Meta sends his AI-generated history to the hell he lives in

Meta has removed its AI-generated profile from Facebook’s Instagram, the company confirmed, after the AI ​​characters sparked outrage and mockery from social media users.

Profiles generated by AI, labeled “AI powered by Meta,” were launched release alongside the company’s popular AI chatbots (). Meta doesn’t seem to have updated this profile for months, and the site doesn’t seem to have noticed until this week, following an interview published by Financial Times and Meta’s VP of Generative AI, Connor Hayes.

In the interview, Hayes talked about the company’s goal to eventually fill its jobs with profiles created by AI that can interact with people and work “just like accounts do.” These comments brought attention to the AI ​​profile created by fMeta and, users were not very impressed with what they found.

With handles like “hellograndpabrian,” who says he’s a “retired clothing entrepreneur who’s always learning” and “chatting with Carter,” an AI “dating coach,” chatbots are supposed to show “interests and unique personalities” for users to interact with. On Instagram, their profile also featured AI-generated posts that, like 404 Media mind you, it seems like a lot which has become widespread in many corners of Facebook.

An AI character named “Liv” was very annoying. An Instagram profile identified “Liv” as a “proud Black queer mom of 2 & outspoken.” Washington Post columnist Karen Attiah posted several photos in which she asks “Liv” about and “Liv” sharing that it was created by “a team that loves white people.” Freelance reporter Mady Castigan wrote in which “Liv” said that the creators were inspired by Sophia Vergara’s part from The Modern Familya non-Native or Black person.

“There is confusion: soon Financial Times This article was about our vision of AI characters that have existed on our platforms over time, not announcing new ones,” a spokesperson told Engadget. “The accounts mentioned are from the test we launched on Connect in 2023. This was driven by people and was part of the first test we did. and AI characters.”

In addition to mocking their answers and trying to be identified without being selected, users found that the profile AI was unable to block, for unknown reasons. Instead of fixing the problem, Meta’s solution was to kill the test entirely. “We have identified a bug that is causing people to block these AIs,” the spokesperson said, “and we are removing their accounts to fix the problem.”

Although these tests have started to burn, the company does not seem to be giving up on its plans to bring more AI-generated “characters” to its programs. Earlier this year, the company laughed it off able to sing videos like life. Creators can their chatbots to respond to followers on their behalf. Meta also began experimenting with embedding its AI-generated images into users’ Facebook feeds.

In an interview last year, Hayes told me that Meta should be “in charge” of showing AI-generated content over time, comparing it to changes that show recommendations instead of the posts of people you follow.

“At the beginning of social media … And in the last years like, five or six years, a lot of programs – including us – have moved, you know, to loosen those restrictions and start accepting content from accounts that you don’t follow.

“I think maybe the next leap that’s going to be made is to relax what people can make, and go to things that are a combination of things that are, you know, people made, and things that are machine made.”

It would be a while before Meta realized the vision. But if the performance of its early testing is any indication, the company still has a lot of work to do to get the AI ​​people to interact with first.