AI가 스스로 만든 열 문제를 해결하다

discovered materials emerged from y combinator with $9m in funding to solve ai's thermal constraints by automating materials science. founders advaith sridhar and akash ramdas built a pipeline using anthropic models to generate leads and proprietary physics simulations to validate candidates, iterating thousands of times daily. the firm aims to bypass the rare-earth supply chain and license patents for heat-efficient materials to gpu manufacturers within a year. despite competition from sandboxaq and cuspai, discovered materials focuses on the 'whack-a-mole' trade-offs between thermal, electrical, and manufacturable properties. this signals a shift toward vertical ai applications solving physical bottlenecks, though the industry awaits the first commercial deployment of ai-discovered materials.