Tech knowledgeable says ‘existential’ fears from AI are overblown, however sees ‘very disturbing’ office threats

Photo of author

By Stacy Connor

A U.Ok.-based tech knowledgeable mentioned he isn’t dropping sleep at night time over the latest progress of synthetic intelligence however argued he does have issues over AI doubtlessly turning into a hellish boss that oversees an worker’s each transfer. 

Michael Wooldridge is a professor of pc science on the College of Oxford who has been a number one knowledgeable on AI for not less than 30 years. He spoke with The Guardian this month relating to upcoming lectures he’ll lead this winter to demystify synthetic intelligence, whereas noting what issues he does have with the tech. 

He informed the outlet that he doesn’t share the identical worries as some AI specialists who warn the highly effective programs may sooner or later result in the downfall of humanity. As a substitute, considered one of his issues is AI morphing right into a hellish boss that screens workers’ emails, gives fixed suggestions and even maybe decides which human workers to fireplace.  

“There are some prototypical examples of these instruments which can be obtainable right now. And I discover that very, very disturbing,” he informed The Guardian.


AI illustration

Synthetic intelligence illustrations are seen on a laptop computer with books within the background on this illustration picture on July 18, 2023. (Jaap Arriens/NurPhoto through Getty Photos)

AI has already staked its declare in a handful of industries, reminiscent of serving to medical leaders diagnose most cancers, or detecting fraud at monetary firms, and even drafting authorized briefs that cite related case legislation. 

“I do lose sleep concerning the Ukraine warfare, I lose sleep about local weather change, I lose sleep concerning the rise of populist politics and so forth,” he mentioned. “I don’t lose sleep about synthetic intelligence.”

Wooldridge defined to Fox Information Digital in an electronic mail that “existential issues about AI are speculative” and that “there are very way more fast and concrete existential issues proper now.”

“Prime of those is escalation in Ukraine – that’s a really actual risk meaning nuclear warfare is definitely nearer now than at any time in 40 years. So, if one desires to lose sleep over SOMETHING, I feel that could be a way more essential subject,” he mentioned. 



Wooldridge did say that the proliferation of AI and its progress in intelligence does convey different dangers, reminiscent of bias or misinformation. 

“It may well learn your social media feed, decide up in your political leanings, after which feed you disinformation tales in an effort to attempt to get you for instance, to vary your vote,” he mentioned.


Wooldridge, nevertheless, mentioned customers ought to arm themselves in opposition to such dangers by viewing AI by way of skeptical lenses and argued firms behind the tech must be clear with the general public.

“I don’t low cost existential issues about AI, however to take them actually significantly would wish to see a genuinely believable situation for the way AI would possibly signify a risk (not simply “it may be cleverer than us”),” he added in remark to Fox Information Digital. 

Building on Oxford University campus

The Tower of The 5 Orders, Bodleian Library, Oxford College, England, Aug. 12, 2014. (Getty Photos)

The Oxford professor will lead a prestigious U.Ok. public science lecture collection this December, the Royal Establishment Christmas lectures, which has explored varied scientific subjects because it was launched in 1825. He’ll sort out explaining synthetic intelligence to the general public this 12 months, highlighting that 2023 marks “the primary time we had mass market, basic objective AI instruments, by which I imply ChatGPT.” 

“It’s the primary time that we had AI that feels just like the AI that we had been promised, the AI that we’ve seen in films, pc video games and books,” he mentioned. 

ChatGPT, the favored chatbot from OpenAI that may mimic human dialog, exploded in use this 12 months, gaining 100 million month-to-month lively customers by January, which set a file because the fastest-growing platform. 

“Within the [Christmas] lectures, when folks see how this expertise really works, they’re going to be stunned at what’s really happening there,” Wooldridge mentioned. “That’s going to equip them a lot better to enter a world the place that is one other instrument that they use, and they also received’t regard it any in another way than a pocket calculator or a pc.”

READ MORE  Bitcoin Soars on Hopes of Funding Fund Approval


The lectures will embrace a Turing take a look at, which investigates whether or not AI demonstrates human-like intelligence. People could have a written dialog with a chatbot, and if they can not inform if they’re corresponding with a human or chatbot, this might present AI has matched human-like intelligence, The Guardian reported. 

ChatGPT student cheating illustration

Illustration image reveals the ChatGPT synthetic intelligence software program, which generates human-like dialog. (Getty photographs)

Wooldridge, nevertheless, pushed again that the take a look at shouldn’t be finest suited to make such a dedication. 

“A few of my colleagues suppose that, mainly, we’ve handed the Turing take a look at,” Wooldridge informed The Guardian. “Sooner or later, very quietly, within the final couple of years, the expertise has bought to the purpose the place it might probably produce textual content which is indistinguishable from textual content {that a} human would produce.”

“I feel what it tells us is that the Turing take a look at, easy and delightful and traditionally essential as it’s, shouldn’t be actually a terrific take a look at for synthetic intelligence,” he added.


The Christmas collection will start filming on Dec. 12 earlier than it’s broadcast on BBC 4 between Christmas and New Years. 

“I need to attempt to demystify AI, in order that, for instance, when folks use ChatGPT they don’t think about that they’re speaking to a aware thoughts. They aren’t!” Wooldridge informed Fox of the upcoming lectures. “While you perceive how the expertise works, it provides you a way more grounded understanding of what it might probably do. We must always view these instruments – spectacular as they’re – as nothing greater than instruments. ChatGPT is immensely extra subtle than a pocket calculator, however it has much more in widespread with a pocket calculator than it does a human thoughts.”

Supply hyperlink

Leave a comment