DooDooLamb News
Dario Amodei expresses discomfort with the ‘overnight’ and accidental concentration of power in AI - Fortune
Brief published February 26, 2026 · Original source published February 25, 2026
Original reporting by feedfeeder at slashdot.org.
Automated brief. Verify important details at the original source.
Anthropic's CEO Has a Case of the Guilt
The man who helped create one of the world's most powerful AI companies is having second thoughts about the whole "accidentally becoming an AI overlord" thing.
Dario Amodei, CEO of Anthropic, just went on record saying he's uncomfortable with how quickly power has concentrated in the AI space. This is coming from someone whose company builds Claude, one of the few AI systems that can hold its own against OpenAI's GPT models. It's a bit like Jeff Bezos in 2010 wondering if maybe Amazon got too big too fast.
Amodei's concern isn't just philosophical hand-wringing. He's warning that AI could mint "trillionaires" and that if the public gets shut out of the benefits, "you're going to get a mob coming for you." The subtext here is pretty clear: we're watching the formation of a new tech aristocracy, and it happened so fast that even the people building it are getting whiplash.
The "overnight" part of his complaint is particularly rich. Anthropic was founded in 2021 by former OpenAI researchers who left because they wanted to build "safe" AI. Three years later, they've raised billions, built world-class models, and landed partnerships with Amazon and Google. That's not overnight success, that's warp speed in an industry that usually measures progress in decades. When your safety-focused AI startup accidentally becomes a major player in the most important technology race of our time, you might reasonably ask how the hell that happened.
The power concentration Amodei is talking about isn't just about money, though the money is staggering. Training state-of-the-art AI models now costs hundreds of millions of dollars and requires computational resources that only a handful of companies can afford. This creates a natural oligopoly where the rich get richer and everyone else gets to use whatever APIs the big players decide to offer. It's like the early days of the internet, except instead of needing a server farm, you need the GDP of a small country.
What makes this particularly uncomfortable is how accidental it all feels. Nobody sat down and planned for three companies to control the future of artificial intelligence. It just sort of happened because the technology advanced faster than anyone expected, the capital requirements grew exponentially, and the best talent kept clustering around the same handful of labs. The result is that we're having conversations about artificial general intelligence and its implications for humanity, while the people building it are basically winging it.
The "mob" comment isn't hyperbole either. We've already seen what happens when new technologies create massive wealth while leaving everyone else behind. Social media platforms made their founders billionaires while arguably making society more miserable. Cryptocurrency created a new class of extremely online millionaires while accomplishing very little for normal people. AI has the potential to be both more transformative and more destabilizing than either of those.
Amodei's discomfort might be genuine, but it's also convenient. Expressing concern about power concentration is a good way to position yourself as the thoughtful, responsible AI company while continuing to accumulate that same power. Anthropic still takes the funding, still builds more capable models, and still competes aggressively for talent and partnerships. The guilt doesn't seem to be slowing them down much.
The real question is whether this concentration was ever avoidable. Building cutting-edge AI requires resources that naturally consolidate power, and the competitive dynamics pretty much guarantee that a few players will dominate. Amodei can feel bad about it, but feeling bad doesn't change the math.
At least he's honest about the awkwardness of the situation. Most tech CEOs would just talk about democratizing AI while quietly building moats around their trillion-dollar opportunities.