Faced with the frenzy of digital players in their mad race for artificial intelligence, AI researchers are warning of the dangers of this technology by even asking for a moratorium, time to think about the future.
You will have noticed: AI is everywhere at the start of 2023. In the space of a few months, artificial intelligence has invaded the media, digital, technological, economic and even educational, artistic and philosophical spheres, carried by its most emblematic representative: ChatGPT, the now famous and essential conversational robot capable of writing essays as well as writing computer code. The success of this new tool is such that it swept away everything in its path in a handful of weeks, surprising both its creator – OpenAI – and the tech giants – starting with Google, which found itself very helpless when the new Bing came…
Since then, a veritable frenzy has gripped the tech planet and all digital players have embarked on a mad race for AI. Not a day goes by without hearing about a new artificial intelligence-based system, with all the major publishers announcing the current or upcoming integration of AI-powered functions, from Microsoft to Adobe in passing Zoom or Canva, to name but a handful. Not to mention the proliferation of “generative” tools, capable of creating fictitious images and videos from scratch from a simple textual description – a promptin the new jargon.
A pause in the AI race
A race that fascinates as much as it worries. Teachers, of course, who very quickly noticed that their students knew how to use ChatGPT to cheat. But also writers, photographers, graphic designers, and, more broadly, all creative people who fear competition from a software “intelligence” capable of imitating their work, even using their works to learn and progress. . And, more broadly still, all the more or less “intellectual” professions which could be assisted and then replaced by devices based on iA, like what happened for manual trades during the revolution. industrial.
But, without waiting for this “great replacement”, many are worried today about the damage that AIs are already causing, in particular with the faked images and invented information that feed social networks and screens in fake news more real than life. And it’s not a bunch of backsliders dreading technological progress: leading researchers specializing in AI have indeed just published an open letter to alert the world to the abuses caused by AI. These recognized luminaries – such as Steve Wozniak, the co-founder of Apple or Jaan Tallinn, the co-founder of Skype – accompanied by various personalities even ask for a break, proposing that nothing more be taken out of the labs before having redefined rules to regulate the use of artificial intelligence-based systems. “We call on all AI labs to immediately suspend for at least six months the training of AI systems more powerful than GPT-4”specifies the text, which calls for a public and verifiable break in “giant experiments”, including all the key players.
“AI labs and independent experts should use this pause to jointly develop and implement a set of shared safety protocols for advanced AI design and development that are rigorously audited and overseen by independent external experts. “, we read again. “These protocols should ensure that systems adhering to them are secure beyond reasonable doubt.” This call for reflection has already collected more than 1100 signatures. Let’s hope he’s heard and that reason prevails over the race for the new Eldorado…