Laude Institute AI Grants Launch

The Laude Institute has launched its inaugural Slingshots AI grants program, designed to boost both AI science and its practical applications. This initiative offers researchers resources often unavailable in academic settings, such as financial aid, computing power, and engineering support, to help turn research into tangible outcomes like startups or open-source projects. The first cohort includes 15 projects, many focused on AI evaluation, including familiar names like Terminal Bench and ARC-AGI. Other projects are developing new approaches to existing evaluation challenges. Formula Code, from CalTech and UT Austin, measures how well AI agents improve existing code, while Columbia's BizBench introduces a benchmark for 'white-collar AI agents'. Additional grants support innovative methods for reinforcement learning and model compression. SWE-Bench co-founder John Boda Yang is also participating, leading the CodeClash initiative, which evaluates code using a dynamic, competition-driven model.