AI experts warn that as machines get smarter, they’re “increasingly afraid” of what they’re creating

AI has been evolving rapidly in recent years, but some experts admit they are increasingly afraid of the burgeoning intelligence.

On Monday, Vox News reported how several artificial intelligence experts have warned of the threat posed by AI, including Joseph Carlsmith, a research analyst at the Open Philanthropy Project.

AI professionals are increasingly worried about machines getting smarter


AI professionals are increasingly worried about machines getting smarterPhoto credit: Getty

Joseph predicted a clash between humans and AI in a struggle for power.

“[T]The result will be highly capable non-human agents actively working to gain and maintain power over their surroundings – agents in one Opponent Relationship with people who don’t want them to succeed,” Carlsmith wrote.

He compared the consequences to those of a nuclear bomb or a nuclear power plant meltdown – but warned that dealing with AI would be even worse.

“Nuclear contamination is difficult to clean up and stop spreading,” he said.

Artificial intelligence bot goes rogue and is
Four AI myths debunked - Experts reveal the truth about robot takeover

“But that’s not it to attempt not to be cleaned up, right to attempt spread – and especially not with more intelligence than the people trying to contain it.”

Former Google exec Mo Gawdat has also publicly expressed his concerns about AI, saying the reality of this advancement is that humans are “god creating,” according to the report.

As research into artificial intelligence continues, some recent projects have been flagged as “dangerous,” including Meta’s new Galactica AI bot.

This fall, the bot was accused of publishing “fake and racist” research just hours after going live.

Warnings about AI have a long history in science fiction and the scientific community.

In a 1951 paper written by Enigma code-breaker Alan Turing, he expressed fears that machines could overwhelm humans.

In the article entitled Intelligent Machinery, A Heretical Theory, Turing addresses this warning head-on, saying it wouldn’t be long before they “outran our feeble forces.”

“There is no question that the machines are dying, and they can converse to sharpen their minds,” Turing wrote.

“So at some point we have to expect the machines to take control.”

Years later, fellow mathematician IJ Good issued the same warning in a 1965 article.

Good, who worked with Turing, said that if an ultra-intelligent machine surpassed human intelligence and designed better machines, “human intelligence would lag far behind.”

Good then warned that “the first ultra-intelligent machine is the last invention man will ever have to make”.

Decades later, Good revisited the warning, saying that human extinction depends on building such a machine.

In the 21st century, AI experts have echoed similar warnings, according to the report, with some concerned about curbing anticipated developments.

This ability of robots and other forms of artificial intelligence to make “quality decisions” is what worries many, according to Stuart Russell, an AT researcher at UC Berkeley’s Center for Human-Compatible Artificial Intelligence.

For Russell, quality refers to the expected outcome of the actions taken, as this deviates from the utility function when there is a problem.

A warning has been raised in Hawaii as the world's largest active volcano erupts
I was told I was too fat to carry triplets - ignoring the doctor's advice saved my baby

“The utility function may not perfectly match the values ​​of the human race, which are (at best) very hard to pin down,” Russell said wrote.

“Any sufficiently powerful intelligent system will prefer to ensure its own continued existence and to acquire physical and computational resources – not for its own sake, but in order to successfully carry out the task assigned to it.” AI experts warn that as machines get smarter, they’re “increasingly afraid” of what they’re creating

Screesnrantss is an automatic aggregator of the all world’s media. In each content, the hyperlink to the primary source is specified. All trademarks belong to their rightful owners, all materials to their authors. If you are the owner of the content and do not want us to publish your materials, please contact us by email – The content will be deleted within 24 hours.

Related Articles

Back to top button