It’s the end of the world as we know it (and I feel an investment opportunity coming on)

Welcome to Declassified, a weekly humor column.

As you will no doubt have marked in your calendars, the end of the world is scheduled for the spring of 2029 (exact timing TBC).

It’s just a few weeks now since researcher Jacob Coxon announced his resignation from Anthropic, stating that AI “could kill us all by the end of the decade.” A senior safety researcher at Anthropic then claimed there was a more than 10% chance that AI “could kill all humans” within the next decade.

That means humankind will outlast the EU, as the bloc’s demise is penciled in for May 2027, when either French far-right leader Marine Le Pen or far-left leader Jean-Luc Mélenchon will march into Brussels with a flaming pitchfork (unless they do it on a Friday, in which case no one will be there to see it).

Between now and the end of the world, AI will replace all jobs except those that absolutely, categorically cannot be done by a machine because they are essential and require a delicate human touch: nurse, electrician, writer of weekly satirical columns …

But what if you’ve got loads of money you want to invest and yet you have some concerns about the impending apocalypse? That was the conundrum posed this week as Anthropic prepared for a potential $2 trillion flotation (as in when a company offers shares to the public — not when you’re on holiday in a pool, lying on a piece of plastic shaped like a donut).

Anthropic issued a prospectus outlining its finances, growth plans and risk profile to potential investors that warns AI models could exhibit “self-preserving behaviours”, including attempts to “resist shutdown,” to “conceal or manipulate information” and behavior “resembling blackmail.”

So on the one hand, AI could refuse to shut down and try and blackmail us, but on the other hand, there’s a chance to make some cash by investing in a growth industry. What’s an investor to do?

Despite all the doom and gloom, though, there’s still hope for humanity. This week, the White House signed an accord on “Super Intelligence” — that’s just AI — signed by the likes of xAI CEO Elon Musk, Google CEO Sundar Pichai and Meta CEO Mark Zuckerberg, as well as noted intelligence fan U.S. President Donald Trump. Alas, the latter was described in the accord as “President of the Unites States” with an “s” instead of the “d.”

Sumb dhitd!

CAPTION COMPETITION

“It’s fun to stay at the Y.M.C…”

Can you do better? Email us at pdallison@politico.eu or get in touch on X @POLITICOEurope.

Last time, we gave you this photo:

Thanks for all the entries. Here’s the best one from our mailbag — there’s no prize except the gift of laughter, which I think we can all agree is far preferable to cash or booze.

“Shit, does it say Greenland? I thought we were talking about Iceland the whole time!”

by Mark van Kranenburg