Outdated assumptions

The design of the animal advocacy movement—the goals we are pursuing, the strategies we use to pursue them, and the division of people and money across those strategies—is built on a set of assumptions about the future which are no longer valid.

Subscribe below and I’ll invalidate even more of your assumptions at no cost to you.

Funding

The movement may no longer be funding-constrained at ~$300 million/year worldwide. Frontier AI companies like OpenAI and Anthropic (makers of Claude) will likely go public in the next couple years. At current projections, this would cause the largest transfer of wealth to rigorously altruistic people in human history, thanks to cross-pollination between AI researchers and the EA community going back to the 2010s.

By the end of next year, even a small fraction of Anthropic employees’ philanthropic funds being directed to address factory farming could multiply the capacity of our movement several times over.

Reasonable people disagree about the extent to which animal advocates have been constrained by funding up to now. But if you expect AI to continue to accelerate (as I do) we should expect funding to stop being our problem. Instead, we will quickly become constrained by our ability to effectively absorb this funding and turn it into programs that actually help animals.

However: here be dragons. As a founder, I know how easy it is to say, “Somebody has extra money they need to spend? I’ve got just the thing.” But we should hold ourselves to a higher standard. We should aspire to direct any new funding in a way that is no less cost-effective than our current approaches, or even more cost-effective.

This should make the challenge clear. Your organization is, I hope, focusing your resources on what you believe are the best opportunities. By default, the marginal next opportunities you could pursue with extra staff, ad spend, etc. are less promising, and the next one even less.

I’m convinced that for most organizations in the movement, doubling our budgets without substantially reconsidering our approach would cause our cost-effectiveness to plummet, or even in some cases reduce our net effectiveness by leading us to spend money carelessly.

But I’m equally confident that we can find ways to become even more effective with new funding. Rather than adding a garnish to our existing programs, ask yourself: what opportunities had we previously ruled out because we simply couldn’t afford them? What would it look like to compete in an entirely new weight class? And this is only compounded by the next faulty assumption:

Labor

Strategies that once required a linear increase of human effort to scale will soon be cheaply automated. Over the coming years, we will need to continuously re-evaluate our model of what types of cognitive labor must be done by humans vs. what can be scaled almost infinitely for low cost using AI.

Video and other forms of content are already being automated. How would your strategies change if you could cheaply design individualized video ads for thousands of narrowly-sliced audience segments– or even for each individual viewer?

Skeptics of individual diet change advocacy have dismissed the strategy in part because it scales too poorly to make much difference. For years, the website for Anonymous for the Voiceless listed the total number of promising conversations AV activists had had with members of the public, and after years of a worldwide movement, it was smaller than the number by which the human population increases in a single day.

AI could change this. In theory, it would be possible to deploy AI vegan outreach agents to talk one-on-one with every human on earth. There are obstacles to putting this into practice, but it suddenly becomes a solvable problem, a question not of “can we?” but “should we?” and “How?”

We need every expert in various subdomains of animal advocacy to think about what automated, infinitely-scalable human-level intelligence means for their specialty.

Institutions

Governments, companies, and basic economic incentives that we previously took for granted may not endure, at least not in a form we would recognize today. How will corporate pressure campaigns work when AI renders all forms of human labor economically worthless, replacing human employees from retail floors to boardrooms? What will consumer demand mean in a world without scarcity– and perhaps without money?

The collective efforts to extract cage free policies from corporations and governments around the world were a masterclass in long-term strategy. Campaigners started with easy targets, using those wins as a launchpad to larger targets. They got the private sector to make commitments almost a decade out, then used those commitments to justify passing state laws forcing compliance from companies who may never have planned to follow through. All this took many years to start mattering for living animals, but the patience paid off.

This kind of strategy is not viable in 2026.

Try thinking of the arrival of transformative AI sometime between now and 2040 as an event horizon. The world until then fits Eisenhower’s parable that “Plans are useless, but planning is indispensable”– unexpected things will happen, specific plans will have to be flexible, but we can anticipate the general direction by extrapolating from current trends.

But at some unknown point, we will cross the event horizon. The world on the other side will be as unrecognizable to us as the present would be to a preindustrial farmer, or perhaps a preagricultural hunter-gatherer, or maybe even a yeast gobbling sugars before the advent of sexual reproduction.

This doesn’t mean we throw up our hands. It means that now is the time to start thinking hard about what strategies could survive across such a horizon– or at least pay off quickly enough to matter beforehand. We have situational awareness that foragers and yeasts never had, but we should still be humble and realistic about what we can predict.

Weighing our options

Humble, realistic… and rigorous! Which brings us back to the point of this post.

I’m asking for your help to generate a rich list of strategies that the animal movement could use to adapt to imminent AGI. This is the brainstorm phase: we need to search across a wider idea space than any small group of strategists could come up with.

After you add your ideas here, I’ll be working with a team of volunteer researchers convened through the Sentient Futures project incubator to evaluate each idea against a diverse set of possible AGI scenarios. As we publish our findings, the strategies that look most promising will be turned into pilot advocacy programs by mid 2026, either launching as new organizations or being taken on within existing orgs.

That means the ideas you submit now could turn into real advocacy work this year. To help you brainstorm, here are a few broad categories that hopefully give a sense of how wide the search space could be.

Leveraging AI for new strategies

For now, most of what AI is doing is not making previously impossible tasks possible. Rather, it is making previously expensive tasks cheap. That opens up strategies that were once bottlenecked on inputs that could only scale linearly with human effort, such as:

Opposition research

Attentive conversation partners

Highly customized media content

What are the most important ways animal advocates could leverage AI to scale strategies that were previously impossible?

Short-timeline big bets

Assume that AI will dramatically overturn the existing social and economic order, with the event horizon in as little as 5-10 years. That leaves us with two broad options: either we focus on delivering tangible change to animals before then, or we bet on effects that will endure.

You might think that even if institutions are upturned, cultural values and popular opinion will influence the mid-term future. Or maybe only technological improvements will matter. Or you may just want to build a strong, agile movement that can best react to new opportunities.

What strategies should animal advocates be focusing on to make quick, durable progress -or- build our strategic capacity for an uncertain future?

Research agendas

Maybe we can try to reduce our uncertainty about the future. We could try to understand what causes species bias in AI models and how to reduce it in future, more powerful models. We could invest in improving their ability to help farmed animals, and even wild animals. We could try to understand how the emergence of superintelligent AI will influence public opinion about other nonhuman minds.

What research agendas should be most prioritized for increasing the likelihood that AGI goes well for animals?

AI company relations

The companies building frontier AI systems are in a position to exercise enormous influence over the future, controlling massive resources and shaping the most important technology in human history.

How should animal advocates go about building relationships with AI companies? What should our goals be in building these relationships?

Modifying existing strategies

All this uncertainty probably doesn’t mean we should throw out our existing strategies, especially the ones that have proven successful. But these approaches at least need an update. They can be streamlined or scaled up with AI, and we should consider whether their goals and priorities should be adjusted for short timelines.

How should animal advocates modify existing primary strategies in light of short AI timelines? Which current strategies could/should be scaled with relatively low friction?

By friction, I mean that if scaling an intervention would require hiring and training new campaigners linear to cost, it will be difficult to scale, whereas “money pits” like mass media and political spending don’t require increased human capital.

Share your ideas

Provide your input at this link:

https://www.tricider.com/brainstorming/36eenMwaMqN

Build on,

Sandcastles

Spread the love by getting your smart friends to respond as well!

Share