OK DOOMER: IS AI ABOUT TO WIPE OUT HUMANITY… OR IS IT THE GREATEST PR CAMPAIGN OF ALL TIME?

OK DOOMER

IS AI ABOUT TO WIPE OUT HUMANITY… OR IS IT THE GREATEST PR CAMPAIGN OF ALL TIME?

Big Tobacco pulled off one of the most successful PR campaigns of all time by persuading customers their product might not kill them.

The AI industry has taken that messaging and flipped it on its head: We probably are going to kill you.

Recently-departed Anthropic researcher Jacob Coxon took to X last week to drop his bombshell. 

He wrote: “I resigned from Anthropic yesterday… The people building AI earnestly believe that it could kill us all by the end of the decade.”

Then, perhaps anticipating this very blog, he wrote: “This is not a marketing stunt.”

His thread was littered with worrying phrases like “hubristic gamble” and “no other level of human activity poses this level of danger.” He described the Hugging Face attack (which we’ve written about here) as a “warning shot”.

Responses to the post ranged from the whimsical (“Well this was a pleasant read right before bed”); to the practical (“Could someone explain to me how AI could “kill us all’”); to the cynical (“So this is marketing for the IPO right”); to the downright – and somewhat irrationally – angry (“So leave just to virtue signal and show humans are too p**** to fight?”).

It was an extraordinary post. But perhaps even more extraordinary was the reply from Anthropic scientist – and current employee – Evan Hubinger, who said: “Jacob is correct here - we really do earnestly believe AI could kill all humans!”

Aside from being among the most ill-advised uses ever of an exclamation mark, Hubinger’s post was simply an astonishing thing for an apparently senior employee to write.  

Coxon and Hubinger had set a hare running that was now gathering pace.

Anthropic co-founder Jack Clark told the BBC a kill switch should be introduced into all AI platforms which could be monitored by a third party.

Anthropic chief Dario Amodei (pictured above) called for the pace of AI development to slow. OpenAI boss Sam Altman and X owner Elon Musk agreed.

Tech stocks dropped on the call for a slowdown, with Japan’s Softbank, a significant investor in OpenAI, down 10 per cent. 

Donald Trump, no doubt with one eye on the markets, sought to calm things down with his unique brand of diplomacy, saying: "AI taking over the World, destroying Humanity, and all other things bad, is a HOAX". He added reassuringly that the only guardrails AI needed was “a strong and smart president”.

At the heart of this story are questions around the rationale.

Are AI developers and bosses genuinely frightened at the rate of progress of AI superintelligence? Or is there an ulterior motive?

And if there is an ulterior motive, is it the desire to land a powerful marketing message? Or a bid to entrench a US incumbent advantage over the Chinese? Or an attempt to manage the speculation bubble?

With trillions of dollars at stake and vested interests at play, the truth most likely lies in a mix of all the above. 

Where fear, uncertainty and doubt were the watchwords of Big Tobacco, look out for more fear, certainty and doomsterism from AI.