On “fringe” ideas

Core claim

For effective altruism to be robustly good — good even if embedded in a society doing great evil — it must combine concrete, accountable work with genuine openness to arguments for caring about beings widely regarded as unworthy of care. Speculative research like wild animal suffering belongs in the movement both on those general grounds and on the specific grounds that welfare biology is a real research field with high-value questions.

Format

Originally a Tumblr answer on The Unit of Caring, republished with permission on the EA Forum. The prompt was a reader’s frustration that people “call themselves EAs” while writing speculative essays on topics like whether wildlife conservation is bad because wild animals suffer.

The historical thought experiment

• The formative question: if effective altruism had existed at other moments in history, would it have done any good, or would it have been stuck in the assumptions of its time?
  • Would an EA movement in the 1840s US have been abolitionist? If not, what must change now to avoid getting similarly big things wrong?
  • Would an EA movement in the 1920s US have been eugenicist? If it would have embraced the Progressive-era sterilisation campaigns, what habits of mind would have prevented that — and are we employing them?

What robustness requires

1. Actually doing things that benefit those who need it most. Donations moved through GiveWell to top charities (excluding Good Ventures) reached $65m in the year before writing. If that figure weren’t impressive and increasing, that would be a sign of community failure — the reality check of being accountable for actual results matters.
2. Continuously monitoring for signs that our work is doing harm — under many possible worldviews, including unintuitive ones.
   • Unhappy recipients are an enormous potential warning sign.
   • Efforts that increase suffering, even in weird hard-to-take-seriously ways, are a warning sign.
   • Forces systematically preventing us from hearing from recipients are a warning sign.
   • The point is to cast a very wide net so that the right answer is at least available.
3. Not expelling people who argue for expanding moral concern. Imagine someone walking into the 1840s EA group saying black people are exactly as valuable as white people, or into the 1920s group saying gay rights matter.
   • The stated principle: if something is an argument for caring more about entities widely regarded as not worthy of such care, then even if the argument sounds pretty absurd, I am supportive of some people doing research into it. And if they’re doing that research with the intent of increasing everyone’s well-being and flourishing as much as possible, then they’re part of our movement.

The resulting shape of the movement

• Three categories of speculative argument worth hearing out:
  1. Arguments for caring about things we wouldn’t normally think to care about.
  2. Arguments that our society is fundamentally and importantly wrong.
  3. Arguments that we are making important mistakes.
• All of this should happen only in conjunction with growth in resources allocated to concrete priorities — most effort on solid, clearly important work, with space alongside for the speculative.
• Why EAs should share a movement rather than a priority: if people had to leave the movement upon changing their minds about what matters most, they would be strongly motivated not to change their minds. Given deep uncertainty about priorities, a movement for everyone trying to do as much good as possible in an impartial, maximising, outcome-oriented way works better.

The specific case for wild animal suffering research

• There is vanishingly little research into what animals’ lives are actually like, especially wild animals.
• Almost no research on climate change, environmental measures or land use examines effects on animal suffering.
• Welfare biology is a new field addressing these questions. Piper expects fundamental welfare biology research — like fundamental development economics or medical research — to dramatically shape our understanding of which interventions are good ideas a decade out.
• This argument stands on its own: it doesn’t depend on the general case for speculative work. Just as vaccine developers and development researchers belong in EA, so do welfare biologists.

On frustration with people who seem to be wasting their time

• Given how much good each person could do, it’s hard not to take it personally when someone seems to be misallocating their effort.
• But being angry at people for being wrong about what’s most important isn’t sustainable long-term.
• Piper describes a major shift in her own thinking: a growing sense of how hard it is to arrive at true understandings, and how complicated and detailed reality is. This makes it easier to look at people who seem wrong-headed and be glad they’re trying, hopeful there is productive fruit down the path they’re walking.
• She takes speculative EA researchers to deeply care about improving the world, and hopes their research teaches them things she can use — even where she thinks they’re off-base.