In 2025, former OpenAI CTO Mira Murati led a team of scientists and engineers to launch Thinking Machines Lab, born from a critical realization about the AI landscape: while AI capabilities had advanced dramatically, the scientific community's understanding of these frontier systems was lagging behind. Knowledge of how these systems are trained remained concentrated within top research labs, limiting both public discourse and people's ability to use AI effectively. Despite enormous potential, these systems remained difficult for users to customize to their specific needs and values.
The founders set out to bridge these gaps by building an organization dedicated to making AI systems more widely understood, customizable, and generally capable. With a team that had created some of the most widely used AI products including ChatGPT, Character.ai, and foundational open source projects like PyTorch, Thinking Machines Lab embraced open science through publications and code releases. They focused on human-AI collaboration rather than full autonomy, building multimodal systems that could adapt to the full spectrum of human expertise and enable a broader range of applications.