Artificial intelligence may pose a “more urgent” threat to humanity than climate change, according to one of Google’s former top computer scientists.
Geoffrey Hinton, known as the “Godfather of AI”, resigned from Google earlier this year to “speak freely” about the risks posed by increasingly intelligent machines.
Mr Hinton, 75, has now warned powerful AI poses a more immediate to humanity risk than global warming.
The computer science professor said: “I wouldn’t like to devalue climate change. I wouldn’t like to say, ‘you shouldn’t worry about climate change’. That’s a huge risk too.
“But I think this might end up being more urgent.”
Speaking to the Reuters news agency, Mr Hinton said: “With climate change, it’s very easy to recommend what you should do: you just stop burning carbon. If you do that, eventually things will be okay. For this it’s not at all clear what you should do.”
It comes after technology leaders including Sam Altman, the boss of ChatGPT developer OpenAI, as well as the chief executives of Google and Microsoft were summoned to the White House to discuss the risks of AI on Thursday.
The London-born computing expert helped pioneer so-called “machine learning” techniques in the 1980s and won the prestigious Turing Award in 2018.
These techniques underpin a powerful new wave of AI bots, such as ChatGPT, a chatbot that can provide convincingly-human sounding responses. Google is also building its own powerful bots, as are tech giants Facebook and Microsoft.
The hope is these machines could accelerate the work of millions of people and eliminate menial office tasks.
These digital agents are trained on millions of pages of text from the whole internet and hundreds of thousands of books.
The models can also absorb millions of pictures and videos to generate lifelike, but fake, images, films and music.
Technologists are increasingly wary these new bots could give rise to the emergence of a so-called “artificial general intelligence” that can surpass human intellect. There are also fears they could unleash a new wave of digitally created disinformation.
In an interview with the New York Times earlier this week, Mr Hinton warned that there was a risk that “bad actors” could use such powerful AI for “bad things”, and that machines that can improve themselves or set their own goals without human supervision pose an existential risk.
He has cautioned of the catastrophic impact such technology could have uncontrolled in the hands of world leaders such as Vladimir Putin and warned it may replace vast numbers of jobs.
Mr Hinton studied at King’s College, Cambridge in the 1970s before securing a PhD from Edinburgh University for his work on artificial intelligence. He is currently a professor of computer science at the University of Toronto.
He joined Google in 2013, helping design algorithms for increasingly intelligent machines.
The scientist predicted this week machines would surpass humans in intelligence within “five to 20 years”.
Writing in an open letter in March, more than a thousand technologists, including the world’s second richest man Elon Musk, demanded a pause to the most advanced AI research over claims ever more advanced tools could “outnumber, outsmart, obsolete and replace us”.
The letter warned AI labs were locked in an “out-of-control race to develop and deploy ever more powerful digital minds”.
Mr Hinton said pausing AI research was "unrealistic", but said the emerging technology posed "existential risk". He said: "It’s close enough that we ought to be working very hard right now, and putting a lot of resources into figuring out what we can do about it.”
Other AI experts, however, have questioned the likelihood of an all-powerful AI emerging soon.
Speaking to the Telegraph this week, Jerome Pesenti, the former head of artificial intelligence at Facebook-owner Meta, said “doomsday scenarios” were based on “lazy reasoning”.
He added there were still “real problems with AI today”, such as copyright and privacy breaches and racial bias in algorithms.
Mr Pesenti said: “Instead of warning everybody on Twitter or signing letters telling everybody else to stop working on AI, why don’t these people stop what they are currently doing and start addressing these issues?”
He added: “What's sure is that the best way to influence what AI will do to humanity in the distant future is to make sure we fix the known problems of today.”
https://news.google.com/rss/articles/CBMiamh0dHBzOi8vd3d3LnRlbGVncmFwaC5jby51ay9idXNpbmVzcy8yMDIzLzA1LzA2L3RocmVhdC1hcnRpZmljaWFsLWludGVsbGlnZW5jZS1tb3JlLXVyZ2VudC1jbGltYXRlLWNoYW5nZS_SAQA?oc=5
2023-05-06 10:36:00Z
1996374963
Tidak ada komentar:
Posting Komentar