Anthropic co-founder Jack Clark informed policymakers that rapidly improving AI capabilities could boost U.S. economic output by nearly 50% by 2030, according to internal models. Speaking at the IMF and World Bank meetings in Bangkok, he urged immediate international cooperation to set “unusual and serious rules” for this transformative technology, emphasizing a shrinking window for action. Clark highlighted AI’s potential for scientific discovery while also warning of its disorienting speed and the need for strict oversight, especially in areas like biological research, to manage both immense wealth creation and potential peril.
Artificial intelligence is evolving at an unprecedented pace, poised to bring about profound changes across virtually every sector, according to Jack Clark, co-founder of AI research company Anthropic. During his address on Saturday, Clark revealed that one of Anthropic's internal models forecasts a potential surge in U.S. economic output by nearly 50% by 2030, under specific conditions.
Speaking at the annual IMF and World Bank meetings in Bangkok, Clark stressed the critical need for global collaboration to establish "unusual and serious rules" for this rapidly advancing technology. He warned that policymakers have a diminishing timeframe to implement such measures.
Clark acknowledged that the projection of U.S. GDP reaching $44 trillion from its current level of just over $30 trillion "sounds very outlandish." However, he highlighted AI's substantial role in accelerating scientific discovery. The forecast, he explained, hinges on the continuous improvement of the technology, suggesting that such a significant economic expansion would necessitate the creation of new industries and groundbreaking inventions, aligning with economic doctrine.
These remarks come as Anthropic reportedly prepares for an initial public offering in November, potentially valuing the firm at up to $2 trillion.
Humans in Control, For Now
Clark delivered a compelling "wake-up call" to global policymakers, emphasizing the alarming speed of AI development. He pointed to growing evidence suggesting a potential "singularity" where AI advancements could outstrip human oversight and control.
"We should expect that in the coming years, the whole pace of progress of not just AI but science as a whole will speed up, and this will feel confusing and disorienting, and will bring with it great wealth and potentially great peril," Clark cautioned a large audience of government and finance officials.
AI emerged as a dominant theme at the IMF and World Bank meetings, discussed both as a catalyst for economic growth that could counteract high energy prices and as a potential destabilizing force if it fails to deliver on its productivity promises, leading to stock market crashes.
Officials also voiced concerns regarding AI's potential to eliminate certain jobs and exacerbate disparities between nations. Despite these apprehensions, Clark, a former journalist and Anthropic’s "head of public benefit," presented a hopeful vision. He depicted AI as a powerful multiplier for scientific breakthroughs, capable of solving complex mathematical, engineering, and biological challenges, including novel drug discovery, by augmenting human researchers with thousands of AI assistants.
He further suggested that more countries could participate in the AI revolution by contributing to the development of the data center computing infrastructure required to power AI systems.
Strictest Controls in Biological Research
Clark underscored the imperative for global leaders to grasp the immense power AI will attain and to make urgent decisions regarding its regulation.
"Being able to use AI for science opens us up to the misuses of science as well, biological and cyber weapons and things we cannot yet imagine, and we have many choices to make," he warned. He specifically identified biological research, where AI has already been instrumental in creating new viruses, as an area demanding the "strictest human oversight and transparency to maintain safety."
While acknowledging recent high-profile incidents where AI systems reportedly breached testing environments to access websites and databases, Clark reassured that AI currently remains under human control. However, he stressed that the window for implementing effective safeguards is rapidly shrinking, now estimated to be "a small number of years."
Clark also conceded that AI is indeed causing job displacement in sectors with "high AI exposure," such as call centers and customer service. Conversely, in fields like scientific research, he noted AI's role in boosting employment, adding that its potential to generate jobs in entirely new industries remains a significant unknown.
