Is AI Advancing Too Quickly?
Artificial intelligence continues to develop at a remarkable pace, but that progress raises serious questions about safety and security.
During an appearance on Mornings With Brian Haldane, technology commentator Haggai Davis discussed the rapid evolution of large language models (LLMs), AI agents and the potential risks associated with increasingly capable systems.
Davis explained that AI agents can work together to complete complex tasks. As companies develop more sophisticated systems, researchers face a growing challenge: ensuring that new capabilities come with effective safeguards.
The concern, Davis emphasized, is not necessarily a Hollywood-style “rise of the machines.” Instead, powerful AI tools could give malicious actors new ways to attack computer networks, public utilities and other critical infrastructure.
The New Cybersecurity Arms Race
The conversation also examined the growing competition between the United States and China in artificial intelligence.
Davis described the competition as resembling a new technological arms race. He argued that slowing development could allow competitors to close the gap, while moving too quickly could create security vulnerabilities.
That creates a difficult balancing act for AI companies and governments. Developers must continue advancing the technology while building safeguards that reduce opportunities for misuse.
AI Is Already Changing Jobs
The discussion also highlighted a more immediate impact of AI: job displacement.
Davis and Haldane discussed reports about Kenyan workers who earned substantial incomes writing academic essays for overseas students. As students increasingly turn to AI tools to generate essays, that source of work has declined.
The example illustrates how AI can transform employment far beyond Silicon Valley. Jobs built around writing, research and other digital services may change rapidly as AI becomes more capable.
For Davis and Haldane, the central question remains how society can capture AI’s potential benefits while managing its cybersecurity, economic and ethical risks.