The AI for Math Fund seeks to advance the pace and impact of math discovery by supporting projects that are important for the field, but no one academic or industry lab has the incentive to do them. The fund supports projects that are less likely to happen in a business-as-usual scenario, and have the potential to advance the field as a whole. These include: developing open-source, production-quality tools; increasing the size, diversity, and quality of datasets required for training AI models; and increasing the ease-of-use of tools so that they are adopted by mathematicians.

The Fund’s first 29 grant awards will go to researchers at universities and organizations to support their efforts to develop systems that help advance mathematical discovery and research across several key tasks.

The inaugural AI For Math Fund portfolio comprises of researchers and universities and organizations developing systems that help advance mathematical discovery and covers a breadth of key topics and bottlenecks in AI for Math research, including:

• Key dataset development

• Improved prover - mathematician interaction

• Core functionality and coverage improvements

• Promising field-building initiatives

• High-ambition moonshots