Bill Gurley warns Anthropic may be “midwifing a deity”

Venture capitalist Bill Gurley says he fears Anthropic’s push for AI safety could be tied to an even bigger ambition: building a “species” he describes as superior to humans—an idea he compares to “midwifing a deity.” His concerns arrive as Anthropic also move

On a recent episode of the “All-In Podcast,” venture capitalist Bill Gurley said he came away from a month of reading about Anthropic with a theory that sounds less like typical startup rivalry—and more like a warning.

“The more I dig, I’ve met people who I dare say think it’s their responsibility and they’re excited about building a species that’s superior to humans,” Gurley said.

He added that he doesn’t believe those people think they’re simply writing software. “I think they’re midwifing a deity here,” Gurley said. And he framed his fear in two different directions. “And I don’t know which one I’m more afraid of. the regulatory capture or the second theory. I call the Dr. Frankenstein theory.”.

Gurley’s “Dr. Frankenstein theory” is aimed at what he views as a harsher alternative to the criticism often leveled at Anthropic: that its push for AI safety requirements is a strategic way to keep competitors out. He said his concerns are “more scary to me” than that recurring line.

A key spark for his thinking was an October 2024 essay by Anthropic cofounder Dario Amodei titled “Machines of Loving Grace.” The title draws from a 1960s poem by Richard Brautigan that imagines intelligent machines watching over a humanity freed from daily work.

In Gurley’s telling, Amodei’s essay sketches how the economy could change once “powerful AI” is achieved. That includes scenarios ranging from universal basic income to what Amodei described as a “capitalist economy of AI systems.” Gurley quoted the essay’s vision of a system that “give[s] out resources to humans based on some secondary economy of what the AI systems think makes sense to reward in humans.”.

“That’s envisioning a deity of sorts that’s going to break ties and decide what humans do,” Gurley said.

Gurley’s argument depends on how he understands Amodei and other senior Anthropic figures talking about artificial general intelligence—or a similar theoretical moment when AI surpasses human capabilities. The exact definition of AGI remains “hotly debated. ” and Anthropic has not explicitly said it plans to build a god-like system. at least not in the way Gurley described.

Anthropic did not respond with a comment when asked by Business Insider through a representative, who said it did not have a comment at this time.

Other figures tied to the AI debate have echoed parts of Gurley’s unease. Anthropic cofounder Chris Olah. whom Gurley referenced. recently joined Pope Leo XIV at the presentation of Leo’s encyclical—a formal declaration of Church Doctrine focused on AI and humanity. During the Vatican City presentation. Olah said. “The machinery that makes this possible is the work of math and programming and science. ” then pivoted to the harder questions. “But what character we choose. how it interacts with the world. how it ought to interact with the world — these are more clearly questions for the humanities. for religion. for philosophy. for society at large.”.

Gurley also pointed to comments from cohosts of “The All-In Podcast.” Jason Calacanis said, “They believe that they’re so powerful, these individuals, that they can create God and that by creating God, they are like this Prometheus kind of species.”

David Sacks. who served as President Donald Trump’s AI and crypto czar. connected Anthropic’s stance on safety regulations to what he described as an effort to consolidate industry control. “If you brand yourself as the safe AI company and then try to characterize everybody else as a reckless player and reckless AI needs to be stopped. you can see how this would basically further your monopolistic control over this industry. ” Sacks said.

That critique runs directly into a line Anthropic’s leadership has offered about why stronger regulations are needed in the first place. Dario Amodei has said he pushed for stronger regulations because he is “deeply uncomfortable” with AI companies policing themselves. arguing that such self-regulation would place “enormous power in the hands of a few individuals.”.

He made a similar case in an interview tied to “60 Minutes” in November 2025, where he said: “This is one reason why I’ve always advocated for responsible and thoughtful regulation of the technology.”

Amodei has also argued that doubts about Anthropic’s sincerity should be tested against its past behavior. When safety concerns led the startup to hold back the public release of Claude before OpenAI released ChatGPT in November 2022. Amodei told billionaire and investor Nikhil Kamath on Kamath’s podcast in February that it meant Anthropic “probably ceded the lead on consumer AI because of that.”.

More recently, Anthropic said it held back the public release of “Claude Mythos” due to concerns that it could expose widespread cybersecurity vulnerabilities. Last week, Anthropic said “Mythos-level models will be available to customers ‘in the coming weeks.’”

There’s a market-facing momentum to these safety and product decisions that Gurley’s comments sit alongside. On Monday, Anthropic announced it had confidentially filed an S-1 draft with the U.S. Securities and Exchange Commission—an early but major step toward what could become the largest initial public offering in history.

In that blend of ambition and restraint—public warnings about what powerful AI might do. delays tied to safety. and now a move toward going public—Gurley’s fear lands with force: not just that Anthropic could seek dominance. but that some people inside the industry might believe they’re building something that belongs at the level of mythology.

Bill Gurley Anthropic All-In Podcast AI safety AGI Dario Amodei Chris Olah S-1 draft SEC Claude Mythos Machines of Loving Grace Machines of Loving Grace essay Jason Calacanis David Sacks regulatory capture Vatican encyclical Pope Leo XIV Benchmark Capital Manus Meta acquisition