You're pledging to donate if the project hits its minimum goal and gets approved. If not, your funds will be returned.
$50,000 left to reach the funding goal.
Modeling Cooperation seeks funding to develop an interactive simulation of how transformative AI could affect farmed animals and how decisions by labs, governments, and animal advocates might change those outcomes.
We’ve developed the software behind Intelligence Rising, a scenario-based simulation of AI futures used to explore competition and cooperation between labs and state actors. We propose applying this experience to animal welfare - an area where, despite the scale of suffering risk for animals, the strategic interactions are poorly understood.
Our immediate objective for this funding is to design and produce a working software demonstration. Depending on the results of an initial demo, we would then seek to collaborate with organizations experienced in facilitating workshops, to bring our wargame to relevant decision-makers and run it in live settings.
We are pleased to share an endorsement from Jasmine Brazilek at CaML for this application:
‘We fully support Modeling Cooperation in creating this game. We have known them for years and their ability to execute is really impressive. We think their ability to reach policymakers with these wargames is one of the most effective levers for change within short timelines.’
The interactions between various stakeholders are difficult to reason about in isolation. A simulation allows participants to explore where incentives conflict and what opportunities for interventions might reveal themselves. Our target group would be animal advocates and policymakers who want to test their theories of change about how rapid AI progression will impact animal welfare.
We propose to:
Design an AI × animals scenario. Using existing work, including Sentient Futures’ simulation as described on the EA Forum, we will identify a set of actors, decisions and strategic interactions suitable for an initial demonstration.
We will produce an interactive simulation in which participants can make decisions, observe their consequences and explore how different strategies produce different outcomes.
Once we have a demonstration, we will approach potential collaborators to explore piloting the software in facilitated workshops.
To be clear, we are not seeking funding to run workshops ourselves at this stage. Our contribution is scenario design and software development. We would work with organizations that have the audiences and facilitation capacity to put the tool into use, similarly to how we’ve worked with Technology Strategy Roleplay previously. We would consider a custom simulation for exploring how rapid AI progress could impact wild animals in a future iteration of the game.
The most tangible measure of success is the production of a simulation that can be used by people outside our development team.
We will assess success against the following benchmarks:
Functionality: The software supports a complete, playable simulation with meaningful decisions and outcomes.
Usability: Someone unfamiliar with the development process can operate it with limited technical assistance, following a short onboarding session.
Information value: The simulation illustrates how decisions by different stakeholders could influence outcomes for farmed animals, including opportunities to reduce suffering.
Scenario credibility: The simulation's assumptions and strategic interactions are reviewed by relevant subject-matter experts such as CaML, with feedback incorporated into the prototype.
Adoption potential: The model gains interest and discussion within AI safety communities, with at least one organization working on AI × animal welfare expressing concrete interest in using the wargame.
We are requesting $50,000 to produce a working simulation over six months. Most of the funding will pay for software development.
Our proposed budget is:
Software development - $25,000: Build the simulation (wages, AI coding agent subscriptions, API costs), including how it works and how people interact with it.
Scenario design - $15,000: Review existing work, for example, Sentient Futures’ AGI × Animals Wargame as publicly documented, and develop scenarios involving AI developers, animal agriculture, governments.
Testing and improvements - $5,000: Test the software and make improvements based on feedback.
Documentation and preparation - $2,500: Prepare the software so we can demonstrate it to potential partners and make it available for future workshops.
Fiscal sponsorship fee - $2,500: Convergence Analysis's 5% fee for fiscal sponsorship and associated administrative services.
Modeling Cooperation began in 2019 following the co-founders’ participation in the AI Safety Camp. We are fiscally sponsored by Convergence Analysis.
Jonas Emanuel Müller leads our work on AI competition dynamics and has over two decades of experience as a software engineer. He previously chaired the board of Animal Charity Evaluators, co-founded two local groups working on farmed animal welfare, and was on the board of other animal/effectiveness-focused NGOs. He is currently on the advisory board of CaML.
Paolo Bova is a PhD candidate in computer science at Teesside University, where his thesis examines how early warning systems can de-escalate AI competition dynamics. He holds a bachelor’s in economics from Trinity College Cambridge.
Tanja Rüegg manages funding reporting and communications, alongside a chief operating officer role at a data analytics firm. She worked for and was on the board of the animal welfare organization Sentience Politics and co-led local groups working on farmed animal welfare.
Siobhan Ballan leads our partnerships and fundraising work, having previously worked at Sentience Politics and Animal Charity Evaluators.
Our research network includes a DPhil candidate in international relations at the AI Governance Initiative at the University of Oxford, and leading researchers including Professor Robert Trager and Dr. Shahar Avin (CSER, Cambridge & UK AISI).
Our track record includes
1. Research software for decision-makers.
Our main project is the software platform behind the Intelligence Rising workshops, run by Dr. Shahar Avin and his team at Technology Strategy Roleplay. These workshops use a structured simulation game to help senior figures in governments, industry and academia develop an intuitive grasp of how AI competition unfolds and what can be done about it. Around 800 participants have taken part since 2022, of which about 200 have experienced our software. We also build and maintain several interactive tools that let researchers and policymakers explore AI competition models directly, including the Safety-Performance Tradeoff web app, developed with Professor Robert Trager.
2. Original research on policy and competition dynamics.
We have a track record of novel research into AI policy. Our report “Safe Transformative AI via a Windfall Clause” shows how a Windfall Clause can be used to de-escalate dangerous competition dynamics even when the situation is uncertain and there is incomplete information. Our upcoming research collaboration with AIWI assesses the consequences of revealing secret information about AI risks and how to make such disclosures of early warnings more credible.
3. Open tools for other researchers.
We share our models so that other governance researchers can build on them. Our repositories are accessible to other researchers, and our tools have been used by PhD researchers studying AI competition.
The simulation may be too speculative. AI’s effects on animals are highly uncertain. We will address this by keeping the initial model simple, making assumptions explicit and treating scenarios as tools for exploring possibilities rather than forecasts.
The software could add little value over existing tabletop exercises. Sentient Futures has already demonstrated that these questions can be explored without dedicated software. Our prototype needs to offer a meaningful improvement in data collection, usability and repeatability.
There could be insufficient demand from potential users. We can build the software, but cannot guarantee that another organization will deploy it. We will therefore limit the initial investment and seek feedback from potential collaborators before committing to going further.
We have not raised funds for this project from any other funder and don’t currently intend to apply elsewhere. We have raised nearly $600k over the past seven years for our other work, from funders such as SFF and Foresight Institute. You can see our full funding record on Trace.
There are no bids on this project.