Our moonshot goal is to build an AI social planner that learns optimal economic policies for the real world. We aim to develop more sophisticated AI policy models that meet usability requirements, such as explainability and robustness, to recommend real-world economic policies that improve social welfare.
This project seeks to scale up the AI Economist (Zheng et al., 2020). We hope to leverage reinforcement learning and machine learning techniques to develop a tool that will automate and improve tasks such as policy design, policy evaluation, and economic objective optimization, which will allow economists and policymakers to focus on the end goal: improving social welfare.
Build and Collaborate
Collaborate on research, contribute code, and interchange ideas.
Develop rich economic simulations, evaluate AI policies, and explore how AI can solve complex economic problems.
Guide the research and tell us which social issues you would like to address using our framework.
Our Trusted AI principles are a key part of our research and publish process. By adhering to them we aim to foster research accountability and results which benefit everyone.
We will seek multilateral feedback from affected communities, civil rights groups, and advocacy organizations for underrepresented groups.
Fair and Representation Data
We will use data that is fair and representative of all groups, including minority groups, being aware of the social context that produced the data.
Our recommendations should not be unduly influenced by funding partners or other financial interests and will only be made after consulting with all stakeholders.
We will provide clear documentation for public data sets and models, for example, through the use of data sheets, model cards, or appropriate alternatives.
We will open-source the code used for research and any real-world policy recommendations.
Ethical Risk Analysis
We will conduct ethical risk analyses for published work, as in Zheng et al., 2020.
We will evaluate economic policies against objective quantitative metrics, such as GDP, Gini Index, and others.