OpenAI isn’t content with just dominating software. President Greg Brockman confirmed in a new interview that the company is actively developing a “family of devices” purpose-built for interacting with its AI models. He teased a timeline only by saying “you can expect them soon,” but was otherwise light on specifics. He didn’t confirm the persistent rumor of a smart speaker slated for 2027 or walk back earlier buzz about a Humane-like AI wearable, leaving the exact form factors a genuine mystery.
Brockman’s vision is a fundamental shift in computing, where we talk to our machines for the “vast majority of what we do” instead of typing. He acknowledged the obvious, awkward social hurdle of an open-plan office filling up with people chatting to chatbots, pointing to the emergence of soundproof “beak” masks as a clear, if inelegant, signal of product-market fit for voice-first AI. “It really shows you where there’s product market fit,” he noted, even while admitting the masks aren’t a great final solution.
The conversation also touched on the elephant in the room: trust. With Apple embroiled in a lawsuit and questions swirling about data privacy, Brockman was blunt about drawing a hard line. “We have no interest in other companies’ trade secrets. We are plenty innovative,” he said, pushing back on the idea that a collaboration with former Apple designer Jony Ive could be legally fraught. He also said OpenAI is pursuing “verifiable, auditable guarantees” that would cryptographically prove only an AI, and not a human, can review private conversations. It’s a technical promise that’s far easier to announce than to engineer.
This is a classic OpenAI power play — signal a hardware revolution without committing to a product. The real test isn’t the device’s design, which Ive will likely nail, but whether the company can solve the deep-seated privacy and behavioral friction that make talking to a computer feel less like the future and more like a privacy nightmare waiting to happen.
