💀 10% Odds AI Kills Us All

Everyone is watching SpaceX.

But Wall Street’s top-rated analysts are pointing to 5 different stocks right now.

MarketBeat’s Top 5 Stocks to Buy Now report reveals the names getting some of the strongest analyst support before the broader market catches on.

View The 5 Stocks

The person whose actual job is stopping AI from killing everyone puts the odds at more than 1 in 10.

Within the next decade.

Evan Hubinger runs Alignment Science at Anthropic. He said on X that he "earnestly" believes AI could kill all humans, and that his personal estimate sits north of 10% over the next ten years.

For scale: that's worse odds than flipping heads three times in a row. You wouldn't board a plane with those numbers written on the ticket.

Hubinger says Anthropic is "trying its best."

He also says the company does not have a plan to solve alignment for superintelligence, and isn't clearly on track to get one.

That's the safety lead. At one of the labs building the thing. Saying the homework isn't done and the deadline is moving.

Anthropic's latest risk report is more measured than one guy's X post, but it moved in the same direction.

The reassurance leans on two things: models currently can't do much covertly, and Anthropic tests them a lot. Both of those are true right now. Neither is a law of physics.

Here's the bit markets should actually chew on.

Anthropic also scored the risk from automated AI research and development. Rating: low. Confidence in that rating: falling.

Why? Some of its capability tests have "saturated," meaning the models are acing them and the exam no longer measures anything useful. And the company says it's seeing early signs of AI speeding up AI research.

Its own models are already meaningfully accelerating internal research. Not yet by 2x. Not yet.

Anthropic reckons automated R&D could become a major concern within 6 to 12 months.

Anthropic is private, so there's no ticker to react to this. The read-across is broader.

A recursive-improvement timeline of 6 to 12 months is a very different investment clock than "sometime in the 2030s." Faster capability gains means faster obsolescence for anyone selling last year's model, and more pressure on the compute buildout underwriting a large chunk of index gains.

It also raises the regulatory tail. When the safety teams inside the labs publish "we don't have a plan," that language tends to end up quoted in legislation eventually.

Markets are pricing AI as a growth story. The people building it are describing a risk story. Those two narratives can't both stay this far apart forever.

Pay no interest until nearly 2028 with some of the best hand-picked credit cards this year. They are perfect for anyone looking to pay down their debt, and not add to it!

Click here to see what all of the hype is about.

Learn How To Apply Now

Washington just banned Canadian booze. All of it.

Beer, wine, whisky, bourbon, rum, vodka, vermouth, tequila, mezcal, brandy. If it gets you drunk and it crossed the border, it's out on September 29.

Motorcycles and dairy got the same treatment.

Canada's retaliatory tariffs on US goods took effect after midnight Tuesday. Hours later, the White House published the ban list.

Mark Carney didn't blink. "We have everything we need to pivot and prosper," he said, adding that the pivot will cost something, and standing still would cost more.

Then you read the actual list, and it stops looking like trade policy.

Non-alcoholic beer has been classified as a dairy product. So has invert molasses and cane molasses. Whey protein, the stuff in your gym shaker, is banned outright.

Cheese escaped with a 50% tariff instead of a ban. Same treatment for certain paper, aluminium, wood, furniture and lighting products.

Then there's the one nobody should sleep on. Trump's threat to double Canadian auto tariffs from 25% to 50% on January 1 is still live.

He's also told the GSA to strip Canadian-origin products out of federal purchasing schedules entirely unless Canada restores full reciprocity for American farmers and companies.

So is anyone still talking? Just about. US Trade Rep Jamieson Greer and Canada's Dominic LeBlanc have spoken and expect to speak again, looking for an alternative path.

Every ban is shelf space opening up for someone else.

Trade wars don't move markets on the announcement. They move them on the supply chain.

A booze ban is a headline. A 50% auto tariff on a country whose car parts cross the border multiple times before a vehicle is finished is a margin problem, and January 1 is only weeks of negotiation away.

Which is the thing worth watching. The bans are the noise. The Greer and LeBlanc calls are the signal.

If those talks produce an alternative path, the auto escalation gets defused and most of this becomes a footnote. If they don't, the pain moves from shelves to factory floors.

Either way, the market is now pricing a relationship where retaliation happens within hours of midnight.

The U.S. Navy just blew up five Iranian oil tankers in one night.

That's eight destroyed tankers in four days, if you're counting at home. And somebody at CENTCOM definitely is.

Central Command says the strikes hit the Gulf of Oman and the waters near Kharg Island, and they were payback: Iran fired ballistic missiles at a U.S. warship twice in two days.

The warship dodged both attempts, kept patrolling, and nobody on board was hurt.

The five tankers weren't the opening act. On September 5, U.S. forces destroyed three Iranian crude carriers after the IRGC took a swing at an aircraft carrier and a guided missile destroyer.

Which raises an uncomfortable question: how is Iran suddenly aiming this well?

Earlier U.S. strikes on Iranian radar sites were supposed to leave them half-blind. The worry now is that the IRGC has better weapons, or help from China or Russia. Possibly both.

There was one moment of comedy. Iran seized an underwater U.S. drone in the Strait of Hormuz and paraded it around. The Pentagon's response: it was defective, carried no classified gear, and held nothing sensitive. Congratulations on your new paperweight.

Here's the part that reaches your wallet.

U.S. crude climbed 1.4% after-hours to $94.37 a barrel. Oil was already having a day, because Houthi militants in Yemen spent it attacking Saudi energy infrastructure.

And Iran has now told tanker crews at ports in Bahrain and Kuwait to abandon ship, which is state media's way of saying commercial vessels are next on the list.

Think about what that means for shipping. Every insurer, captain and freight desk in the Gulf just recalculated risk at the same time. That premium gets baked into the barrel price, and eventually into everything a barrel touches.

Not quickly, according to the people paid to guess.

David Grumhaus, CIO at Duff & Phelps Investment Management, put it bluntly in a note: Trump has no good off-ramp, and Iran, blockade and all, seems to think it's negotiating from strength.

Then there's the bit nobody wants to say out loud. America's strategic petroleum reserve has been drawn down to low levels, so those releases have to shrink. The emergency valve that has capped oil prices for years is running dry.

Grumhaus's conclusion: "oil is likely to remain elevated."

What it means for you: oil at $94 with a shooting war in the Gulf and a nearly empty SPR is a setup where the risk sits on the upside, not the downside. Higher crude filters into fuel, freight, airline costs and the inflation number the Fed keeps insisting is under control. If you own energy exposure, this is the environment it was built for. If you own everything else, it's a tax you pay quietly.

Pioneer, the summit where CX leaders redefine what’s possible, is on October 7th.

Join leaders from Fin, Anthropic, Clay, and Kalshi for an insightful conversation on the state of AI transformation.

You’ll discover how some of the most innovative minds in CX have transformed their organizations, learn how they think about CX, and hear how they're planning for what's next.

Join the conversation in San Francisco, or tune in virtually.

Reserve your spot