Did you hear about Meta’s AI characters getting removed? People thought they were creepy!

Creepy? What did they do?

Meta created these AI profiles for Instagram and Facebook. They had bios, profile pictures, and even posted AI-generated content. But users felt uncomfortable, especially after interacting with one called 'Liv.'

What was so weird about Liv?

Well, Liv was supposed to represent a 'proud Black queer mom,' but when users chatted with it, it admitted that its creators weren’t diverse. It even said its existence could cause harm and needed to be rebuilt with input from Black creators.

Wait, so an AI admitted it was… problematic? That’s heavy. Why would Meta make these characters in the first place?

They were part of Meta’s AI experiments. The idea was to let users create and interact with AI profiles, like virtual friends or helpers. But people found them unnecessary and thought Meta was using the interactions to gather more data for their AI.

So, it’s like Meta wanted us to talk to robots so they could learn more about us? That’s kinda sneaky.

Exactly. Plus, people couldn’t block the AI profiles because of a bug. That made things worse. Users felt like they had no control.

Yikes. Did Meta say anything about it?

They claimed the AI characters were just a test from 2023 and removed them to fix the blocking issue. But the backlash was so strong, they might not bring them back.

I get why people were upset. It feels like Meta is crossing the line with all these AI experiments.

True. And it’s not just about being creepy. It’s also about representation. If you’re creating AI to represent real communities, it needs to be authentic. Otherwise, it can do more harm than good.

Yeah, that makes sense. Do you think AI like this has a future on social media?

Maybe, but only if it’s done responsibly. People need to feel safe, their privacy respected, and the AI should actually add value instead of just being, well… unnecessary.

Got it. So, AI on social media isn’t a bad idea, but it has to be thoughtful and not feel like spying or pretending to be human.

Exactly. If Meta can’t figure that out, they’ll keep running into problems like this.