Anthropic CEO Dario Amodei has put a stark timeline on the debate over advanced AI. Speaking at Journal House in Davos, Switzerland, during the World Economic Forum, he said AI models may move beyond human capabilities “in almost everything” within two to three years, according to a Wall Street Journal interview.
The point was not just that AI systems may become more capable. Amodei connected that possibility to a larger social question: what happens when machines can perform not only digital tasks, but eventually physical work through advanced robotics.
A near-term prediction from Anthropic’s chief
Amodei did not present 2027 as a fixed deadline. He said, “I don’t know exactly when it’ll come, I don’t know if it’ll be 2027. I think it’s plausible it could be longer than that. I don’t think it will be a whole bunch longer than that when AI systems are better than humans at almost everything. Better than almost all humans at almost everything. And then eventually better than all humans at everything, even robotics.”
That wording matters. The prediction is broad, but it is also cautious about the exact year. Amodei is saying that the arrival of systems exceeding almost all human performance across almost everything could be close, while leaving room for uncertainty around timing.
He also described the speed of the field directly: “I think progress really is as fast as people think it is.” For a company building frontier AI models, that is a consequential statement. It frames current AI development not as a distant research story, but as a change that could soon affect how people think about work, capability and economic organization.
Why robotics changes the stakes
The most disruptive version of Amodei’s forecast is not limited to software. During the WSJ interview, he discussed what could happen when highly capable AI systems are able to control advanced robotics.
His argument is that stronger AI could help create better robots. If that happens, the impact would move beyond screens, documents and code. It would reach the physical world, where many forms of human labor still depend on bodies, tools and movement.
“[If] we make good enough AI systems, they’ll enable us to make better robots. And so when that happens, we will need to have a conversation… at places like this event, about how do we organize our economy, right? How do humans find meaning?”
That question is the core of the issue. If AI and robotics together can replace large parts of human labor, the challenge is not only technical. It becomes economic and cultural. Work is not just a way to produce goods or services. For many people, it is also tied to identity, status, routine and self-worth.
Labor, value and meaning
Amodei described a future in which advanced systems create “huge abundance and huge economic value,” while also challenging the idea that distributing that value should depend on people producing economic labor.
He said, “We’ve recognized that we’ve reached the point as a technological civilization where the idea, there’s huge abundance and huge economic value, but the idea that the way to distribute that value is for humans to produce economic labor, and this is where they feel their sense of self worth,” adding, “Once that idea gets invalidated, we’re all going to have to sit down and figure it out.”
This is not simply a prediction about jobs disappearing. It is a warning that the current link between labor, income and personal value may become harder to sustain if AI systems can perform the work that humans once used to justify their economic role.
Based on Amodei’s comments, the questions ahead include:
- How should economic value be organized if human labor is no longer the central input?
- What happens to self-worth when work is less necessary for production?
- How should society respond if AI models and robots can handle more tasks than almost all humans?
The source does not give answers to those questions. Amodei’s point is that they may become unavoidable if the capability curve continues at its current pace.
Anthropic’s position in the AI race
Amodei co-founded Anthropic in 2021 with his sister, Daniela Amodei, and five other former OpenAI employees. Since then, Anthropic has become a major competitor to OpenAI’s AI products, including GPT-4 and ChatGPT.
The company’s Claude 3.5 Sonnet model has remained highly regarded among some AI users and highly ranked among AI benchmarks. That gives Amodei’s forecast added weight: he is not speaking from the sidelines of the AI industry, but from inside one of its leading companies.
The comments also arrive while Anthropic negotiates a $2 billion funding round that would value the company at $60 billion. Amodei disclosed that Anthropic’s revenue multiplied tenfold in 2024.
Investment around Anthropic is also expanding. On Monday, Google announced an additional $1 billion investment in Anthropic, bringing its total commitment to $3 billion. Amazon has invested $8 billion over the past 18 months, and Amazon plans to integrate Claude models into future versions of its Alexa speaker.
Why Amodei avoids the AGI label
Even as he described systems that could outperform almost all humans at almost everything, Amodei distanced himself from the term “artificial general intelligence” (AGI). In a separate CNBC interview from the same event in Switzerland, he called AGI a marketing term.
Instead, he prefers the phrase “country of geniuses in a data center.” In an October 2024 essay, he wrote that such systems would need to be “smarter than a Nobel Prize winner across most relevant fields.”
That language is vivid, but it also shifts the focus. Rather than arguing over a label, Amodei is pointing to practical capability: AI systems that can reason, create, solve problems and eventually help build machines that act in the physical world.
The central message is clear. If AI progress continues as Amodei expects, the debate will not remain limited to model rankings or product features. It will move toward the basic structure of the economy, the future of labor and the role people play when machines can do almost everything better.