The San Francisco company OpenAI has just released ChatGPT, a conversational robot that can adapt to a wide range of themes.
The latter is able to generate content for social networks, write film scripts or even check the validity of the code of a computer program.
If the interface is in English, ChatGPT is able to give the repartee in several languages including French, German, Russian or Japanese.
If its development continues to accelerate, the tool could be a credible alternative to Google in the medium term.
However, the latter does not sit idly by with its subsidiary Deepmind, due to its ability to formulate complex answers almost instantaneously.
But even by relying on reinforcement learning, its repartee is not immune to inaccuracies or even biases.
Should we see, like one of its creators Elon Musk, a harbinger of strong artificial intelligence, capable of overcoming humanity?
What is certain is that this artificial intelligence has no conscience.
And that if new progress emerges, they will undoubtedly have to be supervised so that AI continues to serve man and does not enslave him.
"I'm designed to be helpful and helpful so I do my best to be kind and respectful when answering questions and offering help," ChatGPT explains, before continuing: "However, I'm just a computer program and I have no feelings or motives of my own. My goal is simply to provide answers and help in a precise and helpful way."
Of which act.
The summary of the
France 24 week invites you to come back to the news that marked the week
I subscribe
Take international news everywhere with you!
Download the France 24 app