Live Science poll: 76% want AI development stopped or delayed over safety fears

A new Live Science poll reveals that 76% of over 1,700 readers believe artificial intelligence development should either be stopped immediately or significantly delayed due to safety concerns. However, 30% of respondents think it’s already too late to halt AI’s progression toward superintelligence, with many citing the irreversible nature of technological advancement and the global competitive dynamics driving AI research.

What the poll found: The September survey exposed deep public anxiety about AI’s trajectory toward potential superintelligence, known as the singularity.
• 46% of the 1,787 respondents believe AI development must stop now because the risks are too great.
• 30% think advancement should be delayed until proper safeguards are developed.
• Only 9% believe the benefits will outweigh the risks, while just 5% think superintelligence will never be achieved.

What readers are saying: Many commenters expressed resignation that AI development has passed the point of no return.
• “It is too late, thank God I am old and will not live to see the results of this catastrophe,” wrote Kate Sarginson.
• CeCe responded: “[I] think everyone knows there’s no shoving that genie back in the bottle.”
• One reader noted: “It’s an international arms race and the knowledge is out there. There’s not a good way to stop it.”

The skeptical perspective: Some readers compared current AI fears to historical technology anxieties that proved overblown.
• “For every new and emerging tech there are the naysayers, the critics and often the crackpots. AI is no different,” commented From the Pegg.
• Instagram user alexmermaidtales drew parallels to early electricity fears: “Would you believe this same question was asked by many when electricity first made its appearance? People were in great fear of it, and made all kinds of dire predictions.”

Proposed solutions: Rather than complete halts, some readers suggested regulatory approaches to manage AI development.
• One commenter proposed “heavy taxation on closed-weight LLM’s [Large Language Models], both training and inference, and no copyright claims over outputs.”
• They also suggested progressive taxation on larger model training while allowing smaller, specialized models to remain under consumer control.
• The goal would be “shifting incentives from pursuing AGI into making what we already have more usable.”

Why this matters: The poll reflects growing public awareness of AI risks as the technology approaches potentially transformative milestones, with artificial general intelligence potentially arriving as early as 2026 according to some experts. The divided opinions highlight the challenge policymakers face in balancing innovation with safety concerns while competing in what many see as an unstoppable global AI race.