Anthropic believes its success is key to making AI safe

Anthropic has spent the past five years Warning the world About how advanced artificial intelligence could lead to mass destruction, destabilize society, and cause a series of other massive harms. But at the same time, it has become one of the most powerful forces pushing AI capabilities forward. The company is now among the top developers and distributors of cutting-edge AI models and deals with clients such as the US Army. It has recently been appreciated Nearly $1 trillion.

At first glance, Anthropic’s stark messages and actions seem fundamentally contradictory.

But inside the company, many people don’t see any contradiction. To understand why, you must first understand that Anthropic operates on two basic beliefs. The first is that artificial intelligence is the most transformative technology in human history, and that its arrival is inevitable. The only real question is whether it will lead to disaster or extraordinary prosperity.

The second is that Anthropic believes the world would be better off if it stayed ahead of the AI ​​race, according to several former employees who spoke to WIRED on the condition of anonymity. Internally, leaders and employees at the company often refer to themselves as “the good guys,” the people in charge of AI technology, two of the sources said. The company sees the accumulation of power—whether in the form of capital, computing, research talent, or political influence—not as an end in itself, but as the price for achieving its goals. a task: “To ensure the world transitions safely through transformative AI.”

Helen Toner, executive director of the Georgetown Center for Security and Emerging Technology and a former member of the OpenAI board, uses Simile To describe the human view of the world. She compares powerful artificial intelligence to a forest full of magical treasures and dangerous monsters. All the nearby villagers rush in, attracted by the treasure. In her novel, Anthropic wants to venture into the jungle further than anyone else while investing heavily in taming the beasts—that is, harnessing the benefits of artificial intelligence while containing its disastrous risks.

“What’s special about Anthropic is that they say: ‘People are going into the forest anyway, we have to do that first.’” That’s very clearly their strategy: building cutting-edge AI in order to be a serious player at the table that can talk about what cutting-edge AI systems should look like, what risks they pose, and push for reasonable safeguards. “They’re very clear about this. “It’s just a weird enough strategy that people have a hard time hearing about it.”

Anthropic CEO Dario Amodei explained this approach clearly in a filing A conversation with his co-founders On the company’s career page: “You have to find a way to actually be competitive, to actually lead the industry in some cases, and still be able to do things safely,” he says. “If you could do that, the gravitational force you would exert would be very large.”

He was anthropic Founded in 2021 By a group of former OpenAI employees who defected after losing confidence in the ability of the company’s leadership — especially CEO Sam Altman — to safely bring transformative AI to the world. This sentiment still shapes the company today. Two former employees I spoke with say that in internal discussions, Anthropic executives often cite Altman and OpenAI — and, to a lesser extent, Meta and Elon Musk’s xAI — as cautionary examples that help define Anthropic’s sense of responsibility.

In many ways, Anthropic is like any other Silicon Valley company. Many startups market themselves as David fighting against established, outdated giants in industries they want to disrupt. Google, Facebook, and Apple were all founded on idealistic principles, which later became muddled or abandoned altogether as they became richer, larger, and more powerful.