Elon Musk’s new plan to save humanity from AI

Sitting in the SpaceX Mission Control in Hawthorne, California, SpaceX CEO and Chief Designer Musk Elon watches his space capsule Dragon's progress as it heads for splash down in the Pacific Ocean. SpaceX Dragon capsule successfully traveled to and docked with the International Space Station. Astronauts sent Dragon on it's way back to earth on May 31, 2012.

WASHINGTON — Elon Musk has a new plan to protect humanity from artificial intelligence — if you can’t beat ’em, join ’em.

In October 2014, Musk ignited a global discussion on the perils of artificial intelligence. Humans might be doomed if we make machines that are smarter than us, Musk warned. He called artificial intelligence our greatest existential threat.

Now he is hoping to harness AI in a way that will benefit society.

In a recent interview with the website waitbutwhy.com, Musk explained that his attempt to sound the alarm on artificial intelligence didn’t have an impact, so he decided to try to develop artificial intelligence in a way that will have a positive affect on humanity.

So Musk, who is already the CEO of SpaceX and Tesla, is now heading up a startup called Neuralink. The San Francisco outfit is building devices to connect the human brain with computers. Initially, the technology could repair brain injuries or cancer lesions. Quadriplegics may benefit from the technology.

But the most amazing and alarming implications of Musk’s vision lie years and likely decades down the line. Brain-machine interfaces could overhaul what it means to be human and how we live.

Today, technology is implanted in brains in very limited cases, such as to treat Parkinson’s Disease. Musk wants to go farther, creating a robust plug-in for our brains that every human could use. The brain plug-in would connect to the cloud, allowing anyone with a device to immediately share thoughts.

Humans could communicate without having to talk, call, email or text. Colleagues scattered throughout the globe could brainstorm via a mindmeld. Learning would be instantaneous. Entertainment would be any experience we desired. Ideas and experiences could be shared from brain to brain.

We would be living in virtual reality, without having to wear cumbersome goggles. You could re-live a friend’s trip to Antarctica — hearing the sound of penguins, feeling the cold ice — all while your body sits on your couch.

But many technical hurdles remain. Musk believes it will be eight to 10 years before this kind of the technology will be ready to use by people without disabilities. Musk’s companies have made a habit of achieving what seemed impossible. But he’s also notorious for aggressive deadlines that his companies don’t meet.

Neuralink told waitbutwhy.com that it would need to simulate one million brain neurons before a transformative brain-machine interface could be built. If current rates of progress hold, it won’t reach that milestone until 2100.

In the meantime, there are many reasons for humans to be wary of implanting a computer in their brain. Any digital technology can be hacked. Humans might be unwittingly turned into malicious agents for unsavory causes. Computers crash too. If the interface fails, that could imperil our physical health.

With a brain-machine interface recording our lives, all of our experiences would be stored in the cloud. Privacy would be threatened. Governments or others would have incentives to access that information and track behavior.

If our brains merge with machines, our thoughts would become indistinguishable from what we’d downloaded from the cloud. We could struggle to know if our beliefs and views came from personal experiences, or from what the internet sent to our brains. Humans would be putting enormous trust in the maker of the brain-machine interface to share good information with them.

As Musk sees it, our options are limited.

“We’re going to have the choice of either being left behind,” Musk told waitbutwhy.com, “and being effectively useless, or like a pet.”