On Friday, AI startup OpenAI submitted a new trademark application concerning products tied to its brand—“OpenAI”—to the U.S. Patent and Trademark Office (USPTO). While such filings are common and usually don’t attract attention, OpenAI’s application hints at the development of both immediate and more visionary product lines.
The application specifies a range of hardware, including headphones, goggles, glasses, remotes, cases for laptops and phones, smartwatches, smart jewelry, as well as virtual and augmented reality headsets, all marked for “AI-assisted interaction, simulation, and training.”
Previously, OpenAI acknowledged its collaboration with former Apple designer Jony Ive on a hardware initiative. Over the weekend, CEO Sam Altman informed Korean news outlet The Elec that the aim is to develop AI-enhanced consumer hardware “through collaborations with various companies.” However, he also observed that creating even a prototype AI device may take “several years.”
The trademark application also alludes to robots, specifically “user-programmable humanoid robots” and “humanoid robots equipped with communication and learning capabilities for assisting and entertaining individuals.”
Additionally, OpenAI has recently started recruiting for a new robotics division under Caitlin Kalinowski, who joined the firm last November from Meta’s AR glasses division. As detailed in job postings and reports from The Information, OpenAI seeks to experiment with robots—potentially humanoid—that utilize customized sensors and AI to operate with human-like intelligence in real-world scenarios.
Further into its application, OpenAI mentions custom AI chips and services aimed at “utilizing quantum computing resources to enhance AI model performance.”
There have long been speculations regarding OpenAI’s development of custom chips for running its AI models. The organization has a division dedicated to co-designing chip components, and reports indicate that OpenAI plans to launch a custom chip in partnership with semiconductor giants Broadcom and TSMC as soon as 2026.
OpenAI’s ambitions in quantum computing remain somewhat unclear. However, last year, the startup welcomed a former quantum systems architect from PsiQuantum to strengthen its technical team.
As highlighted by The Register in a March article, quantum computing holds significant promise for considerably enhancing the efficiency of AI model training due to its capability to perform extensive calculations concurrently. With the growing costs associated with AI computing, it seems OpenAI may be envisioning a future where models are trained utilizing hardware that significantly diverges from its current setups.
It is important to note that trademark applications are often crafted broadly and may not accurately reflect a company’s definitive product strategy. OpenAI’s application likely showcases areas the startup is contemplating or has at least briefly considered. However, the timeline for when—if ever—any of the technologies referenced will reach the market remains uncertain.
TechCrunch offers a newsletter focused on AI! Subscribe here to receive it every Wednesday.
Compiled by Techarena.au.
Fanpage: TechArena.au
Watch more about AI – Artificial Intelligence
