“After a period of disillusionment where many researchers lost faith in Deep Reinforcement Learning, the success of O1 demonstrates that there is a powerful place for it when combined with other elements.”