ChatGPT, the enfant terrible of artificial intelligence, blows out its first candle

ChatGPT, a digital and conversational bot, celebrates its first year of existence this Thursday. Available online from November 30, 2022, this artificial intelligence program, developed by the California start-up Open AI, continues to fascinate as well as worry both policymakers, AI experts and the general public.

Published on:

3 minutes

This anniversary of the birth of ChatGPT offers us the opportunity to take a small step back when the whole world discovered, captivated, the first general public artificial intelligence, referred to as generative, which was able to produce different texts on demand, whatever the subject. or respond in writing and now audio to any question in a fairly convincing manner. Less than a week after its launch, more than a million Internet users on social networks confirmed the performance of the system and confirmed that this online program was absolutely “ bluffing “.

ChatGPT, the revolution” comparable to the appearance of a computer »

But the world’s artificial intelligence specialists were very quick to warn us: ChatGPT is definitely a revolution ” comparable to the appearance of a computer » but the use of this highly sophisticated software, which is available for free online, poses certain risks to Internet users. Thus, the first users noticed that the OpenAI chatbot was not getting its answers, which did not allow them to verify their reliability. The creators of ChatGPT state on their website that the program can generate “ incorrect information “,” or biased content “. Some Internet users believe that these highly sophisticated technologies will inevitably be used by conspiracy theorists to spread false information or by computer hackers to create almost undetectable online scams.

The head of social network X, who co-founded OpenAI, is taking the lead in this global anti-AI movement. ” ChatGPT is scary. We are not far from dangerously powerful AI », Elon Musk wrote a year ago. To reassure its users and especially its financial investors, OpenAI is immediately imposing restrictions on the use of its software to limit possible embezzlement. For example, you cannot ask her to search for a person’s personal information or to formulate questions containing sensitive keywords that would be violent, or to generate texts arguing for conspiracy theories, hateful, racist or xenophobic texts.

ChatGPT, the beginnings of absolute AI rule

The artificial intelligence program OpenAI was immediately adopted by many college and high school students to write homework. ChatGPT would encourage the practice of cheating among students, teachers have warned. Some teachers, on the other hand, believed that this software, if used well, is a formidable educational tool enabling the acquisition of knowledge, the most persuasive of them even recommending strengthening its use in educational institutions. A year later, this debate about AI in schools is not absolutely settled.

Read alsoInformation-disinformation: identifying images generated by artificial intelligence, mission impossible?

Governments around the world are still thinking about how to regulate artificial intelligence technologies to regulate the development of this software, which has as profound implications for our society as the industrial revolution did in its time. Especially when some rumors suggest that AI research labs are on the verge of proposing “ general artificial intelligence » which would be able to compete with the intellectual abilities of people. ” A year after the birth of ChatGPT, we have yet to see anything », predicts most researchers in digital science, who argue that our 21st century will be, for better or for worse, the century of absolute rule by artificial intelligence.

Leave a Comment