t fully reflect some of the ancillary benefits of the program—especially ones that are not yet experienced because they will only be obvious when talking to future alumni of the program. For example, in five years, I suspect alumni of the program will say:
It’s very prestigious step on a CV for future work, especially for EAs that are considering policy or academic work outside of the narrow EA world, and would benefit from a boost in getting in.
It gives people (well-funded) time to work on expanding their horizons, and focus on making sure they can do or enjoy doing a given type of work. It can also set them up for their next step by giving them direct experience in almost any area they want to work in.
The network of RSPs is likely to be very valuable in the next decade as the RSP program grows and matures, and the alumni will presumably be able to stay connected to each other, and also to connect with current / past RSPs.
Thanks for this! I agree that a lot of the value of RSP won’t become obvious until after the programme (and also want to flag that as our first cohort is only finishing this autumn, it’s still quite uncertain how large this value will be).
At this stage, the best information we have on how things will shape up for scholars after the programme is what our first cohort have lined up to do immediately after the programme—see here.
What percentage of people would you estimate were chosen from cold applications?
Just under half of the people we took I think nobody on the selection committee knew at all before the application process. For about half of the rest, they’d had a single conversation with me (& I think usually not with anyone else on the committee).
[From memory; haven’t checked carefully.]
I think we had 3-5 conversations prior to the RSP interview, the first one in 2017. Though I think “single conversation” still gives basically the correct impression as all conversations we’ve had could conceivably fit into one very long conversation. (And we spoke very irregularly, didn’t know each other well, etc.)
I also had had a few very brief online conversations with another member of the selection committee, and I had applied to an organization run by them (with partially but not completely overlapping material).
(I was counting you in the other “half of the rest”, i.e. people I’d had more contact with than a single conversation, so probably wouldn’t be regarded as “chosen from cold applications”.)
Ah, makes sense. Sorry, I think I just misread “half of the rest” as something like “the other half”.
One data point that gets at something similar (i.e. to what extent did RSP recruit from people with an existing network in EA):
I was one of 9 people in the first cohort of RSP (start October 2018). Before starting:
0 of the other 8 people I knew even moderately well,
2 people I had met before in person at events once but didn’t know well (to the extent of only having exchanged <10 sentences one-on-one as opposed to in group settings during the multiple-day events we both attended),
2 additional people I had heard of e.g. from online discussions (but hadn’t directly interacted with them online),
4 people I had never heard of.
I was surprised by this, particularly the first point. (Positively, as I tend to think EA is too insular.)
I had been working at EAF/FRI (now CLR) since mid-2016, based in Berlin, and had attended several EAGs before. Overall I’d guess I was moderately well networked in EA but less so than people at key anglophone orgs such as CEA or Open Phil.
I’d also be interested in hearing how competitive places are on the programme.
Typically, how impressive were the backgrounds and ideas of those accepted onto the programme? And what’s the acceptance rate like? I heard that FHI’s shorter summer research programme was extremely competitive.
And, during the programme if you’re struggling with the freedom, e.g. choosing a topic or choosing a methodology, what are the support options available to you like?
Impressiveness: good question, but feels hard to express without going into lots of detail, so I’m going to pass.
Acceptance rate: 9/~150, then 10 out of ~250. We’re planning to take 8 in this round. The summer fellowship was 27/~300.
Some support options, briefly:
Talking with Owen, the programme director
Talking with me or other future project managers on the programme
Peer support
FHI provides opportunities for coaching and other external support
We have various structures that aim to help people with this, like 6-week project cycles, a major project in the second year, a quarterly review process and an advisory board...
For the project cycles and major projects scholars would by default have something like a project supervisor
Apologies for the brief response, writing in haste!
It could be good if someone wrote an overview of the growing number of fellowships and scholarships in EA (and maybe also other forms of professional EA work). It could include the kind of info given above, and maybe draw inspiration from Larks’ overviews of the AI Alignment landscape. I don’t think I have seen anything quite like that, but please correct me if I’m wrong.
I also think that’d be good. To hopefully somewhat address this gap (though a written overview would still be useful), I’ve now created a tag for posts related to Research Training Programs, and tagged a few relevant posts I know of.
Note that those ratios are [number starting on programme]/[number of applications]. In fact a few people were made offers and declined, so I think on the natural way of understanding acceptance rate it’s a little higher.
Out of the rejection pool, are there any avoidable failure modes that come to mind—i.e. mistakes made by otherwise qualified applicants which caused rejection? For example, in a previous EA-org application I found out that I ought to have included more detail regarding potential roadblocks to my proposed research project. This seemed like a valuable point in retrospect, but somewhat unexpected given my experience with research proposals outside of EA.
EDIT: (Thanks to Rose for for answering this question individually and agreeing to let me share her answer here) Failure modes include: Describing the value of proposed research ideas too narrowly instead of discussing long-term value. Apparent over-confidence in the description of ideas, i.e. neglecting potential road-bumps and uncertainty.
One more data point: last year’s Summer Research Fellowship had an acceptance rate of 11/~90.
Do you think that the RSP is interesting for people working on policy engagement—eg writing “grey literature” reports, policy propositions, and feedback on legislation—or do you think it’s a better fit for people working on things in the “peer-reviewed/academic work” category?
I think it can be a good fit for either of those groups. Currently most people are more in the academic work category, but we have a few RSPers who are working on more policy engagement style work, and having a fair bit of success.
It’s also worth pointing out that plenty of RSPers don’t fall neatly into either camp:
Policy people sometimes do academic style things and vice versa
Lots of RSPers are exploring and haven’t yet narrowed down to ‘I am definitely going to optimise for policy engagement/some other style of work’
There are other buckets of activities that RSPers do: software development, teaching and mentorship, community building
Could you describe some of the more policy engagement style / community building projects people have taken on? Would be interested in what people have pursued as that’s closer to my area of interest vs more academic.
Some examples of policy stuff RSPers have done:
Advising governments directly (including the Czech and UK governments on their COVID-19 response, the Mexican government on its national AI strategy)
Networking with relevant government individuals, think tanks etc
Attending relevant conferences, like the Biological Weapons Convention
Writing papers, written submissions and reports on various policy issues
Some examples of community building related stuff RSPers have done:
Co-running programmes like the European Summer Program in Rationality (ESPR), FHI’s Summer Research Fellowship, the Human-aligned AI Summer School, and the Epistea Summer Experiment
Various kinds of advising calls, including lots via EAG/EAGx