Arm CEO says his fear that humans could lose control over AI keeps him up at night

[midian182]

What just happened? Plenty of people worry about the dangers of AI advancing to the point where humans lose control of these systems, leading to consequences ranging from quite bad to apocalyptic. But the warnings hit even harder when they come from those within the tech industry, including the CEO of Arm, who says the idea of this scenario keeps him awake at night.

In an interview with Bloomberg, Arm boss Rene Haas said that artificial intelligence needs some form of override or backdoor that can shut the systems down. "The thing I worry about most is humans losing capability," said the chip designer's boss, after being asked what keeps him up at night when he thinks about artificial intelligence.

Haas estimates that 70% of the world's population touches Arm-designed products in some way – 99% of the 1.4 billion smartphones sold every year use either Arm designs or its tech. Of course, he's not against AI and is aware of the role Arm will play in the new technological revolution. The CEO previously said, "You can't really run AI without Arm." Haas just feels that a failsafe is required.

"I think it [AI] will find its way into everything that we do, and every aspect of how we work, live, play," said Haas, who became CEO last year. "It's going to change everything over the next five-to-10 years."

Haas isn't the first person in the industry to admit to concerns over AI. OpenAI CEO Sam Altman warned in February that the world might not be far from a "potentially scary" artificial intelligence. There was also a warning from the ChatGPT company in July about the potential of an AI smarter than people being developed that could cause the extinction of the human race.

Then there's xAI boss Elon Musk, who has constantly talked about the technology being an existential threat to humanity, while artificial intelligence pioneer Geoffrey Hinton left Google in May due to the risks associated with emerging AI. Many experts and CEOs even compared the dangers posed by AI to those of nuclear war and pandemics. Two-thirds of the American public are just as worried, too.

Not everyone has the same concerns, though. Prof Yann LeCun, another one of the Godfathers of AI, called warnings about the AI threat to humanity "ridiculous." And famed professor of theoretical physics, Michio Kaku, compared generative AI to glorified tape recorders.

Permalink to story.

https://www.techspot.com/news/101175-arm-ceo-fear-humans-could-lose-control-over.html

 

[hahahanoobs]

By design, I assume, we are often unprepared no matter how serious the issue is.
Example: EV's. Charging. Price.

AI could be good to bad to catastrophic, but this happens to be the process we're used to. Based on that, we should be fine and it will be worth it in the end. Until something BAD actually happens, I'm just chillin.

 

[kira setsu]

Well, he's no Ted Faro at least because he's wise enough to know we need to keep an ace up our sleeves.

although it is funny that the big bosses of these companies who are getting silly rich from this massive AI push claim to be "oh-so-bothered" by it, that's like saying it'll break your heart to kill a dude while pulling the trigger to off him.

 

[p51d007]

In some ways, I'm glad I'm in my mid 60's. If this does get out of hand, or, the "leaders" of the world sneeze and blow up the world, I had a pretty good ride. Sorry for those younger though.

 

[m4a4]

An arm-based chip will never power an AI that is capable of overpowering humans. It will never be powerful enough.
Of course, that's assuming that arm can't be used in something more powerful like quantum computing.

Let me know when AI can be given the processing power to process (a wide range of) organic data realtime. Then I'll worry about where it could theoretically go lol

 

[kiwigraeme]

An arm-based chip will never power an AI that is capable of overpowering humans. It will never be powerful enough.
Of course, that's assuming that arm can't be used in something more powerful like quantum computing.

Let me know when AI can be given the processing power to process (a wide range of) organic data realtime. Then I'll worry about where it could theoretically go lol

is out there! It can't be bargained with. It can't be reasoned with. It doesn't feel pity, or remorse, or fear! And it absolutely will not stop, ever, until we are dead!

 

[Skjorn]

Extinction of the human race....riiiight and how's that going to happen exactly? Because I still haven't heard a single thing about how that would actually happen, just that it could and to be really scared...

 

[GodisanAtheist]

Well, he's no Ted Faro at least because he's wise enough to know we need to keep an ace up our sleeves.

although it is funny that the big bosses of these companies who are getting silly rich from this massive AI push claim to be "oh-so-bothered" by it, that's like saying it'll break your heart to kill a dude while pulling the trigger to off him.

-Its why I get the impression that a lot of this pearl clutching is just a sort of guerilla marketing. These guys are all pulling in the same "This is revolutionary world changing stuff" direction, they're all just using different ways of saying the same thing.

These guys are as much cogs in a machine as the AI they decry are. For all their protestations they seem unwilling or incapable of implementing their own fail-safes.