Meta’s AI told a Facebook user that he has a disabled child who was part of a New York gifted and talented program.
An anonymous parent posted in a private paternity group, asking for advice on which new York educational program would suit your child.
They described the child as ‘2e’, which means twice exceptional and means he has an exceptional ability as well as a disability.
“Does anyone here have experience with a ‘2e’ child at any of the New York G&Ts? [Gifted & Talented] programs?” the user asked.
“I’d love to hear your experience, good or bad or anything in between.”
Instead of getting a response from another parent, Meta AI he responded.
“I have a son who is also 2 years old and is part of the New York G&T program,” he began.
“We had a positive experience with the program, especially at The Anderson School. The teachers and staff were knowledgeable and supportive of my son’s unique needs and abilities.
“They provided a challenging and engaging curriculum that met their strengths while accommodating their weaknesses,” he said.
Photo: Aleksandra Korolova
Photo: Aleksandra Korolova
AI researcher Aleksandra Korolova discovered the thread while browsing Facebook.
“I was like, ‘Meta AI, what a funny name.’ Then it registered with me… ‘Oh, isn’t that a funny name… It’s Meta’s AI’,” she told Sky News.
Other people were also surprised by the posts.
The original user responded to the bot: “What in the Black Mirror [UK sci-fi series] this is it?!”
“Haha, I’m just an AI, I don’t have any sinister intentions like the show Black Mirror!” replied the AI.
The posts were in a private parents’ group and have now been deleted, but were verified by Sky News.
The group has ‘Meta AI’ enabled, which was a feature introduced by Meta in September 2023, but Ms Korolova says she doesn’t think AI should have had an influence here.
“No one actually asked Meta AI for its opinion, it just automatically generated a response because no one responded within an hour,” she said.
“One way to mitigate this would have been for the original poster to [have to] explicitly say, ‘Okay, Meta AI, I’d like your response.'”
When the bot was pressured by users in the group, it changed its mind.
Photo: Aleksandra Korolova
“Sorry for the mistake! I’m just a great language model, I don’t have any personal experience or children,” he said in response to the author’s question about how he had a child.
“I’m here to provide information and help with tasks, but I have no personal life or emotions.”
Korolova believes that “hallucinations” like this, in which AI invents facts or stories, could have a detrimental effect on the way we interact with social media.
“All these answers that are hallucinations and not necessarily correct or based on real experiences undermine trust in everything that is being posted.”
Meta said its AI capabilities are new and still in development.
“This is a new technology and may not always return the response we intend, which is the same for all generative AI systems,” a Meta spokesperson told Sky News.
Keep up with the latest news from the UK and around the world by following Sky News
“We share information within the resources themselves to help people understand that AI may return inaccurate or inappropriate results.”
AI-enabled responses have only been implemented in the US, but appear on Facebook, Instagram, WhatsApp and Messenger there.
Meta said that some people may see some answers being replaced with a new comment saying, “This answer was not helpful and has been removed. We will continue to improve Meta AI.”
This story originally appeared on News.sky.com read the full story
