If the anthropy succeeds, a nation can be born of benevolent AI genius

When Dario Amadei He is excited by AI – which is almost always – he moves. The co -founder and CEO rises from a place in a conference room and heads to a whiteboard. He scratches diagrams with hockey curves that show how machine intelligence bends to infinity. His hand rises to his curly hair on his hair, as if caressing his neurons to prevent the system from crashing. You can almost feel his bones vibrate as he explains how his company is, anthropic, is Unlike other AI modelsS He tries to create an artificial general intelligence – or as he calls it, “powerful AI” – which will never become a fraudster. He will be a good person, a utopia employee. And while amodei is vital to the anthropic, he comes second to the company most An important associate. Like other extraordinary creatures (Beyonce, Cher, Pele), the latter passes with one name, in this case pedestrian, reflecting her pliancy and comity. Oh, and that’s an AI model. Hi Claude!

Amadei just returned from Davos, where he ignited the flames in the chats of fire, announcing that in two years, Claude and his peers would surpass people in any cognitive task. He hardly recovered from the trip, he and Claude are now dealing with an unexpected crisis. A Chinese company, called Deepseek, has just launched the latest large language model that has allegedly built up for some of what companies like Google, Openai and Anthropic consume. The current paradigm of an avant -garde AI, which consists of multidirectional hardware and energy costs, suddenly seemed trembling.

Amodei is probably the person who is most linked to the maximalist approach of these companies. Even when he worked at Openai, Amodei wrote internal paper about something he had thought for years: a hypothesis called the big stain of Compute. AI Architects knew, of course, the more data you have, the help your models can be. Amodei suggested that this information be more suggested than they thought; If they fed megatons on things to their models, they could accelerate the arrival of a powerful AI. The theory is already a standard practice and is the reason why the leading models are so expensive to build. Only a few companies with deep pockets could compete.

Now a newcomer, Deepseek– from a country to be controlled by the export of the most powerful chips – there was a rolled out without a large spot. If a powerful AI could come from everywherePerhaps the anthropic and his peers were computing emperors ditchesS But Amodei makes it clear that Deepseek does not support it at night. He rejects the idea that more effective models will allow competitors with low-budgets to jump to the front of the line. “It’s just the opposite!” He says. “The value of what you do is increasing. If you get more dollar intelligence, you may want to spend even more dollars on intelligence!” Far more important than saving money, he claims, is to reach the final line of AGI. Therefore, even after Deepseek, companies like Openai and Microsoft have announced plans to spend hundreds of billions of dollars more on data centers and power plants.

What amodei does obsession is how people can reach AGI safely. It is so hairy that he was forced by the six other founders of Anthropic to leave Openai in the first place because they believe that he cannot be resolved with the Executive Director Sam Altman at the helm. In anthropic they are sprint To set global standards for all future AI models, so they actually help people instead of, in one way or another, to blow them up. The team hopes to prove that it can build AGI, so safe, so ethical and so effective that its competitors see the wisdom in the next suit. Amodei calls this the race to the top.

That’s where Claude comes. Hang around the anthropic office and you will soon observe that the mission would be impossible without it. You never encounter a Claude in the cafe sitting in the conference room or drive the elevator to one of the 10ths of the company. But Claude is everywhere and is from the first days when the anthropic engineers first trained him, lifted him and then used it to produce better Claudes. If Amodei’s dream comes true, Claude will be both our wing model and the Godmodel fairy when we enter the era of abundance. But here’s a cumbersome question suggested by the company’s own research: Can Claude be believed to play nice?

One of those of amodei The anthropic co -founders are no other other than his sister. In the 1970s, their parents Elena Engel and Ricardo Amade moved from Italy to San Francisco. Dario was born in 1983 and Daniela four years later. Ricardo, a leather master from a tiny town near Elba Island, became ill when the children were young and died when they were young adults. Their mother, a Jewish American, born in Chicago, works as a manager of library projects.