In all humility and if asked for a little introspection, ChatGPT finds itself to be “effective”, “very versatile”, of great “availability”, even “brilliant”. But hey, we preferred more objective humans, Internet users from 20 minutesto appreciate the benefits, or the wanderings, of the conversational robot which feeds the planetary buzz by writing poems or even manga.
And among them, some have tried it and then adopted it in their everyday life. This is the case of Liam, 19, who uses it “everyday”, for code, like many, and help with homework. “Speed of responses, constant availability”, for him ChatGPT “has almost replaced Google”, to the point that he finds it profitable to pay the 20 dollar subscription for the formula for large users. “You just have to ask him for the sources of his knowledge and half the work is done,” approves Théo, a student who uses the bot in particular to write his cover letters. He also recently obtained a 14 out of 20 by entrusting him with the entire writing of a biology homework, which therefore proved to be convincing despite “a lack of precision and appropriate vocabulary”.
Arguments that fly for a raise
Antoine, a professional developer, is completely convinced by “this superb tool” as long as it is used correctly. “To correct your code and optimize it, it’s crazy to see how well it works! enthuses the computer scientist who also appreciates the pedagogy of the bot. As for Julian*, he finds the robot downright “incredible” and “very rewarding”. In every sense of the term. Because this 30-year-old disappointed with his last increase says that he had the idea of entrusting ChatGPT with writing advice for a text intended to express his annoyance to his hierarchy. “I just changed some manager names to fit my company context and sent it,” he explains. As a result, he was called for an interview with his management. Meeting for which, now confident, he worked on his pitch using the bot. Bingo! Here he is better paid.
But behind this example of resounding success, there are also much more circumspect users. A physics student challenged the bot on a thermostatic assignment. “And if I can give any advice, it’s to keep using your brain!” “, Moves this disappointed, annoyed, like others, by the too polite mea culpa of the AI when it is contradicted, with its famous “You are right, I apologize for the confusion”. Not to mention her versatility, since she generated different reasoning when this tester treacherously submitted the same problem to her three times.
Confusions, errors and mea culpa
Jean-Pierre, a teasing septuagenarian, admits he was surprised when ChatGPT came up with the right answer to a little riddle that required context when he got stuck on a simple arithmetic question. “However, a computer is above all a calculating machine”, he underlines. Lionel, as an attentive family man, tried to help his son to answer a reading sheet on A happy man, by Arto Paasilinna. Review of operations, the AI confused a city and a protagonist, then alternately described the main character as “short and fat” and “tall and thin”. “In the end, I read the book to disentangle the true from the false and help my son,” says the quadra.
Antoine, finally, who we guess is a bit misanthropic, tried without illusion to have a conversation about everything and nothing with the bot, as he would have done with a person. He found it “boring” and “off-putting”. “Always the same thing but said in different ways”.
Existential questions
Whether they are satisfied or not with its services, ChatGPT does not spare, far from it, a certain vertigo to its loyal or occasional users. Theo, our biology student, admits to feeling “always a little guilty” for using it for his writing, at the idea, “that one day or another AIs of this kind could make us totally useless in this kind of task. For John, who has been tempted to refresh his CV, “companies that hope to do without employees thanks to AI are very likely to fail miserably”. Adrien, while acknowledging that ChatGPT knows how to write, translate and code in many languages, recalls fiercely that “commercial intelligence is not artificial”.
* The name has been changed