▲ Dario Amodei, CEO of Anthropic
While industry leaders driving the global artificial intelligence (AI) sector have unexpectedly raised the notion of slowing down AI development, criticisms against it are emerging one after another.
Critics argue that the companies that have been most aggressively developing AI on the industry frontline are merely serving their own self-interest and acting hypocritically by calling for regulations to slow down the current pace.
The New York Times (NYT) reported on the 13th (local time) that skepticism is being raised in some parts of Silicon Valley regarding the idea of slowing down AI development, as the debate surrounding AI safety spreads into a public spat among key figures.
Previously, Dario Amodei, CEO of Anthropic, expressed the view on slowing down AI development through a lengthy post on the 12th.
In response, key industry figures such as OpenAI CEO Sam Altman, Google DeepMind CEO Demis Hassabis, and xAI's Elon Musk expressed agreement and support for the need to moderate development speed and introduce government regulations.
Consequently, David Sacks, Chairman of the White House Council of Science and Technology Advisers, criticized the stance in a social media post, stating, "Do not act as if the motivation to slow down is purely altruistic."
He pointed out that despite tech companies facing no restrictions on slowing down AI development on their own, it remains questionable why government intervention is supposedly necessary.
"The easiest way not to build artificial general intelligence is to agree not to build it," he said, adding, "Demanding that preferred regulations be crafted in exchange can be perceived as an act of extortion against the public and the political system."
Aidan Gomez, CEO of Canadian AI startup Cohere, also asked in an op-ed on the 13th, "Should we allow a small group of market-dominant AI companies in Silicon Valley to define the rules and safety standards of a generational technology for the entire world?"
He added, "These oligopolistic firms are now demanding the authority to rewrite the rules of competition in their favor and dictate the terms that will apply to everyone else. This is a wolf in sheep's clothing, nothing more than a cartel by another name."
The NYT evaluated that such criticism toward the slowdown argument reveals a deep divergence in views between those who believe the government should regulate AI and those who view regulatory intervention as stifling competition.
Criticisms were also raised regarding Amodei's proposal that AI companies have their products evaluated by third-party "internal evaluators."
Critics questioned whether the two organizations mentioned by Amodei—METR (Model Evaluation and Threat Research), a non-profit evaluating AI safety, and Redwood Research—maintain excessively close ties with Anthropic.
Venture capitalist Bill Gurley stated, "If regulations are introduced, the regulatory body must not be affiliated with the entities being regulated, nor should the regulated companies draft their own rules."
Clément Delangue, CEO of open-source AI model sharing platform Hugging Face, which has previously experienced hacking originating from OpenAI's models, stated in a social media post, "AI safety is not an issue that can be resolved behind closed doors by a handful of advanced labs."
Lawmakers in Washington also seemed to sympathize with the slowdown argument, yet appeared unconvinced about the necessity of regulation.
Texas Republican Senator John Cornyn stated in a post, "If AI developers want to slow down the development of their own products, they are free to do so. One thing is certain: our competitors, including our adversaries, will not be doing that."
President Donald Trump also stated during a press conference on the 13th, "The United States is leading China in the field of AI, and I want to keep it that way."
(Photo: AP, Yonhap News)
※
Copying, redistribution, and unauthorized use in AI training are strictly prohibited.