- The co-founder of Anthropic, Jack Clark, warns that AI could advance without human input.
- Anthropic reports that 80% of its Claude code is written in-house.
- Jack Clark urges the authorities to ensure control over artificial intelligence, despite security issues.
- Major firms continue to develop rapidly with fewer constraints.
As artificial intelligence systems evolve from mere chatbots to full-fledged autonomous agents capable of processing information, coding, and decision-making, questions arise about their rapid development. The discussion received a new edge when Anthropic’s co-founder, Jack Clark, said that artificial intelligence is developing at a rapid pace with minimal human intervention.
Clark stated that the innovation has a gas pedal with no distinct mechanism to apply brakes. It symbolizes that it continues to develop without proper guidance and security measures. Thus, a regulatory check is needed to strike a balance.
Why is Jack Clark Calling for a Slowdown?
Clark warns that artificial intelligence has reached a point where it is self-developing without human deterrence. In an interview with the BBC, he stated that this nuance poses major risks to the industry as a whole. He highlighted that the authorities and government must intervene to exert control over artificial intelligence structures.
He further added that without appropriate policies, society is relying solely on corporate discretion. He drew a historical comparison between the artificial intelligence boom and the early oil industry, in which regulation helps ensure public trust, reliability, and accountability.
What is Happening Inside Claude?
Clark said that almost 80% of Claude’s code was written by the AI system. He mentioned that this percentage could reach 100% within two years, which will affect its reliability, security, and trust. This change is a comprehensive trend towards agentic AI systems that organize and execute tasks and results.
Such infrastructures are already embedded in software and proprietary analysis. As these agents merge with sub-agents and other tools, the predicament increases, making the regulatory process crucial and difficult.
Government and Organization’s Response
Even after Clark’s advisory, the responses remain unsettling. Anthropic welcomed an order on AI from US President Donald Trump but adopted a hands-off approach. The order did not call for government-led security for AI systems, leaving such measures voluntary.
Moreover, major AI developers, including OpenAI and Google, continue to boost their research and development. Anthropic also grows faster and is ready for a public listing with private stakeholders, with a valuation of $1 trillion. These conflicts between public safety and commercial benefit represent a conflict between security and competitiveness.
The rapid development of artificial intelligence could potentially affect the economy. The autonomous AI agents perform long-running tasks, but also instill fear of job replacement. Over the course of time, technology firms have announced large layoffs due to gains from artificial intelligence systems.
On the other hand, Clark stated that creativity is a distinct feature and acts as an advantage. He quoted that people who think comprehensively and originally will benefit from AI. For the upcoming generations planning to take advantage of this AI-driven economy, Clark suggested inquisitiveness and creative hobbies as major factors.
Ever since Dario Amodei assumed office as the CEO of Anthropic, the organization has clashed with the U.S. Department of Defense over surveillance and security issues. Jack Clark suggests that speaking about this issue is not a strategy, but an advisory for everyone. While the development possesses massive benefits, without appropriate regulation and checks, it could turn into a threat without any defenses in place. As AI systems become highly autonomous, the question arises as to what can be done to slow them down.
