• Source:JND

Sam Altman’s aggressive push to stay ahead in the AI race may have stretched far beyond model development. According to a Wall Street Journal report, the OpenAI CEO held advanced talks earlier this year to either acquire or form a deep partnership with Stoke Space — a fast-growing rocket manufacturer — in what would have been a dramatic expansion into orbital infrastructure.

A Serious Bid That Didn’t Materialise

Sources cited in the report say the discussions went as far as exploring a multi-billion-dollar deal that would give Altman a controlling stake in Stoke Space. In the end, both sides walked away without a final agreement, but the talks reportedly underscored how seriously Altman views space as part of AI’s long-term roadmap.

ALSO READ: Apple Watch Hypertension Alerts Now Launched In India: A Major Step In Everyday Heart Health

Stoke Space, founded by former Blue Origin engineers, focuses on fully reusable rockets. The company has been viewed as a potential challenger to SpaceX thanks to its engine technology and rapid development pace. Had the deal gone through, OpenAI would have secured a ready-made entry into the orbital launch business.

Why Altman Was Interested

Altman has repeatedly floated the idea that future AI infrastructure may need to move beyond Earth. His reasoning revolves around two pressures: the exploding energy demands of AI training and the need for more efficient cooling and power distribution.

Placing data centres in orbit is still a theoretical bet, but Altman has already backed startups pursuing unconventional space technologies. One of them is Longshot Space, which is building new launch systems designed to send satellites into orbit without conventional rocket combustion.

If Stoke Space had been added to that portfolio, it would have positioned OpenAI much closer to the tools required to one day build orbital computing infrastructure.

ALSO READ: WhatsApp iOS Feature Update: WhatsApp Adds Quick Voice Message Option After Missed Calls On iPhone

OpenAI is Racing Forward on the Ground

Even without Stoke Space, OpenAI is in the middle of what Altman reportedly called a “code red” push to outpace rivals. Internal reporting suggests the company is testing a new large model codenamed Garlic — positioned as a direct answer to Google’s Gemini 3 and Anthropic’s Opus 4.5.

Early internal benchmarks point to strong reasoning and coding performance, and the model could arrive publicly as GPT-5.2 or GPT-5.5 sometime in 2026.

While the Stoke Space talks have ended, they reveal the scale of Altman’s ambitions. The competition isn’t just about better chatbots anymore — it’s about building the infrastructure for AI systems that may demand more power and resources than Earth can comfortably provide.

Also In News