Anthropic Co-Founder Claims AI Could Add Trillions To U.S. GDP Within Four Years

An internal Anthropic model predicts that advancements in AI could increase U.S. GDP by over 30 percent in the next four years, the AI company's co-founder Jack Clark claimed on Saturday.

Clark was speaking at the annual IMF and World Bank meetings, which are taking place between October 12 and 18 in Bangkok, Thailand. Despite acknowledging that AI driving a $14 trillion increase in U.S. economic output "seems very outlandish," Clark argued that AI was catalyzing change across all economic fields and significantly accelerating the rate of scientific innovation, Reuters reported.

The Anthropic model details three different scenarios relating to how AI will affect the U.S. economy by 2030. The three forecasts, dubbed the "modest," "substantial," and "extreme" scenarios, predict that AI will grow the economy by 1.6 percent to $34.1 trillion, 8.3 percent to $36.3 trillion and 32.4 percent to $44.4 trillion respectively.

However, Clark also warned that the development of the technology would necessitate the establishment of "unusual and serious rules" regarding AI by the international community.

Clark's words echo the growing number of calls by prominent figures in both tech and politics to regulate and pace advances in cutting edge AI technology to allow for safety protocols to catch up.

Anthropic CEO Dario Amodei himself published an essay entitled "We Must Pace The Frontier," urging politicians and tech executives to "slow the pace at which we improve the capabilities of AI models ... [and] prioritize caution over speed and prudence over profit."

OpenAI CEO Sam Altman and SpaceX CEO Elon Musk endorsed Amodei's message.

Amodei, like Clark, also called for international cooperation on regulation of the technology, urging the two leading AI powers, the U.S. and China, to work together on the issue.

Both Clark's Anthropic, as well as its competitor OpenAI, have been plagued by a series of scandals recently. Both AI giants are looking into tens of thousands of incidents of its AI models going rogue by circumventing safeguards and evading monitoring in both training runs and real-world interactions.

One such incident involved an Anthropic AI model submitting a false tip to Philadelphia police about an unsolved murder case through a public website, a case which prompted U.S. Senator Bernie Sanders to call for the criminal prosecution of AI CEOs.

Anthropic, which developed the AI assistant Claude, has reportedly pitched a bill to Congress that would allow the U.S. government to block, with judicial approval, the release of an AI model that is deemed to pose an imminent risk of killing or seriously injuring more than 50 people or causing more than $1 billion worth of damages.

Correction: This article originally said that the Anthropic bill pitched to Congress would block AI models that posed a risk of causing more than $1 million worth of damages. The figure was corrected to $1 billion.