Category: AI | Startups | Physical AI | Tech News
Published: June 13, 2026
Read time: 6 min

---

Jeff Bezos built Amazon from a garage. Then he built Blue Origin to reach space. Now, at 62, he is co-CEO of a startup with a goal that makes both of those look modest: creating an “artificial general engineer” — an AI system capable of accelerating human invention itself.

The startup is called Prometheus. On June 12, 2026, Bezos and co-founder Vik Bajaj revealed new details about what Prometheus is actually building — alongside the announcement of a massive $12 billion new funding round, bringing total funding to $18.2 billion and valuing the company at $41 billion.

With backing from JPMorgan Chase, Goldman Sachs, BlackRock, and Bezos himself, Prometheus is now one of the best-funded AI startups in history. Here is what they plan to do with it.

---

What Is Prometheus Actually Building?

When Bezos first announced Prometheus in November 2025, the company described its focus as “physical AI” — applying deep learning principles to robotics, manufacturing, and the physical world. Details, however, were scarce.

Now, Bezos has put a clearer label on the ambition: an artificial general engineer.

Speaking to The New York Times, Bezos framed the goal in historical terms. “All societal wealth is driven by invention. Six thousand years ago, somebody invented the plow, and we all got wealthier. Then, much later, somebody invented the steam engine, and we all got wealthier.”

His goal for Prometheus is to dramatically accelerate that invention cycle — offering tools that compress what used to take decades of human ingenuity into something far faster and more powerful.

In even bolder terms, Bezos told CNBC that the objective is to produce technological breakthroughs that generate “civilizational wealth” — not just returns for investors or a single company, but advances that benefit all of humanity.

---

What Is Physical AI — And Why Does It Matter?

Physical AI is one of the most important emerging terms in technology right now. It refers to applying the same deep learning techniques behind large language models and generative AI to the physical world — specifically to robotics, manufacturing, engineering, and autonomous systems.

Think of it this way:

Physical AI is the frontier that companies like NVIDIA (with its Isaac robotics platform), Figure AI, and now Prometheus are racing to define. The idea is that the same breakthroughs that made AI remarkable at language and images can be applied to engineering and manufacturing — fundamentally changing how physical things are designed and built.

---

The “Artificial General Engineer” — What It Could Mean

Co-founder Vik Bajaj offered a more grounded explanation of what Prometheus is working toward.

Designing genuinely new technology, Bajaj said, “takes a thousand human minds creatively working together” and is “one of the most complex things we do as a species.” Moreover, he noted that the engineers behind those breakthroughs “use tools that really haven’t changed for decades.”

Prometheus wants to change that. The goal is to arm engineers with AI tools that allow them to reach design breakthroughs far more quickly than current processes allow — compressing years of iterative engineering work into dramatically shorter timelines.

An artificial general engineer, therefore, would not replace human engineers. Instead, it would act as a powerful AI collaborator — one capable of simulating, testing, and iterating on engineering designs at a speed and scale no human team could match.

---

The Funding — And What It Will Buy

The numbers behind Prometheus are striking:

The investors include some of the most significant names in global finance — JPMorgan Chase, Goldman Sachs, and BlackRock — alongside a substantial personal contribution from Bezos himself.

What is all that money for? Bezos was direct: compute. “One of the reasons we’ve had to raise a significant amount of funding is because what we’re doing is very compute-intensive and we need to create that data.”

Training physical AI systems requires enormous amounts of data — simulated and real — about how the physical world works. Generating that data and training models on it, demands computing infrastructure at a scale comparable to the largest foundation model training runs in existence.

---

The $100 Billion Investment Fund

Beyond what Prometheus is building internally, the company’s ambitions extend further. Earlier this year, The New York Times reported that Bezos and Bajaj are working to raise a $100 billion investment fund — separate from Prometheus itself — to back companies that could directly leverage and benefit from what Prometheus produces.

That fund could include investments in Bezos’ other ventures. Blue Origin, his space company, would be an obvious beneficiary of AI-accelerated engineering tools. However, the fund is designed to be broader — backing the entire ecosystem of companies that could apply Prometheus’s physical AI capabilities.

This structure — building the foundational AI technology while simultaneously funding the companies that will use it — mirrors the approach Anthropic has taken with its Claude models and enterprise partnerships. It is a way to both build and monetize a platform without being limited to a single product or market.

---

The Competition — Prometheus Is Not Alone

Prometheus enters a crowded and well-funded field. Numerous startups are already exploring physical AI applications:

• Figure AI — humanoid robots powered by AI, backed by OpenAI and Microsoft
• Physical Intelligence (Pi) — training general-purpose robot policies
• 1X Technologies — AI-powered humanoid robots for industrial use
• NVIDIA Isaac — physical AI platform for robotics simulation and deployment

Furthermore, established players like Boston Dynamics, ABB, and Fanuc are integrating AI deeply into their robotics and manufacturing platforms.

However, what Prometheus has that most competitors do not is funding at a scale that allows it to generate the massive proprietary datasets that physical AI requires. With $18.2 billion raised and a $100 billion investment fund in progress, Prometheus has a significant financial advantage over virtually every other player in the space.

---

What This Means for the AI Industry

Prometheus’s emergence adds another dimension to an already remarkable moment in AI development. Consider what is happening simultaneously:

• Google is building agentic AI that operates across digital workflows
• NVIDIA is putting AI computing power into personal devices
• SpaceX is pitching orbital AI data centers
• Anthropic and OpenAI are racing toward AGI in language and reasoning
• And now Bezos is pursuing an artificial general intelligence for the physical world

Each of these efforts targets a different frontier of what AI can do. Together, they represent a comprehensive push to apply AI to every dimension of human activity — digital, physical, and beyond.

For businesses and professionals, the Prometheus story is a reminder that physical AI is not a distant future concept. It is being actively funded, built, and deployed right now — and the companies that understand it earliest will have a significant advantage as it matures.

---

Bottom Line

Jeff Bezos has a track record of making audacious bets that seem implausible until they are inevitable. Amazon, AWS, Blue Origin — each appeared over-ambitious at the time and ended up reshaping entire industries.

Prometheus may be his most ambitious bet yet. An artificial general engineer that accelerates human invention itself — backed by $18 billion and the engineering talent to pursue it — is not a product. It is an attempt to change how humanity creates things.

Whether Prometheus delivers on that vision in five years or twenty, the direction it represents is clear. Physical AI is the next major frontier — and the race to define it has officially begun.