EA needs more “rooms of one’s own”

Midtermist1214 Oct 2025 19:14 UTC
25 points
14 comments5 min readEA link

[Draft Amnesty—this is incomplete and I’m looking for feedback on whether this direction is worth pursuing]

The observation

Virginia Woolf argued that to write fiction, a woman needs money and “a room of one’s own”—not just physical space, but financial independence and protected time for creative work. Looking at intellectual history more broadly, a striking pattern emerges: many significant breakthroughs came from people who had this kind of freedom.

Charles Darwin spent decades at Down House, financially secure from family wealth, meticulously gathering evidence for evolution without any professional obligations. Adam Smith received a pension from the Duke of Buccleuch that let him spend nearly a decade writing The Wealth of Nations. The 1956 Dartmouth workshop that founded the field of AI was enabled by a modest Rockefeller Foundation grant that gave researchers freedom to explore.

Not all breakthroughs required wealth—Marie Curie made her discoveries despite extreme poverty. But the pattern is clear: financial independence, when present, enabled specific kinds of work that are hard to pursue otherwise: long-term evidence gathering, high-risk exploration, synthesis across domains, ideas that challenged institutional orthodoxy.

The problem in EA

Today’s EA movement has largely structured itself around a model where a small number of individuals with significant resources (major funders, organization leaders) set priorities, and then funding flows to specific projects and organizations that address those priorities. This has produced substantial impact.

But this structure has a significant limitation: it concentrates both power and the capacity for exploration among a narrow group. Who gets to spend significant time thinking about new cause areas, exploring novel approaches, or developing ideas that don’t yet fit into fundable project categories?

Largely, it’s people who happen to have financial bandwidth—those with inherited wealth, tech industry windfalls, academic positions with light teaching loads, or partners who can support them. This is a privilege filter that operates invisibly. We don’t notice how much of EA’s “independent thinking” comes from people who can afford to do it.

Meanwhile, many talented people who could make important contributions are locked in a different situation: they’re working full-time jobs in unrelated fields, thinking about EA in whatever spare hours they can find, unable to properly develop promising ideas because they’re cognitively exhausted from financial precarity.

What this costs us

The psychological research on scarcity is stark: financial precarity doesn’t just reduce available time, it imposes a “bandwidth tax” on cognitive capacity. Studies show that contemplating a significant unexpected expense temporarily reduces cognitive performance by the equivalent of 13-14 IQ points. The chronic stress of making ends meet depletes exactly the mental resources needed for the kind of deep, strategic, long-term thinking that EA requires.

This means we’re losing potential contributions in at least three ways:

Lost exploration of new directions. Ideas that could become important cause areas or methodologies never get developed because the people who had them lack the bandwidth to pursue them properly. They remain as interesting forum posts or passing conversations, never fully fleshed out.

Lost “unlocking” ideas. Some ideas don’t just have direct impact—they unlock additional resources or capacity. Historical examples: GiveWell’s methodology made giving legible to analytical donors, unlocking substantial funding. Early AI safety work created a field that now attracts billions. Wild animal suffering research opened an entirely new cause area. How many similar ideas are currently stuck because their originators lack freedom to develop them?

Movement capture by default. When only people with existing privilege or those funded by a small number of sources can afford to explore, the range of ideas investigated naturally narrows toward what those groups find intuitive or interesting. We lose intellectual diversity not through active exclusion but through structural constraint.

Historical evidence this matters

Recent research on the relationship between financial independence and intellectual breakthroughs reveals several key patterns:

The mechanisms are specific. Financial independence doesn’t guarantee breakthroughs, but it enables particular things:

  • Removes the “scarcity mindset” cognitive burden, freeing mental bandwidth for deep work

  • Provides time for long-gestation projects (Darwin’s decades of evidence-gathering, Smith’s decade writing)

  • Enables risk-taking on high-uncertainty projects that traditional funders reject

  • Frees people from institutional constraints and the need to conform to departmental priorities

  • Creates psychological security to explore unpopular or controversial ideas

Modern programs confirm this. MacArthur Fellows consistently report that unrestricted funding enabled them to take on more ambitious projects and pursue riskier, more theoretical ideas that were difficult to fund through traditional mechanisms. Academic sabbaticals, despite unclear impact on publication metrics, are valued by recipients specifically for enabling the kind of unstructured exploration that doesn’t produce immediate measurable outputs.

But financial independence is an accelerant, not fuel. The successful historical cases shared key traits: intrinsic motivation that predated their financial freedom, demonstrated track records of excellence, and access to intellectual communities. Darwin’s wealth enabled his mission; it didn’t create it. This suggests that providing “rooms of one’s own” is most effective when granted to people who are already proven but constrained.

The proposal

EA should experiment with providing financial independence—“rooms of one’s own”—to individuals who have demonstrated promising ideas or capabilities but lack the bandwidth to fully develop them.

This would be different from current EA funding in several ways:

Not project funding. Traditional grants fund specific projects with defined deliverables and timelines. This would fund people for open-ended exploration, with the understanding that the most valuable outputs might not be predictable in advance.

Not organization building. We’re not funding people to start and run organizations (though that might emerge). We’re funding the earlier stage: developing ideas to the point where they could become organizations, projects, or frameworks that unlock additional resources.

Longer time horizons. Probably 3-5 years minimum, not 1-year renewable grants. The goal is to remove financial precarity and the cognitive burden it creates, which requires enough runway that recipients aren’t constantly worried about what comes next.

Different success metrics. Not papers published or projects completed, but: novel research questions generated, paradigms shifted within fields, new cause areas or methodologies identified, ideas that unlock additional funding or capacity, and valuable knowledge from “successful failures.”

Design challenges I haven’t solved

Selection is genuinely hard. How do you identify people with the right combination of intrinsic motivation, capability, and promising ideas? An open application process would be noisy. A nomination-based system (like MacArthur) might work better but raises questions about who gets to nominate and whether this just replicates existing networks.

Measuring success is contentious. The whole point is to enable work that doesn’t fit conventional metrics, but “trust us, patient qualitative evaluation” is hard to defend in a movement that values quantification. How do you evaluate 5 years later whether this was worthwhile?

Movement capture risk is real. History shows that funders shape movements even with the best intentions. The NAACP in the 1920s shifted from anti-lynching work to education litigation because that’s what the Garland Fund would support. How do you fund exploration without inadvertently narrowing what gets explored?

Any program would need: genuinely independent governance, intellectually diverse selection committees that rotate regularly, explicit communication that challenging funder priorities is encouraged, and tolerance for work that might make funders uncomfortable.

Opportunity cost is high. The money could fund proven interventions or established organizations. The bar should be: does this create more expected value than those alternatives? I’m genuinely uncertain, which is why I’m posting this for feedback.

Questions for discussion

  • Is “early-stage idea exploration” actually a bottleneck in EA, or are we limited more by execution capacity?

  • Are there better interventions than this for addressing the concentration-of-power and privilege-filter problems?

  • What would convince you this is worth trying as an experiment (say, 2-3 people for 3 years)?

  • Can anyone point to current examples of promising ideas/​people that are clearly bottlenecked by lack of bandwidth?

  • What failure modes am I not seeing?

I’m particularly interested in hearing from people who feel they’re in this situation—talented and motivated but locked out by financial constraints—and from people who think this is a bad idea and can articulate why.

(note: the ideation, drafting, and revision was done with assistance by generative AI)

Midtermist1214 Oct 2025 19:14 UTC
25 points
14 comments5 min readEA link