Thinking Machines, an AI lab founded by Mira Murati, has published a mission statement arguing that concentrating AI power in a small number of labs is a genuine danger, framing its argument around the concept of "distribution" of knowledge, values, and ownership. The company applies economist Friedrich Hayek’s logic against central planning to AI, contending that a model trained once in a lab and shipped everywhere flattens expertise into "a standard offering." Thinking Machines wants organizations to own and fine-tune their own weights instead of leasing intelligence from APIs, which is the premise of its fine-tuning product Tinker, launched in October 2025. The post argues that most flagship models inherit the temperament of their predecessor, stating, "A single alignment spec suppresses creativity and diversity," and proposes model weights that organizations can shape according to their own values. The company’s reasoning runs in contrast to Geoffrey Hinton’s warnings about open-sourcing powerful models, asserting that leaving all frontier capability in closed, centralized hands creates the real risk.