A wave of doomsday warnings about artificial intelligence has been making the rounds online, but not everyone inside the industry is convinced humanity is doomed. The BBC spoke with several current and former employees of major AI companies, including OpenAI, Meta, and DeepMind, who pushed back on the idea that AI development will inevitably produce tools capable of mass destruction.
The renewed panic followed viral comments from Jacob Coxon, a former employee of Anthropic (an AI safety-focused company), who suggested that future AI agents could one day design and deploy biological weapons. Other industry voices, including some at OpenAI, DeepMind, and Elon Musk’s startup xAI, echoed similar warnings, calling for a slowdown in AI development.
But several workers who spoke to the BBC anonymously, since they weren’t authorized to talk to the press, said such claims tend to be vague or rely on far-fetched hypothetical leaps rather than concrete evidence. One former OpenAI employee said the argument that AI puts all of human life at risk is rarely backed by real detail. Nvidia’s CEO also dismissed the fears as “unnecessary” and “irresponsible.”
Still, experts agree that AI does pose real, more immediate risks, from hackers bypassing safety systems to the technology’s growing use in military applications.
Curious about the full debate? Check out the original BBC report.
Source: Not all AI workers think the tech could kill everyone (feeds.bbci.co.uk).
Image: Gabriel GM, BY 2.0 (via Openverse).