“Artificial intelligence systems can pose serious risks to society and humanity” and therefore “we invite all artificial intelligence laboratories to immediately suspend training for at least six months». Signed Elon Musk.
Elon Musk, Steve Wozniak and Yuval Noah Harari
The new broadside against Ai comes from leading figures in the sector, first of all the owner of Tesla. He is the one who is among the founders (but has now left the board) of that OpenAI (his story here) which gave life to ChatGpt, the artificial intelligence which was the first to let us experience the power of “pseudo-human” thinking. He who uses artificial intelligence on his cars for the Autopilot and autonomous driving system two years ago he presented robots Tesla Bot and Optimus. Then there are the founder of Apple Steve Wozniak, the bestselling essayist Yuval Noah Harari and many other experts such as Yoshua Bengio, of the University of Montreal, and Stuart Russell of Berkeley, one of the major protagonists of the revolution.
«Let’s stop at Gpt-4»
The names appear in an open letter posted online calling for a moratorium for all systems more powerful than Gpt-4, the latest version of the system that is incredible. It goes beyond the texts and can also create images, it can create a videogame like pong (the progenitor of the sector) in less than a minute, draw a website from a simple diagram drawn on a paper and create a recipe “looking” at the photo of the ingredients (as well as many other things, as we tell you here).
The reasons for the request
It is precisely the cultural stature of the signatories that gives strength to a letter that takes on the classic tone of «what will happen if it falls into the wrong hands» already seen for so many innovations. “Powerful AI systems should only be developed when we are sure that their effects will be positive and their risks will be manageable», reads the text, and hence the request for a stop in development for six months «until shared security protocols for such projects are developedimplemented and verified by independent experts».
Why right now
The date of publication of this letter is not accidental. On the one hand, even non-experts have now become aware of the enormous potential of artificial intelligence and of «dramatic economic and political disruptions (especially for democracy) that AI will cause», as stated in the letter. We got a taste of it with the realistic photos of Donald Trump’s fake arrest or the Pope’s duvet. Many believed them to be true (and there are those who passed them off as such) yet they had been created by an AI.
In addition, last Monday Europol published a document showing the abuses of Gpt 3.5 (the previous version of Gpt-4 on which the current ChatGpt is based) and how it could be used by cybercriminals to generate phishing emails without spelling and grammatical errors, generate credible texts of propaganda and disinformation, independently program software by compiling code through numerous programming languages. Software that could also include malware of course.
Sam Altman also calls for caution
In all of this Sam Altman, the brilliant CEO of OpenAI (his story here), has not yet spoken but basically he had already had his say. On February 24, he signed a post in which he underlines «a serious risk of misuse, accidents and social unrest» of «Agi», the general artificial intelligence that emulates human thought. Altman admits that the benefits «are so great that we do not believe that it is possible or desirable for society to halt its development forever» and therefore he and his team are becoming «increasingly cautious in the creation and dissemination of these models».
The Terminator scenario
Its fears are the same as those of Europol (abuse, computer crime, disinformation, propaganda) and can be summed up in a brilliant definition, the “Terminator scenario”. Theorized, among others, by Altman and Musk, it recalls the apocalyptic scenarios of James Cameron’s film, that is what could happen to a humanity rendered obsolete from the development of an artificial intelligence endowed with consciousness and cognitive abilities superior to ours.
And then there’s Microsoft
There is another curious aspect of the punctuality of this letter. It just comes in the aftermath of the alliance between OpenAI and Microsoft of February 7 (details here). Despite Bill Gates’ fears, Satya Nadella’s company has invested 10 billion in Altman’s company to clearly beat the competition (especially Google’s AI Bard, here is the comparison between the two). He has already implemented ChatGpt in his Bing search engine (with some oddities), in Teams and in Windows 11 and is now working to bring artificial intelligence to Office 365 as well. ..