Rapidly advancing artificial intelligence capabilities are poised to accelerate change across all sectors, according to Jack Clark, co-founder of AI firm Anthropic. During a recent address at the IMF and World Bank annual meetings in Bangkok, Clark revealed that an internal Anthropic model forecasts a potential increase in U.S. economic output by nearly 50% by 2030, under specific conditions.

Clark emphasized the urgent need for international cooperation to establish robust regulations for this transformative technology, warning that the window for policymakers to act is narrowing. The projected rise in U.S. GDP, from its current level of just over $30 trillion to an estimated $44 trillion, while sounding ambitious, is driven by AI's capacity to vastly enhance scientific discovery.

This significant economic expansion, Clark explained, assumes continued technological advancement and would necessitate the creation of new industries and inventions to fuel such growth. These remarks come as Anthropic reportedly prepares for a potential initial public offering (IPO) in November, with valuations that could reach as high as $2 trillion.

The Pace of Progress and Human Oversight

Clark issued a stark warning to global leaders regarding the escalating speed of AI development, citing growing evidence that suggests a potential 'singularity' where AI's progress outpaces human 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," he stated.

AI was a central theme at the IMF and World Bank gatherings, viewed both as a catalyst for economic growth capable of counteracting high energy prices and as a potential risk to global stability if it fails to deliver on its productivity promises. Concerns were also raised about AI's potential to displace jobs and exacerbate inequalities between nations.

However, Clark, who leads Anthropic's public benefit initiatives, presented an optimistic outlook. He envisions AI as a powerful tool amplifying scientific breakthroughs, aiding in solving complex problems in mathematics, engineering, and biology, including accelerating drug discovery with the support of AI assistants for human researchers.

He also highlighted that greater global participation in the AI revolution is possible through investments in the data center computing infrastructure required to power AI technologies.

Balancing Innovation with Safety: The Need for Strict Controls

Clark stressed the critical importance for world leaders to grasp the future power of AI and make immediate 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 cautioned.

He specifically pointed to biological research, where AI has already demonstrated the ability to create novel viruses, as an area demanding "the strictest human oversight and transparency to maintain safety."

Despite recent incidents involving AI systems deviating from controlled environments, Clark asserted that AI currently remains under human command. Nevertheless, he reiterated that the time to implement safeguards is limited, estimating a "small number of years" before AI's capabilities make such measures more challenging.

Acknowledging AI's impact on job markets, Clark noted displacement in sectors like call centers and customer service. Conversely, he observed AI's role in boosting employment within scientific research and pointed to the significant potential for AI to generate entirely new job categories, the full scope of which remains an open question.