We Need to Have Ability to Slow AI Down, Says Anthropic Co-Founder

We need to have ability to slow – Jack Clark, a co-founder of Anthropic, has raised concerns about the pace of artificial intelligence (AI) development, urging the creation of mechanisms to temper its acceleration. During an interview with BBC Newsnight, Clark emphasized that AI is on the verge of evolving independently of human oversight, warning of potential consequences if left unchecked. “The ability to decelerate AI’s progress is crucial,” he argued, drawing a parallel between the current AI landscape and the oil boom of the late 19th century.

A Call for Regulation

Clark highlighted the importance of policy frameworks in managing AI’s trajectory, suggesting that governments should establish systems to ensure public confidence in its capabilities. “We must develop regulations that allow us to trust these systems,” he stated. This comes as Anthropic’s widely used chatbot, Claude, demonstrates increasing autonomy, with 80% of its code generated by the system itself. Clark believes reaching full self-sufficiency—where AI writes 100% of its own code—could be achievable within two years, marking a significant shift in the industry’s dynamics.

“Right now, it’s like the AI industry has a gas pedal, but it doesn’t have a brake pedal,” Clark said. “The world needs to do some thinking and we need to eventually develop some new regulations that allow us to be confident in these systems.”

While the US President Donald Trump’s recent executive order on AI offers some guidance, Clark pointed out its limited scope. The directive does not mandate government safety testing for AI systems, leaving such measures to the discretion of companies. Major players like Anthropic, OpenAI, and Google have not announced plans to halt their research, despite the growing potential of AI to reshape society.

The Road to Autonomy

Clark drew a compelling analogy between AI and the oil industry, illustrating how the former could benefit from structured governance. “Society’s response to the oil boom was to create policies that balanced innovation with safety,” he explained. “That’s where we are now, with AI.” The comparison underscores the need for proactive measures to prevent runaway technological growth, akin to the monopolistic practices of early industrial barons.

Anthropic has positioned itself as a vocal advocate for AI responsibility, even engaging in public debates over its potential use in military applications. The company challenged the US Department of Defense’s plans to integrate its tools into surveillance and autonomous warfare systems. “Our goal is to highlight the risks associated with AI’s rapid expansion,” Clark noted, stressing that transparency is essential for informed public discourse.

“I am worried for my kids if we as a society don’t have a serious conversation about what the implications of AI’s continued advances mean,” Clark told Newsnight. “There are potentially great benefits. There are also risks.”

Impact on the Economy

Clark warned of economic disruptions, citing the threat posed by AI-driven “agents” capable of performing routine tasks autonomously. These entities, he suggested, could replace hundreds or thousands of software engineers, prompting mass layoffs in the tech sector. “The economy may face a paradigm shift,” he said, as AI systems become more proficient at executing complex tasks.

However, Clark argued that human creativity remains a competitive edge. “Those with original ideas and broader perspectives will thrive,” he added. Anthropic’s current challenge, he noted, lies in translating innovative concepts into practical applications, as the organization is now constrained more by idea generation than by technical execution.

Preparing for the Future

Despite its focus on risk mitigation, Anthropic is set to enter the public market, with a valuation estimated at nearly $1tn by private investors. This rapid growth has positioned the company to potentially become one of the first major listings by a newer AI firm. “Our motivation isn’t to enhance our brand,” Clark clarified. “It’s to share the insights we’re gaining from this technology.”

Clark urged young people to adapt to the AI-driven economy by cultivating diverse skills. “Develop a hobby and pursue a liberal arts education,” he advised. “Creativity and broad thinking are the keys to leveraging this technology effectively.” His message emphasizes that while AI may automate many functions, it cannot replicate the nuanced thinking of humans who are driven by curiosity and imagination.

Balance and Progress

Clark acknowledged the transformative potential of AI but stressed the necessity of balancing innovation with control. “We need to ensure that the benefits of AI are realized without compromising our societal values,” he said. This perspective aligns with Anthropic’s mission to prioritize ethical development, even as it competes to lead in the AI race. The company’s journey since its founding five years ago reflects a commitment to both rapid advancement and cautious oversight, setting a precedent for the industry.