Transformative AI and Animals: Animal Advocacy Under A Post-Work Society

By Kevin Xia 🔸 @ 2025-05-25T18:32 (+53)

Thanks to Max Taylor, Irina Gueorguiev, Robert Praas, Albert Didriksen, Mark Rogers and Justis Mills (EA Forum) for feedback on this post. All mistakes are my own. This post does not necessarily reflect the views of my employer.

Executive Summary

Contextual Notes 

This piece explores the evolving roles of industrial animal agriculture and farmed animal advocacy in a society shaped by transformative AI (TAI).[1] As an illustrative example, I explore mass automation, universal basic income (UBI) and the emergence of a post-work society (i.e., a society in which people can live well without being economically compelled to work).[2] I chose to explore this scenario, because I find it to be the most compelling among genuinely transformative scenarios - and because I find the implications particularly illustrative.

However, this piece does not intend to be read as a forecast. I don’t intend to claim or defend the likelihood of post-work, the constraints and milestones to get there, nor any particular implications of post-work for farmed animal advocacy efforts. Instead, I hope to pose questions that I believe to be underexplored at the intersection of post-TAI economics, industrial animal agriculture and farmed animal advocacy, and to illustrate their potential implications and relevance through this example. 

Introduction

With AI systems exponentially improving in many areas, forecasts of when we develop artificial general intelligence have generally become shorter, with many predicting AGI in the next 5-10 years. These systems, then, could have transformative effects on our society, the economy and how we do things, including our efforts to end factory farming. Some authors have talked about “a century in a decade” or similarly accelerated rates of development, which, if we can sufficiently address the risks associated with such compressed lines of events, could bring us reason for excitement.[3] But even in cases where we avert many of these risks, things don’t necessarily look rosy for non-human animals. Preparing for these types of scenarios seems increasingly urgent. This piece examines a generally positive outcome post-TAI (a UBI-powered post-work society)[4] and explores possible changes in the dynamics of farmed animal advocacy efforts and industrial animal agriculture.

What becomes of systems like factory farming, which are justified by its perceived efficiency, and the desire to protect agricultural workers? And on the flip side, what becomes of non-profit work when money, wages, and time are no longer the central constraints?

Post-Work Farmed Animal Advocacy Work

Non-profits today, especially in the farmed animal advocacy movement, are largely constrained by funding. Notably, this funding is primarily used to pay staff members, as we need people to dedicate their full-time work hours to helping animals.

A post-work society changes this calculus. When people are no longer bound to paid employment, they may choose to contribute to causes they care about out of alignment rather than necessity. We may see passionate, capable individuals engage in advocacy full-time without requiring salaries—massively increasing the labor pool for mission-driven work and in some ways leveling the playing field. 

This shift could also challenge traditional organizational design in many ways, some for the better, some for the worse. With AI staff available, new non-profits may be started left and right (including by representatives of industrial animal agriculture), effectively drowning each other out as they compete for stakeholders’ attention. Rather than recruiting staff, non-profits might cultivate distributed networks of contributors. Movement governance might shift toward decentralized collaboration, with AI supporting coordination and knowledge sharing.

However, our current spending on advocacy efforts may not be reflective of the post-TAI scenario. If cutting-edge AI remains paywalled for organizations, spending may simply shift from labor to AI capabilities, leaving us funding-constrained nonetheless. Moreover, the very same mass automation through TAI that enables this talent pool would also automate much of the work that nonprofits do. So what remains for humans to do in farmed animal advocacy?

I am skeptical of attempts to predict which exact roles will remain better done with humans in the loop, but some loose guesses of the type of work we could be looking at include:

Notably, it remains unclear whether and to what degree these roles even persist. For example, how important is major donor fundraising, if we are not primarily constrained by funding? The question, then, is about finding, preparing for and effectively leveraging human-scalable strategies[5] that are robust against or even greater in potential under the economic system that enables them. Furthermore, it would be helpful if these strategies, then, would be asymmetrical, i.e., could not be just as easily deployed by the industry or a movement to defend the industry.

Post-Work Industrial Animal Agriculture

At first glance, one may be hopeful that in a post-work, abundance-oriented society, factory farming may just cease to exist. If animal products can be replaced by cheaper, more ethical alternatives—and if profit no longer drives production—why would factory farms still exist? A couple of reasons seem plausible:

None of these are certain, but I think they are important to investigate as part of the broader question of how the system of industrial animal agriculture may (want to) be upheld under transformed economies.

Open Questions for Farmed Animal Advocates

As noted before, I don’t have any answers or specific recommendations, other than the mere fact that we should think more deeply about how the system around us will change under transformative AI and how that may affect our predictions on leveraging AI interventions. To re-emphasize, my intentions are primarily to pose these questions, such that other people can provide answers and meaningful recommendations to the movement. As such, here is a list of questions that came up as I explored the dynamic of a transformed economic system and how it changes the dynamics of industrial animal agriculture and farmed animal advocacy work:

Closing Thoughts

At its core, a post-work future would not just be a policy shift, but a moral turning point. If we can eliminate suffering at scale, if no one needs to work in exploitative conditions, if we can replace harm with compassion, then failing to do so becomes a deep ethical failure. The legitimacy of factory farming today largely rests on a perceived trade-off: economic efficiency versus harm. For many, this trade-off is already indefensible, so what if it disappears?

The challenge, then, is not just to make alternatives economically viable, but to make compassion culturally inevitable. In a world where no one has to work, some will still choose to do so—not for wages, but for meaning. Farmed animal advocacy may be one such calling. Yet to remain relevant, it must evolve. We must prepare for a world where traditional leverage points, such as funding, labor, or outrage, may lose potency. We must anticipate the narratives that legacy industries will adopt. And we must invest in cultivating moral imagination, cultural influence, and future vision.

  1. ^

    In line with the EA Forum Topic, I define Transformative artificial intelligence (TAI) as artificial intelligence that constitutes a transformative development, i.e. a development at least as significant as the agricultural or industrial revolutions. This does not, technically, require artificial “general” intelligence in the sense that it does not need to be able to perform all or almost all tasks that humans can perform.

  2. ^

    A post-work society isn’t necessarily one where no one works—it’s one where work is no longer a prerequisite for survival or a decent life. People may still choose to work for meaning, identity, status, or social contribution, but not to meet basic needs. I find it helpful to log different economic scenarios on a scale that compares “average work week” vs. “quality of life” - you’d have 60 hour work weeks to "survive" a century ago, to 40/35 hour work weeks today and would continue to scale towards 20 hour, 10 hour and ultimately “0” hour work weeks, first to survive/to live a decent live (i.e., post-work)  and ultimately to thrive (i.e., post-scarcity). 

  3. ^

    I, for one, would have expected the end of factory farming in less than 100 years. The prospects of getting there in a decade feels absurd, but we should probably expect absurd things to play out under transformative AI.

  4. ^

    The idea here is usually that AI systems would become increasingly capable at cognitive tasks and remote tasks, including helping design robots and other physically embodied systems that would further replace physical tasks. The idea that such AI systems would actually lead to mass automation, that mass automation would lead to job losses and that job losses would lead to UBI are all far from obvious, but assumed for the purpose of exploring this radically transformative system.

  5. ^

    I.e., strategies that increase significantly in effectiveness with the number of humans involved.