Periodic Labs, co-founded by Liam Fedus and Ekin Dogus Cubuk, has successfully raised $300 million in seed funding to innovate in the field of material science using artificial intelligence. The funding round, led by Felicis, attracted numerous high-profile investors, reflecting a strong belief in the potential of AI to transform scientific research. The startup aims to automate the discovery of new materials by combining advanced machine learning simulations with robotic experimentation. This approach is expected to generate valuable data, even from failed experiments, thereby enhancing the AI training process. The founders emphasize the importance of integrating real-world experiments into the AI loop, which could shift the focus of scientific research from merely publishing successful results to exploring new hypotheses. With a talented team and a well-equipped lab, Periodic Labs is initially targeting the discovery of new superconductors, which could have significant implications for energy-efficient technologies.