[Digital Today reporter Jinju Hong] Prominent venture investor Bill Gurley (빌 걸리) has launched a direct criticism of the development philosophy of AI company Anthropic. He raised concerns, arguing that Anthropic has a mindset aimed at creating a being that surpasses humans, beyond simply developing high-performance AI.
On June 1 (local time), Business Insider reported that Gurley recently appeared on the All-In podcast and set out critical views on Anthropic’s AI safety strategy and long-term vision.
Gurley said, “After intensively reviewing materials related to Anthropic for a month, my perspective on the company completely changed.” He claimed some people “believe it is their responsibility to create a species superior to humans, and they seem excited about it.” He added, “They are not simply building software; they are summoning God here.” Gurley said he calls this perspective the “Frankenstein Theory.”
The criticism is drawing attention as a step beyond the regulatory debate that has been raised around Anthropic. Some in the industry have repeatedly argued that Anthropic’s push for strong AI safety regulation may be intended to make it harder for competitors to enter the market.
But Gurley pointed to what he called a more fundamental issue than regulatory strategy. He claimed there is a view inside Anthropic that looks positively on a future in which AI more powerful than humans takes on a core role in running society and the economy. As a basis, he cited an essay published last year by Anthropic co-founder and CEO Dario Amodei (다리오 아모데이) titled “Machines of Loving Grace.”
In the essay, Amodei presented multiple scenarios for how the economy and society could change after the emergence of ultra-high-performance AI. Gurley focused in particular on a passage describing an economic structure in which AI allocates resources. He said, “In some ways it is a story that assumes a godlike being,” and argued it “depicts a structure in which AI decides what it will provide to humans and what rewards it will give.”
Anthropic did not issue a separate official stance on the remarks. The company has stressed AI safety as its top value. In an interview last year on the CBS programme “60 Minutes,” Amodei said he feels “deep discomfort” about the way AI companies control themselves and argued for the need for responsible regulation.
In another interview earlier this year, he said safety concerns led the company to delay releasing its model “Claude” before the 2022 launch of ChatGPT. He said it ultimately lost a chance to get ahead in the consumer AI market, but prioritised safety.
Anthropic has recently delayed the release of some models for safety reasons. The company previously put on hold the release of its next-generation model “Claude Mythos,” citing concerns it could broadly expose cybersecurity vulnerabilities. It said last week it plans to provide the model to customers within weeks.
Policy controversy is also continuing. David Sacks (데이비드 색스), who oversaw AI and cryptocurrency policy in the Trump administration, recently criticised Anthropic by saying that if it defines itself as a “safe AI company” and portrays competitors as relatively risky, it could ultimately expand monopolistic influence over the market.
All-In podcast co-host Jason Calacanis (제이슨 칼라캐니스) also claimed, “It seems there is an atmosphere inside Anthropic that believes it can create God.”
The controversy is drawing more attention as it coincides with Anthropic’s push for an initial public offering. Anthropic announced on June 1 that it has begun listing procedures by submitting a confidential draft S-1 to the U.S. Securities and Exchange Commission. The industry sees that if the listing proceeds, it could be among the largest AI company IPOs of recent years.
The industry expects scrutiny and debate around Anthropic to intensify as AI safety regulation, philosophies on artificial general intelligence, model release policy and listing plans converge at the same time.