Sentient Futures: building at the forefront of transformative AI and nonhuman welfare
AI systems that will affect trillions of animals and potential artificial minds are being developed now, mostly in the Bay Area, by a small number of people who are setting norms quickly. Sentient Futures builds the community and talent pipeline that brings nonhuman welfare into that process.
We run courses to prepare people to contribute, incubators and residencies to turn ideas into projects, and summits to get everyone in the same room. Our Slack community with 2,000+ members is one of the largest and most active coordination hubs for nonhuman welfare.
Our track record:
- Built AnimalHarmBench, the first benchmark of how language models treat animals (FAccT 2025).
- Ran the coalition that got nonhuman welfare named as a systemic risk in the EU’s General-Purpose AI Code of Practice, the first mention of animals in an AI regulatory text.
- Advised multiple frontier AI labs on inclusion of nonhuman welfare considerations in their alignment docs, eval suites, and training data.
What marginal funding contributes:
We intend to scale up all our programs and events in 2027, with a moderately ambitious funding gap of about $1.6M. At the margin:
- $150 funds one AIxAnimals course participant
- $200 funds one incubator participants; $500 more covers compute and project support
- $10,000 funds one resident-month in San Francisco, covering allowance, food, compute, and project support.
- Larger gifts let us run the residency year-round instead of a single 3-month cohort. They also enable us to pay more competitively for both new hires and existing staff.
You can learn more about us here and reach out to me (Managing Director) or @Constance Li (Executive Director) for questions. You can also directly support us here.
Insecticides kill more animals directly than any other human activity. What should we do about it?
Insecticide use on agricultural land kills more animals directly than any other human activity, plausibly in the quadrillions of animals (not including springtails, mites, and other very small invertebrates, which might push these figures much higher). Despite this, there is practically no research or work on the animal welfare impacts of these practices. And insecticides aren't the only highly neglected animal welfare issue on agricultural land — other agricultural practices use or kill billions or trillions of animals annually, such as the use of parasitoid wasps for biological control, managed honey bees used for pollination services, and wild residents of cropland killed by harvesting practices or stubble burning.
I've started a new organization that works to reduce the harms of these practices on animals. Our initial planned intervention is a large-scale program to induce farmers to use less painful insecticides. Our work is complicated to talk about publicly so we aren’t making a top-level post, but we think that helping animals on agricultural land is plausibly among the most cost-effective things that can be done to help animals, and that very promising interventions are on the near horizon.
We’d love to share more about our work with donors interested in learning more. Our core budget funding gap is around $350k for 2027, our ambitious budget gap is much larger, and we’re aiming to deploy initial interventions in the near future. Complete this brief form, and we’ll share a document with more information. You can also support us directly here.