The Future of Humanity Institute is launching a Research Scholars Programme, likely to start in October 2018. It is a selective, two-year research programme, with lots of latitude for exploration as well as significant training and support elements. We will offer around six salaried positions to early-career researchers who aim to answer questions that shed light on the big-picture questions critical to humanity’s wellbeing.
At this stage, we’re collecting expressions of interest in the programme. See Applying for details.
Humanity today may be in a position to affect the long-term future: whether and how intelligence spreads through the universe, and what use it makes of it. Working out which actions in the world are particularly valuable for the long-term — and working out how we should think about this — are hard but crucial topics. They are studied within the Future of Humanity Institute under the name macrostrategy. We think a robust understanding of macrostrategic considerations helps people undertake better (for the long-term) projects and make better (for the long-term) decisions at many different scales.
In this programme we aim to foster this skill of choosing actions for their all-things-considered effects, and give people space to exercise it. We will give them tools for thinking about the long-term impact of research and other activities. In most cases we foresee participants coming out of the programme with views on topics such as: what the future might look like at timescales of tens, thousands, or billions of years; what the largest determinants of different possible futures are; which technologies are most likely to be transformative at a timescale of decades; what the largest levers in the world are for affecting these. We will then encourage participants to explore different research ideas they have, and practice making judgements about what to pursue. (This need not be macrostrategy research; the idea rather is that the choices are macrostrategy-driven, i.e. informed by thinking about long-term consequences first. We think research in any of the other areas studied at FHI can fall within this category, as well as some other topics.) We will support this with mentorship and the best other feedback loops we can construct, and a collaborative, curious working environment to encourage exploration and challenging of ideas.
Why choose this rather than the traditional academic model where people start by gaining deep expertise in an area?
We see two reasons:
- Academia often forces people to choose early, committing to many years for a PhD before having an opportunity to spend a few months on each of a few different types of research. For people who know exactly what they want to do this has the virtue of bringing them to the point of making technically impressive contributions faster, but for others it can be costly by making them commit early. And given the incentives in the academic job market, if one might want to take time to explore areas, it is often better to do so before a PhD than after.
- Macrostrategy research is highly multidisciplinary, and there is not just one type of relevant expertise. Similarly, we think that — because of diminishing marginal returns — many of the most valuable questions to work on fall between traditional disciplines and have no real experts.
Owen Cotton-Barratt will be heavily involved in running the programme. He has run previous experiments in getting people more involved in FHI’s research, for example organising this seminar series in 2015.
Why participate in this programme?
This is a unique opportunity for people who intend to soon assume research and research-adjacent positions critical to addressing the world’s most pressing problems.
In particular, we expect that you will get a lot out of the programme if you:
- Have lots of ideas, but are unsure which ones to pursue and could benefit from research mentorship;
- Are interested in multiple different research areas, or in macrostrategy work itself;
- Would benefit from the freedom to explore your options (for instance before committing to a topic for a PhD or postdoctoral project, or working out whether you can contribute best through direct research or an adjacent activity); and/or
- Are excited by being part of a cohort of exceptional individuals with shared curiosity and desire to identify and pursue the highest-value opportunities.
On the other hand we don’t expect this to be right for everyone interested in impact-driven research. In particular this might not be the right fit for you if:
- You have a clear idea of what you want to work on long-term (and it isn’t one of FHI’s research areas). In this case you might well be better working on it immediately, and doing so with experts (for instance by doing a PhD).
- You would prefer to be part of a larger research team, so that you don’t need to be able to self-direct and could be better off directly joining an organisation in such a role, or becoming excellent at the type of research tasks that you might be given.
Since this is the launch of a new program, the closest we have to a track record are two examples of people who worked for around a year as junior researchers with significant freedom in research directions and mentorship from Owen Cotton-Barratt. Max Dalton now heads the research and content team at the Centre for Effective Altruism. Ben Garfinkel is now a researcher at FHI.
People in the programme may work on topics in FHI’s research areas, including global priorities research, AI strategy, AI safety, and reducing catastrophic risks from biotechnology. We hope that many will go on to careers working in or adjacent to these areas, either directly pursuing or otherwise facilitating research. Direct research roles are a natural next step, but we believe that many important roles (such as working in community-building or grant-making) are helped by a good understanding of important research directions.
The programme will hire people as researchers, likely within the Future of Humanity Institute in Oxford, UK in the joint offices with the Global Priorities Institute and Centre for Effective Altruism. We plan to start in October 2018, although we have not ruled out delaying until January if this would result in a stronger cohort. We expect the cohort to have around six participants, but it could be as few as four or as many as ten. We want to make the best version of the programme, which means that these decisions (and others) will be sensitive to our getting a better sense of what strong candidates for the positions look like (see Applying).
The programme is somewhere between a research degree, a residency, and a research fellowship. The programme will pay participants salaries, likely of at least £28,000. We also expect to be able to provide visas for successful applicants.
It will accept exceptionally talented individuals who are interested in research, and in choosing research questions based on global-scale impact, but not necessarily committed to a particular approach. Participants need not have significant research experience, but they must be able to demonstrate aptitude. For applicants with PhDs, or other significant experience, we expect to be able to offer higher salaries and more senior job titles. There will be little structured internal division according to level of prior experience, but a lot of the work will be self-directed and mentorship will be personalised, so we expect those with more research experience will have a slightly different experience.
We seek participants eager to spend their time on a combination of research, curriculum, and applied work, with priorities between these activities largely determined by the participants themselves.
- Research activities are meant to develop participants’ methodological toolkit, engaging with topics within cause prioritisation, existential risk, and artificial intelligence strategy to both obtain new skills and generate new insights. They include cursory field overviews, quantitative impact evaluations, and theoretical argumentation.
- Curricular activities supplement this work, augmenting research efforts with formal thinking techniques, advanced domain knowledge, and general academic skills. While topical focus will vary greatly depending on one’s background, we will offer resources to cover both substantive and procedural topics, such as Fermi estimation, mechanism design, long-term-future modeling, and data-driven analysis.
- Applied activities are meant to make researchers deft at ensuring people use their work. They include pursuits such as conference presentations, workshop design, project management, and networking.
We want to provide the structure and support which will be most useful in the long-term for participants. Because this is the first year of the programme, the details below offer just a current best guess. We expect the plans to change as we go and in consultation with programme participants — indeed we hope that the participants will join us in taking ownership of making the programme as valuable as possible. For this reason there will be few if any obligatory aspects (although there may be some we strongly recommend).
Our guess is that the ideal structure for the first year will include:
- Weekly or twice-weekly reading and discussion groups covering a range of macrostrategy-relevant topics, with access to established researchers to provide comments or give talks.
- Weekly or fortnightly research exercises to help train research-relevant skills (e.g. Fermi estimate and calibration exercises; ways of writing clearly without getting bottlenecked).
- A majority of time kept clear for people to pursue their own projects, which could include research ideas as well as more applied research-adjacent activities (e.g. organising a workshop).
- At least weekly individual mentorship sessions to discuss projects and directions.
- Regular opportunities to share things with the group, and perhaps occasional collaborative group projects.
- Every couple of months, a research workshop of one to four days to dive into a topic more intensely and generate ideas.
- Trips to visit partner organisations within the UK and internationally.
We guess that a number of mutually beneficial collaborations will arise naturally between people on this programme and researchers at FHI or partner organisations such as the Centre for Effective Altruism and OpenAI. We think people will typically need less structure as the programme goes on. We will be very happy for people to make an extended visit away from Oxford to work with collaborators, or to leave for another opportunity part-way through the second year.
In general, we expect programme participants to be active stakeholders in decisions about what will best benefit them. We will therefore avoid prescriptive curriculum, enforced project formats, and forced collaborations. Instead, we focus on offering infrastructure and mentorship as guidance and support. We expect participants to identify and pursue appropriate goals — for example we think in some cases people will choose to publish academic papers, but in some cases this will not be the most effective way of disseminating research.
Ideal candidates will:
- Be sharp and curious. The best researchers are often primarily motivated by captivation. This likely describes you if you read papers in a range of disciplines for fun, or are inclined to dive into fields even (and especially) if they’re confusing or seem wrong.
- Pay attention to scope and long-term impact. Habitual quantitative thinking is a powerful tool for informing decisions. Focus on long-term outcomes is central to macrostrategy, so intuitive access to it will be useful throughout the programme and helpful to others in the cohort.
- Be engaged and conscientious. Crucial research questions are often in uncharted territory. It’s therefore critical that candidates are not paralysed by too much choice, but happy to strike out and try things. Similarly, since the programme is so new, we’re seeking candidates who see a flexible structure as an opportunity, and instinctively do something to fix problems they notice.
Note that we do not list a specific degree requirement for ideal candidates. We expect to be able to accept students both in the middle of and after having obtained degrees at the undergraduate and graduate levels.
If you are interested in participating in the programme, we encourage you to submit an expression of interest by 25 May. We will attempt to provide estimates of the likelihood of acceptance to potential candidates who do so, particularly for those candidates considering other options (such as graduate programmes) with imminent decision deadlines.
Submitting an expression of interest will also help us better gauge the expected cohort size, research priorities, and levels of experience. This will directly feed into our planning for the first year.
Official job applications will likely open in June. We will announce application process dates as soon as they become available, and keep everyone who has expressed interested updated. To be alerted when the application becomes available and/or submit an expression of interest, please submit this form.
We will follow up with people who demonstrate interest on a rolling basis, so potential candidates with imminent decision deadlines are encouraged to contact us quickly and indicate urgency.
If you have any questions not addressed here please email firstname.lastname@example.org.