AI leaders keep building powerful systems, then warn that their creations could end the world. The latest example is Anthropic’s Claude Mythos, which they say can find cybersecurity bugs better than humans and poses serious risks if misused.
This isn’t new. OpenAI once delayed GPT-2 for safety reasons, only to release it later. Leaders like Sam Altman and Dario Amodei have repeatedly talked about extinction risks while racing to ship products and raise billions.
Why the Fear Narrative?
Critics say it’s smart marketing. By painting AI as an almost supernatural threat, companies make themselves look like the only ones capable of controlling it. This distracts from today’s real problems: massive energy use, deepfakes, mental health impacts, and labor issues.
It also discourages tough regulation after all, who wants to slow down the heroes protecting us from doomsday?
Skepticism and Reality
Experts like Heidy Khlaaf question some of the bold claims, noting missing standard metrics. History shows Silicon Valley often overhypes: remember the Metaverse or Bitcoin replacing all money?
Who benefits? AI companies gain higher valuations, investor interest, and lighter oversight.
Who loses? The public faces anxiety, delayed regulation, and unaddressed current harms.
The Bottom Line
Advanced AI does carry real risks, especially in cybersecurity. But treating it like an ungovernable force of nature serves company interests more than public safety. We’ve successfully regulated nuclear tech and biotech before.
The smarter approach isn’t panic it’s clear-eyed oversight, transparency, and treating AI as the powerful tool it is, not a myth. (198 words)