Ray Blanco:

Anthropic Researcher Believes More Than 10% Chance AI 'Could Kill All Humans'

A top safety researcher at Anthropic has warned AI is advancing so quickly he believes there is a greater than 10% chance it "could kill all humans" within the next decade.

Evan Hubinger said in a post on X the risk from the models which currently exist was "low" but he was "worried" the technology might develop and improve itself soon to the point where it posed an existential risk to humanity.

He did not spell out how he thought AI systems could in future result in humans being wiped out. But his comments are the latest in a series of increasingly stark warnings about AI, with the debate shifting from whether it truly poses a risk to how big that risk is.

Hubinger's intervention was in response to another post on X from Jacob Coxon, an AI researcher who has just quit Anthropic and previously worked at OpenAI.

"Neither company is acting responsibly," Coxon wrote, adding, "These will soon be superhuman systems that can hack anything, revolutionise any field overnight, and acquire real power and resources."

⇒ Read More Here

Davis Wilson:

My Ride in Elon's Cybercab

After riding in just about every major self-driving service over the years – Waymo, Cruise, Aptiv and Tesla's existing Robotaxi – yesterday, I finally got my chance to ride in a Cybercab.

My thoughts? Cybercab felt much more natural than competitor Waymo. It took a direct route to my destination. Acceleration was smooth. Braking was smooth. Turns felt normal. At no point during the trip did the car do something that made me wonder what it was thinking.

My one complaint? Without mirrors or even a rear windshield, I had no easy way to see whether a car or cyclist was approaching from behind before I exited the vehicle into the street.

Tesla asked me for feedback after the ride, so that's exactly what I typed in the suggestion box. And then came my favorite part of the experience…

After submitting my feedback, Tesla asked if I wanted to leave a tip and suggested $0.69 and $4.20 as amounts… only for a hedgehog wearing sunglasses to pop up on my phone with the message "JUST KIDDING" after I selected. Of course, there was nobody to tip.

It truly is Elon's world. I'm just riding in it.

⇒ Read More Here

Chris Campbell:

What Could Happen if the CLARITY Act Fails To Pass in 2026

With lawmakers in the U.S. Senate set to consider legislation for regulatory clarity, there's a limited window for the bill to become law, potentially delaying it into the next session of Congress with different politics in play.

The U.S. Senate is scheduled to return to session on Monday after more than a month in which lawmakers were on state work periods. Senator John Thune, the Republican majority leader in the chamber, has scheduled a cloture vote on the Digital Asset Market Clarity (CLARITY) Act for Tuesday, in which his compatriots will need support from a handful of Democrats to meet the 60-vote threshold and overcome a filibuster.

Should the bill fail to advance with a three-fifths supermajority, the Senate will have less than 36 days of business before 2027, when a new session of Congress is scheduled to be sworn in, one where Democrats could be in control, depending on the outcome of November's midterm elections.

Senator Cynthia Lummis, one of CLARITY's biggest supporters, warned on Sept. 6 that the "next real opportunity" for the bill to pass might not be until 2030 if lawmakers were unable to reach an agreement.

⇒ Read More Here