Author: Deep Tide TechFlow

On November 17, 2025, 61-year-old Jeff Bezos became the CEO of a company again. This is his first return to a leadership position since stepping down from Amazon in 2021. The new company is called Project Prometheus, with a startup fund of $6.2 billion, primarily focused on "physical AI," targeting the manufacturing sector.

Seven months ago, 70-year-old Eric Schmidt took control of a rocket company called Relativity Space and assumed the role of CEO. He did not explain why he decided to step back into the fray at this age; perhaps the unspoken answer is that "every day in the AI era is crucial."

In June of the same year, Laurene Powell Jobs, the widow of Steve Jobs, gave a rare public interview. She sat next to Jony Ive, discussing the prototype she saw at the io company. It was a device described as an "AI device" acquired by OpenAI for $6.4 billion in stock, with no screen, and its shape reportedly resembled a player worn around the neck. Her assessment of the prototype was, "Watching an idea turn into reality is an incredible thing."

Three individuals, three different postures. Yet they are all placing bets in the same casino.

In the past three years, the top investors in Silicon Valley have been doing the same thing: pouring money from family offices, venture capital, and charitable foundations into AI. Schmidt, Bezos, and Powell Jobs are just the three most prominent figures. But if you carefully break down their target lists, you will find that this is not the same game; they are betting on three completely different futures.

Schmidt: Treating AI as the Next Cold War

According to data referenced by Wikipedia and The AI Insider, Schmidt's family office Hillspire has invested in over 22 AI companies since 2019, totaling over $5 billion. The list includes Anthropic, SandboxAQ (a quantum + AI company spun out from Alphabet), Inworld AI, Holistic AI, and Altera. These are targets that "industry insiders" would list.

But what truly exposes his underlying intentions is another list.

White Stork: A company producing AI drones in Ukraine. Rebellion Defense: Defense AI. Istari: Simulation. Swift Beat: Military software. This is a family office treating AI as the next generation of military equipment.

Since 2016, Schmidt has served as the chairman of the Defense Innovation Board and co-led the National Artificial Intelligence Security Commission from 2019 to 2021. He is a player who handles AI policy, defense procurement, and energy infrastructure as one and the same. In January 2024, Forbes reported that he launched the drone project White Stork simultaneously in the U.S. and Ukraine, viewing the Ukrainian battlefield as a "laboratory for AI weapons."

Then there's infrastructure.

In January 2026, he co-founded a company called Bolt Data & Energy with Texas Pacific Land, becoming its chairman. This company does not rent data centers or buy power grid electricity; it plans to build its own natural gas power plant in the West Texas wilderness, directly supplying power to data centers. The plan is to reach 1 gigawatt initially, eventually escalating to 10 gigawatts, equivalent to the electricity consumption of 7 million households. Texas Pacific Land invested $50 million, also including priority rights for water resources. In a response to Fortune, Schmidt stated, "The biggest bottleneck faced by AI is not the algorithm, but energy."

In March of the same year, he took control of Relativity Space. This company is developing a reusable rocket called Terran R, aiming to disrupt SpaceX's monopoly on mid-to-low orbit launches, with orders amounting to $2.9 billion at that time.

Putting all these pieces together, the logic becomes quite clear.

Schmidt does not believe in the strategy of "betting on a basket of large model companies." He believes that the outcome of AI will ultimately depend on three things, namely computing power (data centers and electricity), transportation (rockets, satellites, drones), and policy (defense boards and congressional hearings). He also invests in model companies, and after the release of DeepSeek he publicly called for the U.S. to increase investments in open-source, but that is merely a chess piece on his board, not the entirety.

His reaction to DeepSeek is quite revealing. After DeepSeek's release in early 2025, Schmidt immediately wrote an article in The Washington Post stating that it was a "turning point in the global AI race." His prescription was not to retreat but to double down, including more open-source investments, more Stargate-like infrastructure, and greater sharing of training methods among model laboratories.

In other words, he views AI as a marathon between nations, and he himself is already standing on the sideline, also serving as a member of the organizing committee. At 70, returning to serve as CEO of Relativity may seem like turbulence to outsiders, but he explains: working at 100, like Kissinger, "requires responsibility during a period of major change, not withdrawal."

Bezos: A Full-Stack Control Freak

Bezos's approach is completely different from Schmidt's.

According to data from StartupHub citing TechCrunch, The Information, and Bezos Earth Fund, by mid-2026, Bezos had deployed over $19 billion in AI. This figure is still rising.

Breaking it down, the funds can be divided into three main parts.

The first part is Anthropic. Amazon started investing $8 billion in September 2023, and in April 2026 committed to an additional maximum of $25 billion. Anthropic runs on AWS using Amazon's Trainium chip. This binds Amazon's cloud infrastructure, Bezos's model layer, and Anthropic's research capacity into a triangle — it is not merely a financial investment. When Anthropic's valuation soared above $60 billion, Amazon had already captured the largest piece of the external cake.

The second part consists of scatter investments from Bezos Expeditions. Bezos Expeditions is raising a multi-billion dollar dedicated AI fund, upgrading "Bezos's personal angel" into "institutional investor." Among its investments, Perplexity, an AI search company, saw its valuation rise from $520 million in January 2024 to $20 billion in September 2025.

The third part is Project Prometheus.

In November 2025, Bezos, together with former Google X executive Vik Bajaj, announced the founding of this company, with a startup fund of $6.2 billion, a team of nearly 100 members, with team members coming from OpenAI, DeepMind, and Meta; the founding advisor list includes Ashish Vaswani and Jakob Uszkoreit, the authors of the 2017 paper "Attention Is All You Need." The company's goal is to apply AI in manufacturing, including automobiles, aerospace, and chips.

Why manufacturing? Because it coincides perfectly with Bezos's other businesses. Amazon has the Kuiper satellite constellation, and the first batch of customers after the implementation of manufacturing AI will be right at home.

Musk referred to Project Prometheus on X as a "copycat."

But structurally, this is not an act of plagiarism.

Bezos holds the model layer through Anthropic, the application layer through Perplexity and Figure, and the computing power layer through Amazon. Now, he is creating another Prometheus, integrating AI into manufacturing, also capturing the "execution layer of the physical world." This is a full-stack approach; from training chips to deployment in factory workshops, every layer has its own card.

About ten days after launching Project Prometheus, it quietly acquired a company called General Agents. This company specializes in "computer agents" that can directly operate entire computers through AI. WIRED later revealed that this acquisition was completed just four days after inception.

Harsha Abegunasekara, CEO of Donely, commented, "General Agents genuinely solves the speed problem; Ace runs on your computer almost instantaneously." His company was originally a competitor of General Agents.

From angel investing to building a dedicated fund, to personally assuming the role of CEO, Bezos accomplished all this in just 18 months. He is essentially constructing a system larger than Amazon.

Powell Jobs: The Low-Key Faction

When looking at these three individuals together, Powell Jobs appears to be the least like an "AI investor."

According to CNBC, citing private wealth data platform Fintrx, her family office, Emerson Collective, has invested in at least 9 AI-related startups since 2022, with total financing rounds exceeding $1 billion. This figure is not on the same scale as Schmidt or Bezos.

But what's interesting is the list itself.

Proximie: Remote surgery connection platform; Atropos Health: Clinical data AI; Formation Bio: AI pharmaceuticals; Curipod: A Norwegian AI teaching tool; Mistral: A large model company from France, the only player in Europe competing against OpenAI.

No defense, data centers, or rockets.

Emerson Collective's official website clearly states its investment direction: education, energy and environment, digital health, financial technology, and media. AI is merely an interwoven tool among these themes. She holds a majority stake in The Atlantic and is very skilled in Colombian-style "soft power" investments.

However, her most significant investment is not aligned with the previous ones.

After Jony Ive left Apple in 2019, Powell Jobs invested in his design company LoveFrom through Emerson Collective. Ive later stated in an interview with Financial Times, "Without Laurene, there wouldn't be a LoveFrom." A few years later, Ive founded a hardware company called io, focusing solely on AI devices, and Powell Jobs invested once again. In May 2025, OpenAI acquired io with a full stock transaction worth $6.4 billion, making Ive a billionaire on paper. Emerson Collective also benefited from this.

Another critical investment is that Emerson Collective was one of the early investors in Mistral AI. At that time, this French company was the remaining spark in Europe regarding large models.

When piecing these together, her AI bets are concentrated in two directions: either "using AI to solve specific human problems" or "reshaping human-machine interactions" (the io device, Ive's design).

VC Sheet described Emerson Collective in an assessment as: "a deliberately opaque LLC that combines venture capital, philanthropy, policy advocacy, art and media ownership, utilizing any of the most effective tools for grants, policy lobbying, or investments."

Philosophically, she is closer to the old-generation East Coast family offices, where influence is more important than returns, long-term is more critical than short-term, and the microphone is more important than the spotlight.

Three Investment Philosophies

When placing these three lists together, you will see three sets of judgments about the future of AI.

Schmidt bets on national competition and infrastructure bottlenecks. In his world, AI will ultimately be determined by "who has the most electricity, who has the fastest rockets, who has the strongest drones." Models are just admission tickets; the real moat lies in the physical layer. Therefore, he personally runs Relativity and Bolt; what he seeks is not returns but control.

Bezos bets on industrial revolution-level application diffusion. He believes AI will eventually permeate every machine tool, every aircraft, every satellite like electricity. Thus, he locks in the model layer through Amazon, the manufacturing layer through Prometheus, and embeds the consumer application layer through Expeditions. He isn’t betting on whether a specific company can win, but on whether "the whole structure can win."

Powell Jobs, on the other hand, is betting on something else. She is wagering that people will ultimately be unable to tolerate the current model of human-machine interaction. She and Ive repeatedly emphasized in their interview with Financial Times that "humans deserve better." Her investments in io, LoveFrom, medical AI, and educational AI stem from the same judgment: the largest market in the next decade will be to "fix the side effects caused by the internet over the previous decade."

Three sets of judgments, three different strategies.

Which one is correct? No one knows. Schmidt may overestimate the weight of geopolitics in the AI economy. Bezos may underestimate the capital consumption of the "full-stack" heavy asset model; a typical example is that Prometheus is rumored to need another $10 billion before its first delivery. Powell Jobs faces an even more awkward question: the io device won’t start mass production until 2027, while OpenAI’s own financial model has been repeatedly questioned by the market.

But one thing is certain. As the winners of the last generation of the internet collectively turn their family funds toward AI, this is no longer a small breeze in a certain track. Bolt has already raised $150 million in startup funds, and Anthropic alone has consumed Amazon's promise of $33 billion. This scale of capital flow will shape the industry geography of the next decade.

As for who will ultimately succeed, we will have to wait until 2030 to look back. Until then, all three veterans are still at the table, and the stakes are still rising.