Jeff Bezos stepped in front of cameras Thursday to talk about Prometheus, the artificial intelligence startup he co-leads, after the company announced a $12 billion funding round that values the venture at $41 billion. The former Amazon CEO, who left that post in 2021, sat down with CNBC's David Faber at Prometheus headquarters in San Francisco alongside co-CEO Vik Bajaj, and insisted the company is not being "secretive."
That claim deserves scrutiny. Prometheus launched in November with $6.2 billion in initial funding. Since then, it has remained tight-lipped about its products, disclosed little about what it has built, and now nearly tripled its total capital base in a single raise. Bezos himself called it "premature" to reveal what the company has accomplished so far.
So what, exactly, is $18.2 billion in combined funding buying? Bezos says Prometheus is developing AI tools to help engineers design and manufacture physical products faster and easier. The company is focused on building AI models for physical tasks, engineering, manufacturing, and drug design. And the latest round, Bezos explained, will go largely toward acquiring more computing power.
"That is a big chunk of the funding we've raised. And 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, you know, create that data."
That is a staggering amount of capital for a company that has not publicly demonstrated a product. Investors are being asked to trust the Bezos name, the Bajaj pedigree, and the promise that something "really quite remarkable" is taking shape behind closed doors.
Bajaj brings serious credentials. He is a professor at Stanford University's School of Medicine and previously co-founded Alphabet's Verily, a research lab dedicated to life sciences. The pairing of Bajaj's scientific background with Bezos's operational record is the pitch: AI that moves beyond chatbots and into the physical world.
Prometheus has been recruiting aggressively from the top of the AI talent pool, pulling engineers from OpenAI, Google DeepMind, and Nvidia. That hiring pattern signals ambition. It also signals cost. The best AI researchers command enormous compensation packages, and the compute infrastructure they need does not come cheap.
The broader question for anyone watching the AI sector is whether this level of spending produces proportional value, or whether it represents the kind of speculative excess that has defined previous tech cycles. Big Tech's combined AI spending spree has already drawn warnings from analysts who see echoes of past bubbles.
Consider the math. Prometheus launched roughly seven months ago. It has disclosed no revenue figures, no customer base, no product demonstrations, and no timeline for commercial availability. Yet it now commands a valuation higher than most publicly traded industrial companies.
Bezos told Faber the results so far are "really quite remarkable." But he would not say what those results are. The public is left with a paradox: the richest man in the world asks for trust while withholding the evidence that would earn it.
That dynamic is not unique to Bezos. The AI industry broadly has operated on a "trust us, the breakthrough is coming" model that has attracted hundreds of billions in capital with limited accountability for outcomes. What makes Prometheus different is the sheer concentration of money and reputation in a single private company with no public product.
Bezos, for his part, has also weighed in on public policy questions beyond AI. He recently backed a proposal to make income under $75,000 tax-free, a position that drew attention given his own vast fortune.
Bezos pushed back on the characterization that Prometheus has been operating in secrecy. But the facts tell a different story. The company launched with billions in funding and immediately went quiet. It has not released demos, whitepapers, or technical benchmarks. Its co-CEO describes current progress as "premature" to discuss.
In Silicon Valley, that pattern is familiar. Companies raise at enormous valuations, hire top talent, burn through capital on compute and research, and delay public accountability until they either succeed spectacularly or fail quietly. The difference here is scale. Twelve billion dollars is not a seed round. It is a bet that would rank among the largest private funding events in history.
The company's focus on physical-world AI, engineering, manufacturing, drug design, is genuinely distinct from the large language model race dominated by OpenAI and others. If Prometheus can deliver tools that accelerate how physical products are designed and built, the economic implications would be substantial. That is a real and worthy ambition.
But ambition is not execution. And $41 billion in implied value demands more than a co-CEO saying the work is "really quite remarkable" while declining to show it.
Bezos also touched on AI regulation during the interview, saying "reasonable" rules are needed. He did not elaborate on what specific regulations he would support or oppose. That vagueness is itself telling. Tech founders routinely call for "reasonable" regulation in the abstract while lobbying against specific proposals that would constrain their operations.
The question of how AI companies should be regulated is one of the defining policy debates of the moment. Prometheus, with its focus on manufacturing and drug design, would face a different regulatory landscape than a chatbot company. Physical-world AI touches FDA oversight, industrial safety standards, and intellectual property law in ways that pure software does not.
Bezos's San Francisco headquarters also places Prometheus squarely in a state that has driven many wealthy founders and businesses toward lower-tax states. Whether Prometheus stays rooted in California as it scales remains an open question, one that other tech leaders, including Uber co-founder Travis Kalanick, have already answered by relocating.
Several critical questions remain unanswered. Who led the $12 billion funding round? Which investors participated? What specific products is Prometheus building, and when will they reach market? What milestones has the company hit internally that justify the "really quite remarkable" label?
None of these were addressed in the interview. Bezos and Bajaj offered broad strokes, AI for physical tasks, compute-intensive work, aggressive hiring, without the specifics that would let outside observers evaluate whether the valuation makes sense.
CNBC also flagged a potential connection between Prometheus and Amazon, noting that Amazon "could work with Prometheus." No details were provided. If Amazon becomes a customer or partner, it would raise obvious questions about self-dealing and conflicts of interest, given that Bezos remains Amazon's largest individual shareholder and executive chairman.
For now, the public knows this: Jeff Bezos has raised $18.2 billion for a company that won't show what it has built, won't say when it will, and insists it is not being secretive. The market has decided to take him at his word. Whether that faith is rewarded or punished will say a great deal, not just about Prometheus, but about whether the AI investment boom is building something real or simply burning capital on promises.
Twelve billion dollars buys a lot of compute. It should also buy the public a straight answer about what the money is for.