Project Incubator

This project is centered on the Ψ Interspecific Affect GPT, an experimental AI tool developed by the Welfare Footprint Institute to support interspecific affective comparisons within the Welfare Footprint Framework. The tool is available through WFI’s AI Tools and Applications page. The scientific rationale, methodological approach, and current workflow are described in this article. The initial focus is on provisional, evidence-informed hypotheses about the maximum plausible intensity of Pain that members of selected animal taxa may be capable of experiencing, expressed relative to human-anchored reference categories. These estimates concern potential affective capacity, not the intensity that animals ordinarily experience. For each taxon, the tool produces an explicit reasoning dossier addressing taxonomic scope, the plausibility of sentience, relevant neurobiological, behavioral, cognitive, pharmacological, evolutionary, and ecological evidence, competing hypotheses, arguments against the preliminary conclusion, uncertainty, a provisional affective-capacity ceiling, and priorities for further research. The project would test both the substantive interspecific conclusions and the broader research workflow. Expected outputs could include:

• critically audited dossiers for approximately three to five selected taxa;
• a structured dataset recording AI claims, citation checks, errors, omissions, human corrections, and unresolved uncertainties;
• an analysis of recurrent AI failure modes and variation across repeated runs;
• recommendations for improving the tool’s instructions, evidence standards, knowledge base, and evaluation protocol; and
• depending on the maturity of the results, a technical report, preprint, academic manuscript, EA Forum post, or a combination of these.

Build, help others build, and help advertise interactive experiences that inspire wonder. This is part of an AI Safety Museum Exhibition project.

• We're a team of 10 volunteers.
• There’s room for machine consciousness and animal welfare projects as well.
• See CivAI’s We Need to Talk for inspiration.

I am a PhD student working on global environmental land use modelling and integrating animal welfare into the models. Additionally, I am working with researchers from Wageningen University to create a benchmark for large language models to assess their behavior and attitude towards animals.

The project involves jumping on opportunities as they emerge, so exact details will come in due time. We are doing whatever we can to reduce friction and incentivize frontier labs to address animal welfare in their alignment training. This includes a technical component (benchmarks and training data) as well as a social component.

Carp are the most farmed fish on earth — roughly 15–20 billion individuals a year in China alone — and almost all of it is sold live into the domestic market, so there's no export or certification lever and very little outside attention. Before sale, carp are commonly held for 30–45 days with no feed, at five to six times normal stocking density. The practice has been described in passing but nobody in animal welfare is working on it.

There's a promising crack in it. Removing the muddy off-flavour that drives depuration (geosmin and 2-MIB) appears to take only days, while the multi-week protocols are doing a different job — improving flesh texture and quality. If the long purge is discretionary quality work rather than technical necessity, it becomes far more challengeable — and at least one documented alternative already exists in the literature.

The mentee will produce the first clear-eyed assessment of the practice, to inform whether and how Myrias intervenes. The review will:

• Establish, with sources and evidence grades, why the purge is done and how its prevalence and duration vary by region and by tier of the supply chain
• Synthesise what it costs the animal — the welfare effects of prolonged fasting combined with crowding
• Assess the alternatives on both welfare gain and commercial plausibility

Deliverable: a standalone review plus a short prioritisation memo. No coding or domain expertise required — research synthesis, clear writing and honest evidence-grading matter most.

👤 This project is co-mentored by Henrike Gätjens. To apply, select only Philip Kratz as mentor.

Create an engaging, interactive info website in German, with the first goal of making knowledge about AI×Animals accessible to a German-speaking audience. Later, we might extend this into field-building and in-person meetings/summits. The website should:

• Showcase positive and negative scenarios, chances and risks for nh animals through AI
• Be catchy, picturesque, precise and informative