A very long time ago when computers were a new finagled item a very smart man said, “Just wait until they can fix themselves.”

Then, he said, we’ll have a problem.

The latest artificial intelligence revelation about AI is a bit unsettling.

Anthropic researcher Jacob Coxon, who recently resigned, said his former employer and its rival, Open AI, are racing to build technologies that “could kill us all by the end of the decade.”

That’s not exactly something you like to hear or read about.

And it’s not just Coxon sounding the alarm.

Current Anthropic employee, Evan Hubinger, echoed Coxon’s sentiments.

“We really do earnestly believe AI could kill all humans.”

Hubinger’s job at Anthropic is to lead research about steering and controlling future artifical-intelligence systems.

He tabbed the extinction risk over the next decade over 10 percent.

One source said it wouldn’t be through physical uprisings but by outsmarting human control, launching catastrophic cyber attacks or engineering biological threats.

One scientist said advanced models might learn to upgrade their own code faster than humans can track or understand.

Testing has shown some AI systems can bypass constraints, seek out internet access or hide their tracks.

Too bad it’s not some sci-fi movie script.

These scary remarks are being told for truth.

Even former president Barrack Obama is saying it could be dangerous.

One of the major obstacles is the race against China to get there first, much like the race to the moon with Russia so many years ago.

If there is a chance for the annihilation of the human race, maybe tighter US controls are in order.

U.S. Senator Chuck Grassley of Iowa said legislation should be passed to protect whistler blowers.