The Quiet Power Behind the Lab
In the world of high-stakes AI, we spend a lot of time talking about who has the best model or the most compute. But for founders, the real story is usually found on the cap table. A recent report highlighting Eric Schmidt’s ownership in Anthropic is a textbook example of how the old guard of Silicon Valley maintains a grip on the new guard of intelligence.
Schmidt, the former CEO of Google, reportedly owns roughly 1% of Anthropic. To some, that sounds small. In the context of a company valued in the tens of billions, it is a massive piece of leverage. More importantly, it is about half of what CEO Dario Amodei reportedly holds. When a single outside investor holds that much weight relative to the founder, the power dynamics shift in ways that builders need to pay attention to.
The Value of Early Access
Why does this matter to someone building a startup today? It shows that access still beats talent in the early rounds. Schmidt wasn't just another check; he was a strategic signal. In the early days of Anthropic, having a veteran like Schmidt on the cap table provided a level of institutional credibility that even the smartest ex-OpenAI engineers couldn't manufacture on their own.
For the builder, this is a double-edged sword. You want the big names because they open doors to enterprise contracts and hardware pipelines. But you also have to realize that these individuals are playing a different game. They are diversifying their influence. Schmidt isn't just betting on Anthropic; he is betting against his former employer, Google, while simultaneously sitting on boards and committees that influence national AI policy.
The Dilution Reality Check
Let’s talk about the math of the 1% versus the founder’s stake. Building a foundation model is perhaps the most capital-intensive endeavor in human history. To compete with GPT-4 or Gemini, Anthropic has had to raise billions from Amazon and Google. Every time a new billion dollars flows in, the founders’ slice of the pie gets smaller.
Dario Amodei holding roughly double what Schmidt holds is a testament to how quickly founder equity evaporates in the AI sector. Most founders in traditional SaaS expect to own a much larger chunk by the time they hit these valuations. In AI, you are essentially trading your ownership for the electricity and H100s required to stay in the race. If you are a founder looking at this, you have to ask yourself: are you building a company, or are you building an R&D lab for the cloud providers?
The Strategic Skepticism
I’ve always been skeptical of the "alignment" narrative when it comes to the business side of these labs. Anthropic positions itself as the safe, ethical alternative to OpenAI. Yet, their cap table is a who’s who of traditional tech power. When you have the former head of Google owning a significant stake, the "public benefit corporation" status starts to feel like a very thin layer of paint on a very traditional corporate structure.
Investors like Schmidt aren't there for the altruism. They are there because they know that whoever controls the weights of the most efficient models will control the next twenty years of software. For builders using Anthropic’s API, you need to realize that you are building on a platform that is increasingly influenced by the same people who built the ad-tech monopolies of the 2000s.
What Builders Should Take Away
- Equity is a weapon: Use it to bring in strategic partners, but realize that every percentage point given to a legacy tech titan is a vote for the status quo.
- The Compute Trap: Anthropic’s dilution shows that if your business model relies on massive compute, you will likely lose control of your company faster than you think.
- Watch the exits: When people like Schmidt start leaking or reporting their stakes, it’s often a sign that the private market is looking for liquidity or a new round of hype to justify the current valuation.
We are seeing a consolidation of power disguised as a revolution. The players haven't changed; they’ve just moved to a different table. If you're a founder, don't get blinded by the celebrity of your investors. Look at the percentage, look at the control, and ask yourself if you're actually the one in charge.
The goal of a founder is to build something that lasts. The goal of a legacy investor is to ensure that whatever comes next still looks a lot like them.
The lesson here isn't that Schmidt is a genius for buying in early—everyone knew Anthropic was a powerhouse. The lesson is that even the most innovative AI labs are eventually tethered to the old money and old minds of Silicon Valley. If you want to build something truly different, you have to find a way to grow without selling half your soul to the people you’re trying to disrupt.
Read the original at TechCrunch Startups →