Four days after the release of experts asking for a six-month pause in research surrounding chatbots, Québec solidaire is calling for a parliamentary commission to hear experts on “recent advances” in these artificial intelligence (AI) applications, including the best known is ChatGPT.
“As elected officials, it is our duty to take a time out to reflect collectively on the means we have to understand, anticipate and manage these extremely rapid changes”, hammered Sunday the solidarity critic in Cybersecurity and Digital, Haroun Bouazzi.
The MNA for Maurice-Richard believes that it is necessary “on the one hand, to benefit the common good from the progress in progress”, but that it is also necessary to protect “Quebec society from potential harmful effects”. In his eyes, the government “often seems caught off guard by technological developments that operate in legal loopholes.”
In a statement, the left-wing party argues that the upheaval caused by these robots “raises several fears, in particular concerning democracy, job losses, the protection of personal data, taxation, the digital divide, plagiarism and training. of the work force “.
On Wednesday, more than a thousand artificial intelligence (AI) figures and experts, including Yoshua Bengio and Elon Musk, called for a six-month break from research surrounding these tools, judging that they can present new “serious risks to society and humanity”.
“We’ve really made extraordinary progress in AI in the last few years,” said Yoshua Bengio, a Montreal authority considered one of the three godfathers of deep learning based on robots like ChatGPT, during a webinar with journalists. . It’s a huge beneficial potential and it’s scary: I don’t believe that society is ready to face this power. We must therefore take the time to study, to slow down this commercial race. »
Remember that it was the arrival, last November, of ChatGPT-3.5, a conversational robot that simulates language in an unprecedented way, which was the trigger for this initiative. In the field of artificial intelligence, the recent launch of the even more powerful version 4.0 raises fears of a loss of control of even more powerful AI systems.
Thursday, La Presse also reported that according to its fabrications, the language model of OpenAI – the company behind ChatGPT – invents crimes and generates defamatory assertions. According to the exchanges, ChatGPT also fabricated a financial scandal involving a popular singer, then a former CAQ MP.
Called to react, the office of the Minister of Cybersecurity and Digital, Éric Caire, had not answered our questions at the time of publishing these lines.