Comments
That is....a giant picture of my head haha But thanks for featuring me!
The role of artificial intelligence (AI) in our everyday lives is rapidly accelerating. Take ChatGPT: released less than three years ago, it now has over a billion weekly users, and is already vastly more capable than the original version. The same goes for other popular AI models, like Google’s Gemini or Anthropic’s Claude. And that’s just the tools that most of us have access to: behind the scenes, companies are increasingly using even more powerful AI for applications as diverse as financial trading, policing, and drug discovery.
Animal agriculture is no exception: major producers are already exploring how to integrate AI into their farms. In a couple of decades, the typical farmed pig might be overseen by an AI-powered CCTV system rather than a human farmer; China’s pilot ‘pork skyscrapers’ give a taste of what this might look like. This could bring benefits: for example, commercially available systems today can already monitor pigs’ movements and vocalizations, then notify staff to intervene if they detect signs of disease or distress. But AI-powered animal farming also poses enormous risks: given the animal agriculture industry’s track record, we should expect them to use new technology to maximize their own profits, regardless of what this might mean for the animals themselves.
Likewise, AI could help us monitor wild animals in far more detailed and sophisticated ways, greatly improving our understanding of other species and how we can help them thrive. But given the tiny amount of funding that goes towards improving wild animals’ wellbeing, this is far from a given, and it’s easy to imagine AI being used to exploit them, rather than support them.
Fortunately, AI also holds immense potential for animal advocacy. To help get a better sense for what this looks like in practice, we asked five of our Recommended Charities and Movement Grant recipients how they are currently using AI, how they expect AI to change their work in the future, and how they’re preparing.
Richie is Director of Research at Bryant Research, a food-system think tank that works to accelerate the transition to alternative proteins. He is also Lead Data Consultant at Greener by Default, an organization working to make plant-based foods the default across all foodservice.
As a research organization, Bryant Research staff are pretty heavy users of the latest wave of AI research tools, and two of the projects we’ve delivered would have been almost impossible without AI. For research papers, we’re big fans of Perplexity (an AI-powered search engine), and Elicit (an AI research assistant). We also regularly use the Deep Research modes in ChatGPT Plus and Google Gemini.
It’s hard to overstate how much AI deep research tools have changed research. They shortcut weeks of reading Google search results. Of course, everything that goes into our research reports is checked by a human, but the time savings are still enormous. We’ve been able to deliver some projects more than 25% faster, which saves our clients money and lets them make strategic decisions quicker.
The impacts of AI on research work is a huge question for us. Chris Bryant (our director) and I have even discussed whether Bryant Research might soon be rendered obsolete! If AI research tools continue to progress at their current rate, we may reach a point where the animal advocacy movement no longer needs external research organizations to do their desk research. I feel like this could even happen in the next 1-3 years. As a result, we’re thinking about how we can continue to offer unique valuable services to the animal movement. We would rather shut down than continue to extract resources from a movement that could do without us.
As such, we have a few plans to ensure we stay useful: first, we’re investing even more in services that AI cannot easily replace, such as our survey and polling work, focus groups, advanced data analysis, and movement-wide strategic research. Second, we’re leaning in to become the biggest power users of AI: not just staying on top of the latest research tools, but innovating on what research even looks like. That might be providing custom chatbots, dashboards, landing pages and more to allow our clients to engage with their research in exciting new ways.
Experiment constantly! Don’t wait to be told how to use AI by some “expert”. What’s unique and exciting about this current wave of AI developments is that you don’t need technical skills or expertise. You just need curiosity and willingness to experiment. No one fully knows all the things that AI is capable of, even the engineers who build ChatGPT and Claude. And no one knows your work better than you. That means you might just be the best person to transform your job with AI.
What worries me personally about AI is that adoption is so uneven. I’m concerned that we’ll see some people and organizations pull so far ahead that others won’t be able to catch up. I know plenty of people who use AI every hour of the day, but plenty of others who have never even opened ChatGPT. I know some organizations that ban AI usage, and others that are pulling off projects that would be totally impossible without AI. Our movement faces many threats, and I genuinely believe that lagging on AI is one of the largest.
Aaron is Co-Founder and Chief Operations Officer at Shrimp Welfare Project, which supports the seafood industry in its transition to higher-welfare shrimp farming practices.
I think that a major shift coming soon is that AI will become more entrenched throughout organizations, rather than it simply being a tool to help you do your work. Rather than managing people in the traditional sense, we might end up predominantly managing AIs, or managing people who get most of their work done through AIs.
The way we’re trying to prepare for that at Shrimp Welfare Project is by building an internal team culture that prioritises and normalizes working with AIs, so that we’re preparing for this shift. I found this blog post really helpful for understanding this idea and how to implement it.
I think that just as strong organizational health is a real hallmark of the best organizations (for more on that, I strongly recommend reading Patrick Lencioni’s The Advantage), I think the best organizations in the future will all have this strong AI culture underlying everything they do.
Right now I’m really focused on how Precision Aquaculture technologies (high-tech systems used in the farming of aquatic animals, like underwater cameras, sensors, and software) could enable continuous individual monitoring of individual animals, and how this can be harnessed to improve animal welfare. For example, smart feeders can adjust automatically when shrimp show signs of stress, which helps reduce overfeeding and poor water quality, which are both major welfare issues. I tend to think of this as “Precision Welfare.”
Most of these technologies are being developed for technically advanced aquaculture systems like the inland systems you find in Europe, the U.S., and certain parts of Asia like Singapore or China. However, there might also be a trickle-down effect where technologies initially get developed in these situations but then gradually become available in less intensive, more pond-based systems.
Overall, I think AI is going to enable intensification. Unfortunately, this just seems unavoidable. But I do think that AI could also provide a level of data and individualized monitoring that may actually allow us to improve shrimp welfare, if we engage with this in the right way.
I’m very excited that we’re at the beginning of another revolution in terms of how the world works. It’s really rare to feel like you’re not only living through history, but can hopefully direct it in a positive way.
But there’s also a flipside to this. There’s a lot of uncertainty about how the future might look, and I think it’s going to be very difficult for us to keep up with what’s happening. I think it’s also going to be very difficult to change course once certain paths are locked in. It feels like we’re at the beginning of something incredibly world-changing, and that is both exciting and also slightly terrifying.
Hannah is a Research Analyst in the animal welfare department at Rethink Priorities, a think-and-do tank that works to find tractable and neglected opportunities for impact across animal welfare and other cause areas.
My work is increasingly focused on the intersection of AI and animal welfare, especially how AI will impact animal agriculture. As such, my research questions are evolving from “How can we improve welfare in current systems?” to “How do we ensure AI-designed systems prioritize welfare from the start?”
Many feel like we’re on the cusp of something big, but no one yet knows how this will materialize. AI development brings so many large, important, open questions, and it feels like we need a lot more researchers working to try to answer these questions. But even working out the methods for how to try to predict AI developments is tricky. Since we don’t yet know if the default AI trajectory will be overall positive or negative for animals, it’s hard to recommend how advocates can promote AI developments improving animal welfare.
Another area that seems more feasible is working out what not to do: for example, it seems increasingly risky to carry out interventions with long timelines. When evaluating interventions, I now think about “AI resilience” – that is, will this still matter if farming becomes highly automated? Priority goes to robustly good interventions and foundational work that will probably remain useful across different AI trajectories.
Capacity building work also looks increasingly relevant. Advocates are having to widen their skillset to understand what AI tools are and what they can do. A lot of our forward-facing work involves trying to get a better sense for how AI might affect animal welfare or, failing that, sketching out plausible scenarios to be prepared for multiple ways AI could impact animal welfare. We don’t want advocates to feel like they have to watch AI happen and then wait for the dust to settle before they can work out what to do next; instead, we want them to feel able to work in the face of uncertainty and get involved in making AI go well for animals.
Similarly, I expect stakeholder engagement to increasingly focus on spreading awareness of AI developments, highlighting its impacts on animals, and engaging with AI developers. It’s also essential for us to increase our own knowledge of AI and technological concepts so that we can engage productively with those developing AI farming systems.
Janire is a Researcher at Wild Animal Initiative, an organization that advances research and fieldwork to better understand and improve the welfare of wild animals.
I currently use AI as a support tool for summarizing scientific literature and reviewing written materials, which greatly helps speed up research preparation and data synthesis. I am hoping to start to use it to also help analyze large sets of behavior data (like movement patterns or social interactions) which can help us understand whether wild animals are thriving or struggling.
I see AI playing an important role in how we assess animal welfare through more complex data types, particularly video footage and acoustic recordings of the sounds they make. These could enhance our ability to detect subtle stress indicators and behavioral anomalies in real time. This could help us spot things like unusual pacing, reduced activity, or vocal signs of pain right as they happen, instead of weeks later during manual reviews.
I’m particularly excited about AI’s potential to automate complex data processing tasks and incredibly time-consuming processing like “video labelling.” Right now, it can take hundreds of hours for researchers to go through video footage and tag different behaviors, so that we can spot patterns linked to poor welfare or evaluate whether an intervention is working. AI can speed this up dramatically by automatically flagging relevant clips.
AI is moving fast, and that can make long-term planning tough, especially in research where consistency and ethical boundaries really matter. I see AI as something that can support human judgment, not replace it. That mindset helps me stay open to new tools without losing sight of what makes good science reliable.
Animal advocates must be thoughtful and deliberate. AI offers powerful capabilities, but it’s not neutral; it reflects the data and values fed into it. In animal welfare contexts, poorly applied AI could lead to misinterpretations or even harm. My advice is to treat AI as a collaborative tool that requires careful design and oversight. If used responsibly, it can bring about significant improvements in how we assess and protect wild animals.
Sagar Shah, Rethink Priorities
Sagar is a Senior Researcher at Rethink Priorities, focused on farmed
animal welfare.
AI tools speed up my writing tasks by around 50-75%. For general writing, I use the voice-to-text tool Superwhisper to describe what I want to write, then I have the Large Language Model (LLM) Claude Pro produce a draft. I then follow up with voice prompts to provide feedback and corrections.
I no longer really use Google search. Instead, I use voice mode to ask LLMs questions, providing them with context about my interest level and the details I need.
Sagar is a Senior Researcher at Rethink Priorities, focused on farmed animal welfare.
I find that AI tools have massively increased the speed at which I understand new topics, and I can now complete tasks that previously took days or weeks in just hours. This means I can spend less time aggregating information, and more time on strategic thinking about the different actors involved in actually bringing about change.
I use Deep Research tools (like Gemini Pro) two or three times a week to summarize information across 50-60 websites. My key use cases include:
For quick literature reviews, I’ve replaced manual searches and bibliography reviews with:
I expect current AI tools will improve significantly, greatly reducing the human effort needed for writing, summarizing, and information gathering. However, at least over the next few years, I believe human work will remain essential for:
My preparation strategy is to actively use available AI tools while focusing on value-added work that’s beyond current AI capabilities. It’s also important to keep abreast of developments in AI tools and make good use of them as they’re released and developed. For example, when I first tried Deep Research tools around the start of the year, the hallucination rate was sufficiently high that I didn’t trust the answers, but that’s changed now.
I advise readers to try the tools extensively. After months of heavy use, I can’t imagine working without them. I believe they offer incredible value for the price. There’s a real risk of falling behind: Those who don’t use AI effectively might soon be significantly less productive than those who master the technology.
AI is already having major impacts on advocates’ work. All signs point to continued AI progress, meaning that these impacts are only likely to increase. As a movement, we can’t afford to get left behind: it would be extremely damaging if those people and organizations causing enormous animal suffering harnessed AI while animal advocates failed to keep up. If you take just one thing away from our interviewees, it’s that we must all trial AI tools ourselves and think about how best to use them for the good of our organizations and the broader movement. Implicit throughout all of the interviews is that we are on the cusp of something huge; let’s not waste this opportunity to steer AI in a positive direction for all species.
Here are some steps to get you started:
Executive summary: This exploratory piece gathers perspectives from five animal advocacy leaders on how AI is reshaping research, farming, and organizational practices, highlighting both risks (e.g. intensification of animal agriculture) and opportunities (e.g. faster research, precision welfare, advocacy tools), and urging advocates to experiment with AI now to avoid falling behind.
Key points:
This comment was auto-generated by the EA Forum Team. Feel free to point out issues with this summary by replying to the comment, and contact us if you have feedback.