A look at OpenAI's interest in robotics, from investing in startups like Figure and Physical Intelligence to, sources say, considering making a humanoid robot
Over the past year, OpenAI has dropped not-so-subtle hints about its revived interest in robotics: investing in startups developing hardware …
Context & Ripple Effects
OpenAI’s robotics interest already included a partnership with Figure after Figure’s $675 million financing and OpenAI model-development tie-up. This report broadens the question from backing and enabling robotics companies to potentially building a system itself.
The later decision to rebuild OpenAI’s robotics research team makes this an early marker of a more sustained physical-AI push, rather than an isolated startup investment.
First-order effects
- OpenAI’s consideration of a humanoid robot puts its role in robotics under closer scrutiny: it could remain a model and technology partner, or move toward a finished product.
- Figure and Physical Intelligence gain validation from OpenAI’s continued attention to robotics, while also facing more uncertainty about how closely their interests will align with a potential in-house effort.
Second-order effects
- Humanoid-robot developers and their investors will have to distinguish between partnerships that provide AI capabilities and relationships that could eventually create platform-level competition.
- A move toward building hardware would require OpenAI to confront integration, manufacturing, and deployment questions that are largely separate from supplying models to robot makers.
Third-order effects
- The story points to an emerging split in physical AI: labs may either sell intelligence to a broad hardware ecosystem or vertically integrate into robots themselves; the latter path could concentrate more of the stack inside a few AI companies.
- Whether that consolidation occurs will depend on whether general-purpose models translate into reliable real-world operation, a question that continues to underpin the sector’s investment enthusiasm.
The trend: Generative-AI labs are extending their ambitions from digital interfaces into physical systems, forcing a choice between enabling robot makers and becoming one.