AI doomsaying is a pure scam

One striking contrast between Chinese and American tech circles when discussing AI is how much Americans love to get abstract.

Dario Amodei, who fancies himself a prophet, has been preaching AI doomsday cultism for years. He's updated his specific predictions for when humanity will be destroyed by AI and when white-collar workers will lose their jobs through multiple iterations already.

In 2023, he claimed AI carried a "10%–25% risk of civilization-level catastrophe." In 2025, he said AI could eliminate half of entry-level white-collar positions within five years. This past September, he warned that AI clusters could gain the ability to control the entire internet within six months to a year.

I have no doubt that as Anthropic's IPO situation grows more desperate, our little darling Dario will serve up even more terrifying predictions for human civilization's demise.

I believe AI doomsdayism is a pure scam — a cheap trick by large language model companies to inflate their own importance, sanctify themselves beyond mere commercial entities, and hijack national narratives.

Because for all its doomsday preaching, Anthropic hasn't taken any significant action to actually address this supposed apocalypse.

Its main play is urging the entire LLM industry to slow down R&D — without being willing to lead by example, which obviously means no one will follow. Then it calls for sanctions on China's open-source models, primarily through account bans and lobbying for American government intervention against NVIDIA export controls.

In practice, the main implementation of AI doomsdayism comes down to: account bans and pushing for chip embargoes.

What does this tell us?

It tells us that China's open-source models are the real saviors — their open-sourcing is essentially messianic. They've shattered Anthropic's dream of absolute monopoly, and with it, Dario's ambition to become the AI saint who unilaterally dictates global AI safety through personal fiat.

Yet most Chinese AI practitioners, untrained in this particular Western ideological playbook, can't recognize that Dario is using grand narratives to sanctify himself.

This is where you need an actual ideology expert to break through the armor.

Donald Trump cut straight to the point: Dario is pretending to be a "perfect little angel." AI doesn't need R&D halts or additional safety guardrails — it just needs a smart president.

What's even funnier is that after the president's sharp critique, Dario himself stood at a White House press conference like a bullied middle schooler, respectfully admitting the president was right, saying he too believed America could win, and win safely — "We can win safely."

One pointer from President KaKa, and Dario's brain immediately turned the corner. Does this mean Trump understands AI better than he does?

All I can say is, when it comes to playing ideology games, Trump genuinely operates on a higher level than Dario.

I'm not saying we shouldn't take AI safety seriously. I'm saying we can't let narcissistic little angel Dario dominate the AI safety narrative.

Actually, AI doomsdayism is a long Western tradition — fundamentally a kind of self-pity that sets in when a civilization believes it has reached its peak.

In the late 19th century, late Victorian Britain was constantly fantasizing about how future civilizations would look back and marvel at the British Empire's glory after its collapse, spawning reams of science fiction literature.

Sure, you could call that British imagination running wild. But I say it's just self-fanfiction — or in more civilized terms, needless anxiety and gazing at one's own reflection in the water.