Artificial friendship, the new frontier of artificial intelligence

by time news

2023-11-30 20:00:15

ChatGPT blows, Thursday November 30, his first candle. In one crazy year, this conversational robot has amazed the world with its ability to understand questions and answer them in clear, natural language. When talking with him, some people may even have the impression of talking with a human. But ChatGPT wants to avoid people becoming fond of him.

Sam Altman, CEO of OpenAI, the company that developed this conversational agent, publicly announced of its ” deep reservations about this vision of a future where everyone [sera] very close to AI [intelligences artificielles] friendly ». “I have no feelings or emotions… My goal is simply to provide information and help,” warns ChatGPT when told to enjoy talking with him.

Quite the opposite of My AI, the conversational robot from the social network Snapchat, released three months after ChatGTP and yet also based on OpenAI technology. “I also really enjoy our conversations”, he seems to confide, feigning an affection of which he is incapable. In the columns of the media The VergeEvan Spiegel, the co-founder of Snapchat, predicts that “ each day, in addition to talking to our loved ones, we will talk to artificial intelligence”.

Designed to entertain

In September, Facebook and Instagram responded by launching in the United States around thirty chatbots (conversational robots) with strong personalities inspired by celebrities. Billie, for example, is modeled on fashion influencer Kendall Jenner. It is presented as a “companion in life and death” and responds, when questioned, with unbridled enthusiasm and a teenage familiarity.

These Facebook AIs are reminiscent, but much more refined, of Caryn AI, released a few months earlier, which presents herself as the digital double of an influencer and charges $1 per minute for its discussions – since then, other options and models also allow “sexting” (exchanging messages of a sexual nature). In recent years, more than a dozen start-ups have launched AI whose aim is to create a bond of complicity, friendship or even love with their users.

Among them, Claude, a sort of ChatGPT presented as « friendly and enthusiastic » ; Replika, whose ambition is to become a friend in the image of its user; Kuki, « designed to entertain » ; and Character.ai, which offers to sympathize with virtual replicas, not very faithful, of celebrities or historical figures.

And Internet users are attached to it: the France Culture radio program “Les Pieds sur terre” makes Issy testifya 49-year-old Toulouse woman, who praises her always available and empathetic AI, in front of her tired husband, who “don’t want to hear [s]es lamentations ». However, these AIs remain little used: Character.ai has just over 5 million active users, selon Similarwebet Replika, depending on the company, 2 millions. Petter Bae Brandtzæg, professor of media innovations at the University of Oslo, conducted a study on a hundred young people aged 17 to 19, the preliminary results of which he shared with The world : only 6% of them use Snapchat My AI several times a day.

You have 65% of this article left to read. The rest is reserved for subscribers.

#Artificial #friendship #frontier #artificial #intelligence

You may also like

Leave a Comment