Coefficient Giving's Global Catastrophic Risks (GCR) division — including our work on navigating transformative AI and reducing catastrophic biorisk — is on track to spend well over $1 billion in 2026, and might spend multiples of that amount in future years. Timelines to transformative AI seem shorter than they did a few years ago, and we may have only a few remaining years to make a difference. At the same time, the amount of philanthropic capital available for helping with the transition to transformative AI looks much larger. The situation is changing rapidly, and we'll have to make many high-stakes decisions in the coming years. To ensure we make those decisions well, we need the support of exceptional researchers to improve our strategy, identify cruxes, model possible AI trajectories, and assess different interventions to avert global catastrophic risk and improve the long-run future.

About the teams

We're hiring for senior research roles across the GCR Central team (led by Emily Oehlsen) and the Short Timelines Special Projects team (led by Claire Zabel). We're open to a wide range of experience levels, and the title and shape of the role will be tailored to the successful candidate. We're particularly excited about candidates who could manage others and lead a focused research team, in addition to doing research themselves.

GCR Central

Emily Oehlsen and George Rosenfeld lead Coefficient Giving's GCR division and the teams within it: AI Governance and Policy, Technical AI Safety, GCR Capacity Building, Short Timelines Special Projects, and Biosecurity and Pandemic Preparedness. We set the GCR division's high-level strategy, manage the program teams, and own high-priority special projects to ensure the division achieves its goal of navigating transformative AI and reducing catastrophic biorisk. We are ultimately responsible for the division's success, and work on projects that determine how to direct our giving and allocate scarce staff capacity across sub-fields.

Short Timelines Special Projects

Claire Zabel leads the new Short Timelines Special Projects (STSP) team. Our remit is to drive forward new projects that look especially important if timelines to transformative AI are short — and, more specifically, to make sure that wherever funding can vastly improve outcomes in a short-timelines world, it does.

We do what it takes to set up the highest priority projects. That sometimes means making grants, but it often requires other efforts in addition to allocating funds: writing a proposal for an organization that doesn't yet exist and finding someone to start it; convening a small group of people who should be talking to each other; doing the macrostrategy research that tells us what the right grant programs even are; or working with for-profit and unconventional levers when they're the best vehicles for impact. STSP takes on work that falls between the remits of the other GCR teams: identifying what should exist before it exists, finding the right people, and helping it get off the ground.

(The STSP team is also hiring for other roles and may be willing to shape them around the right candidate – for more information on this, see the Appendix.)

About the roles

While all roles share a similar profile in terms of responsibilities and expected skills, there are some differences depending on the team.

Senior Researcher / Research Director (GCR Central)

We're hiring a research lead or senior researcher to own the strategic questions that sit above any single team and shape what the whole division does, working as a close thought partner to Emily and George. The decisions we make over the next few years (e.g., which sub-fields to back, how fast to spend, when to start something new) may determine how well we can help humanity in navigating the transition to transformative AI. Many of the questions that should inform those decisions are ones we don't currently have the time or the specialized depth to answer well. We aim for this role to help close that gap.

This is a senior position, reporting to and working directly with Emily and George. We're ideally looking for someone who already has enough context on AI strategy to engage the hardest questions quickly. We expect to make one senior hire, and could build a small team around a strong lead over time.

Here are some examples of the questions we'd expect this person to own:

• Where should we set our funding bar, given our best read on investment returns, future fundraising, and the most promising large-scale spending opportunities? Where we set the bar could change how we spend billions of dollars.

• How should we update our views of AI timelines and the most important threat models, and how should this shape what we prioritize?

• How should we trade off priorities across the division, e.g. how should we allocate scarce staff capacity to different sub-teams or projects, and how should we effectively draw input from advisors inside and outside CG?

• Which areas should we prioritize for expansion, e.g. by spinning up a new team?

Beyond any single one of these, we'd want this person to be the first port of call when Emily, George, the GCR team leads, or CG's wider org leadership have a strategy question, holding flexible research capacity for whatever is most decision-relevant at the time. We expect this role to be plugged into organizational needs (versus pursuing a more open-ended research agenda).

We're looking for someone to maintain situational awareness across the division and the wider field and surface considerations that should change GCR or sub-team strategy before we'd otherwise notice. As a result, we anticipate that an excellent research lead will spend significant time building relationships with external experts and designing efficient ways to draw out their views on decision-relevant questions.

Research Fellow / Research Officer / Senior Researcher / Research Director (STSP – SSSOs)

We are hiring for one or more researchers to support us in identifying and evaluating super scalable spending opportunities (SSSOs) — high-impact giving opportunities where billions of philanthropic dollars could be effectively deployed — that might emerge in the next couple of years.

The key question we expect researchers to seek answers to is how a foundation or coalition of funders could support specific large projects (at scales of $10s of billions or more) to help the transition to a post-transformative-AI world go better. As the list of potential SSSOs is long and varied (and this hire would plausibly make it even longer), the role could involve investigations into topics ranging from how nonprofits can secure access to compute before and/or during an intelligence explosion, what post-ASI philanthropic spending opportunities may emerge during such a period, and how public institutions (both existing and new ones) can become more prepared for a world shaped by transformative AI. For each of these questions, we are most interested in findings that generate insight into practical implications for current and future philanthropic spending and investments.

Right now, we don't have the capacity to explore many of the most plausible-seeming ways in which huge amounts of philanthropic resources could be leveraged to make the transition through transformative AI go better. If timelines are short, that might mean that some of the largest-scale opportunities can't be realized, and vast resources that could've made a difference are used much less effectively than they could have been, or not at all. We hope that filling these roles will substantially contribute to closing that gap, increasing the probability that these considerable resources are used more effectively than they otherwise would have been.

Research Fellow / Research Officer / Senior Researcher / Research Director (STSP – RTT)

We are hiring for one or more researchers to support us with Research on Transformative Technologies (RTT) that may emerge during a potential intelligence explosion. An intelligence explosion could accelerate progress not only in software and AI R&D, but in other sciences and engineering disciplines that could themselves be terminally decisive — for example, technologies that could cause catastrophic risk themselves, provide a durable advantage to a small set of actors over other actors, or otherwise facilitate "lock-in" of bad outcomes. AI-driven biological risk is an obvious example, but there might be others, for example risks related to advances in nanotechnology or robotics that advances in AI may enable. Currently, our picture of when, to what degree, and in what order AI will unlock the most important of these technologies is unclear. If timelines are short, that might mean we are avoidably caught off-guard and miss opportunities to prevent bad actors from accessing such technology, develop defenses, or take other measures. We hope that filling these roles will substantially contribute to closing that gap, increasing our preparedness for what could be the most consequential physical technologies in human history.

We want researchers to investigate questions like (1) which physical technologies could become decisive, (2) when they might arrive, (3) what bottlenecks their arrival (other than AI capabilities), if anything (4) what their effects would be, and (5) what we can do about it to make the transition to advanced AI be better and safer.

This likely involves investigations into topics ranging from whether AI can meaningfully accelerate nanotechnology or atomically precise manufacturing, how feasible self-replicating systems (from autonomous manufacturing to space probes) are and on what timescale they might arrive, to which material inputs (such as particular rare earth metals) might bottleneck advanced manufacturing, to how quickly energy generation could be scaled once AI is very capable. For each of these questions, we are most interested in findings that generate insight into practical implications for current and future philanthropic spending, investment, and preparedness.

About you

Researchers at CG tend to have a broad remit, but in most cases (and for all of the advertised roles) you will do some combination of the following:

• Conduct research and minimal-trust investigations to inform program strategy and grantmaking by helping to investigate a new grantmaking area or model the expected cost-effectiveness of a given intervention.

• Stretch across disciplines to form a holistic view on strategic questions.

• Generate hypotheses, gather evidence and talk to domain experts to synthesize and distill isolated data points to come to an all-things-considered view, including in areas you know little about.

• Write up your findings in an accessible and action-oriented way.

• Build back-of-the-envelope models to estimate the returns and cost-effectiveness of different interventions and philanthropic opportunities, or to improve our understanding of crucial questions.

• Conduct prioritization exercises to identify the next big priorities the team should be working on.

You might be a good fit for these roles if most of the following applies to you:

• You are open to working on a wide variety of topics, depending on the needs of the team and an ever-evolving list of research questions, rather than specializing deeply in one sub-area. You are excited about working in relatively unscoped areas with little-to-no established evidence to rely upon.

• You are comfortable working in a fast-paced environment and can prioritize getting to tentative and imperfect answers quickly. You are a pragmatist who is able to make and justify reasonable estimates with limited data and come to concrete conclusions. You prioritize the most important questions to inform a decision rather than trying to answer everything.

• You have a relatively practical and concrete orientation towards research, i.e. you are excited to be part of a decision-making process that results in real-world change and affects high-stakes decision making rather than pursuing a particular predetermined or exclusively conceptual/theoretical research agenda.

• You have strong analytical and critical thinking skills, especially the ability to quickly grasp complex issues, find the best arguments for and against a proposition, and skeptically evaluate claims. You should feel comfortable thinking in terms of expected value and reasoning quantitatively and probabilistically about tradeoffs and evidence.

• You have good judgment, can identify and focus on the most important considerations, have good instincts about how promising the additional marginal hour of research is on a given topic, and can form reasonable, holistic perspectives on complex questions and wicked problems. "I take short timelines seriously" and "I have real uncertainty about timelines and impact" are not mutually exclusive to you.

• You enjoy original thinking, writing memos that change minds, picking apart claims, and going to first principles when the field's consensus is thin or wrong.

• You have a bias towards action: People who've worked closely with you describe you as some combination of agentic, calibrated, and fast.

• You have a relatively deep and broad understanding of AI strategy and AI futurism.

• You communicate in a clear, information-dense, and calibrated way, with good reasoning transparency, both in writing and in person. You are able to communicate your thinking clearly and concisely and explain what you believe and why, including why you chose to weigh particular evidence over other evidence and where you're uncertain.

• (For senior versions of these roles:) You have research management or leadership experience and know what it takes to mentor and support junior researchers, and to drive forward a research agenda with relatively little oversight and steering.

Signs this might not be the right seat:

• You strongly prefer working in a well-defined role with clear deliverables and a narrow scope. You mostly want to row, not steer.

• You'd be unhappy if your team's focus shifted significantly in rapid succession, or you'd feel sad if you had to abandon a project you've worked hard on. You prefer to take a question and go away for a year to work on it.

• You have a strong need for fast and heavily empirical feedback loops, and the ability to cleanly attribute impact in the world to your actions.

• You are trying to build a public portfolio and want to prioritize publishing interesting conclusions rather than privately informing decision makers.

None of these are deal breakers on their own, but if several of them describe you, we suspect you might be happier elsewhere — and we'd rather you find that out before applying than three months in. The application process is designed so that you can find out if that's the case, and we are happy to talk through any uncertainties you might have.

This job description is intentionally broad and covers a range of potential outcomes in terms of scope and seniority. This reflects our willingness to shape the role and title around specific candidates.

Logistics

• Compensation for candidate based in San Francisco or D.C.: We anticipate total compensation for these roles to be as follows (including an unconditional 401(k) grant of $24,500):

  • Research Fellow: $254,000

  • Research Officer: $280,000

  • Senior Researcher - Research Director: $319,500 - $416,000

There would be a 10% downward adjustment for candidates not based in one of our hubs (SF or D.C.).

• Time zones and location:

  • The GCR Central team has no hard location requirements for their role, though we prefer candidates based in (or willing to relocate to) the San Francisco Bay Area, Washington, D.C., or London, but we are open to considering candidates in other locations.

  • The STSP team prefers candidates to be based in the San Francisco Bay Area, but we are open to considering candidates in other locations.

  • We'll support candidates with the costs of relocation. We'll also consider sponsoring U.S. work authorization for international candidates (though we don't control who is and isn't eligible for a visa and can't guarantee visa approval).

• Benefits: Our benefits package includes:

  • Excellent health insurance (we cover 100% of premiums within the U.S. for you and any eligible dependents) and an employer-funded Health Reimbursement Arrangement for certain other personal health expenses.

  • Dental, vision, and life insurance for you and your family.

  • Four weeks of PTO recommended per year.

  • Four months of fully paid family leave.

  • A generous and flexible expense policy — we encourage staff to expense the ergonomic equipment, software, and other services that they need to stay healthy and productive. This policy also includes a productivity benefit, which provides a set amount for staff to expense items that enhance their productivity.

  • A continual learning policy that encourages staff to spend time on professional development with related expenses covered.

  • Support for remote work — we'll cover a remote workspace outside your home if you need one, or connect you with a Coefficient Giving coworking hub in your city. We currently have offices in San Francisco and Washington D.C., and multiple staff working from several other cities in the U.S. and elsewhere.

  • We can't always provide every benefit we offer U.S. staff to international hires, but we're working on it (and will usually provide cash equivalents of any benefits we can't offer in your country).

• Start date: The start date is flexible, and we may be willing to wait for an extended period of time for the best candidate, though we'd prefer successful candidates to start as soon as possible after receiving an offer.

Research staff are typically employed by Coefficient Giving's 501(c)(3), and as such are likely eligible for public service loan forgiveness programs.

Appendix: More about us and our work

Why CG, and why now

GCR Central and STSP are small, fast-moving teams, with mandates built around filling the most important gaps quickly. We have substantial resources and ambitious plans — and right now, hiring is the binding constraint on getting it done. Our guiding mission: If transformative AI comes soon, and money could have saved us from a catastrophic outcome, let's make sure it does.

The case for importance of our work rests on two claims:

1. Transformative AI may arrive much sooner than many had expected. Recent work by METR finds that the 50%-time horizon of the most recent Claude Mythos Preview model is at least 16 hours (approximating levels that the current task suite can't even reliably measure); scenarios like AI 2027 sketch out in concrete detail what a relatively short-timelines trajectory could look like. We're also beginning to see early evidence of recursive self-improvement, with AI systems taking on a growing share of the work of building their successors. We know that reasonable people can disagree on timelines and their implications, and we take these disagreements seriously. But conditional on shorter timelines, a meaningful share of the philanthropic impact possible this decade may be concentrated into a small window — and we think building teams and hiring around that possibility is the right move to improve our preparedness.

2. A lot of high-leverage AI work is upstream of "the next marginal grant". Some of the most useful things we can do — identifying gaps in the ecosystem, recruiting people into roles that don't yet exist, pushing on macrostrategy questions the field hasn't articulated, deciding where to set the funding bar for a division spending over $1 billion a year — aren't well-served by grantmaking alone. These roles are meant to create space for someone to launch minimal-trust investigations into questions like "What does the philanthropic menu look like 3–6 months ahead of a transition to a world shaped by transformative AI?" and have that work translate quickly into concrete moves.

3. Coefficient Giving has decided to invest seriously in this. We have deep relationships, context, convening ability, and substantial financial resources, and the organization is committing meaningful headcount and capital to this work, including by scaling STSP and adding dedicated research capacity at the division level. The mandate is unusually broad: we can choose between many levers — grants, original research, founder recruitment, ecosystem coordination, working with for-profit entities — rather than being committed to any one of them. We expect the resources behind individual workstreams to grow significantly as the team grows.

Why this work matters

There are projects on our list right now that we believe are important and that aren't getting done — either because they're unstaffed, or because the one person on them is too stretched to do the work justice. Every additional strong hire can make the difference between taking another bet and leaving it on the table; between running another grant program and not running one; between scoping out a critical initiative and not getting to it in time. In other words: The world needs more "general managers" for the most important problems, and we want to support our own staff or our external partners in filling that gap. Research is a critical input to that, as it will affect many influential decisions we take further down the line.

Both teams are small enough that the next several hires will substantially shape what they become. For someone with the right combination of judgment, ambition, and ability to operate without much scaffolding, this is a huge opportunity: many strong performers should expect to grow into leading a sub-team, founding a new organization, or owning a major workstream within a year or two.

Snapshot: What the work looks like on the STSP team

STSP roles are unusually varied, and there is a mix of breadth and depth – we both value flexibility and acknowledge the necessity of specialization. New joiners on our team can generally expect to work on some combination of the following:

• Identifying what should exist and making it happen. Writing proposals for new orgs or programs, recruiting and developing founders, helping new initiatives spin up, and following through with funding and support once they exist.

• Original strategy and macrostrategy research. Scoping research questions, writing memos that change how we (and other funders, researchers, and decision-makers) think about a topic, and turning that thinking into concrete asks, programs, or hires.

• Active grantmaking. Running focused grant programs, designing RFPs, evaluating proposals where we have specific views about what should be funded, and helping promising grantees be more ambitious.

• Talent identification and recruiting. Finding the right people for important roles — at grantees, at orgs we'd like to see exist, and on our own team. You won't be doing this work alone, as you will have the support of a recruitment team and our Chief of Staff.

• Coordination across the ecosystem. Convening small groups of senior people working on adjacent problems; brokering collaborations and making sure the right people are talking to each other; identifying unmet needs across funders, researchers, and labs.

• Working with for-profit and unconventional levers. Where traditional nonprofit structures alone aren't the right tool, we're open to working with for-profit entities, investing, supporting industry coordination, and other mechanisms.

Most strong candidates will be a blend. If you think there is a chance you can do any of these well, please get in touch with Moritz (moritz.vonknebel@coefficientgiving.org), and we'll figure out together where you might add the most value. We expect that new joiners can both specialize over time (taking ownership of a part of our portfolio) or develop more breadth.

FAQ

Do I need to choose between the GCR Central and STSP roles?

No. The application form lets you indicate interest in any or all of the three roles, and the evaluation process is heavily overlapping by design: we're looking for similar candidate profiles across all three. Tell us if you have a preference and we'll take it seriously, but staying open won't count against you; we'll work out together where you'd add the most value as the process goes on.

What do you mean by short timelines? Are short timelines really likely enough to organize a team around?

We think that transformative AI is plausible within one to four years, and that short timelines are plausibly likely enough to act on. Sensible people have a wide range of views, and so do we. The case for organizing a team around shorter timelines doesn't require certainty, and (i) the evidence has shifted toward shorter timelines over the last couple of years; (ii) the cost of being prepared if timelines are short is likely much lower than the cost of being unprepared; and (iii) much of the work we do is robust across timeline scenarios.

Is the work actually counterfactual?

Often, very much so. A meaningful share of what we do is upstream of the next marginal grant — proposals for organizations that don't yet exist, scoping work, headhunting, ecosystem coordination — where progress is often slow or nonexistent absent active intervention. And as Luke Muehlhauser put it in a recent post, a single new hire can expect to move up to hundreds of millions of dollars in a single year, and that "hiring one fewer grantmaker usually means those millions will just sit in an account for another year rather than being deployed to useful ends". Especially if timelines are short, that might make the difference between the money going unused and the money being used to support a top-priority activity. The research roles are highly counterfactual as well: they help shape high-stakes strategic decisions we have to make one way or another, and strong research capacity means we make better, more calibrated decisions.

Can't you just hire whoever you want? I feel like I'm more needed elsewhere.

Hiring is genuinely difficult for us. We've had rounds where we've under-hired relative to what we wanted, and lowering our bar would have diffuse but real costs — particularly for our agility and culture, which matter a lot under time pressure, and our ability to do sound decision making in complex and sign-unclear areas. So if you turn down an offer, that probably does mean we have one fewer person doing this work. If counterfactual fit is the deciding factor for you, tell us — and if we make you an offer, we'll give you our candid read on how counterfactual you are.

Why work here rather than at a frontier AI company or in government?

It depends on the person and on comparative advantage; we think many people should pursue roles in those places and others. Coefficient Giving sits in a useful zoomed-out position, and STSP specifically is unusually free to act across many levers — grants, research, recruiting, coordination, for-profit work. If your strengths are around prioritization, scoping, ecosystem-building, and acting fast across many levers, this is one of the better seats we know of.

What does growth look like here?

Pretty varied. Strong performers tend to grow into a portfolio shaped around their strengths. Likely paths include taking on bigger and more ambitious workstreams and becoming a "general manager" for a given problem or intervention, working with a variety of other funders and philanthropic entities, leading hiring rounds, founding or seeding new organizations, managing people (each of who could move tens or hundreds of millions of dollars every year), or setting strategy for an area of the portfolio. We are particularly keen to hire people who could grow into senior leadership positions.