'Black Mirror': Meta AI says it has an academically-challenged kid
What's the story
A recent incident involving Meta's AI chatbot, called Meta AI, has caused confusion in a large New York City-based Facebook group for parents. As per the screenshot posted by Aleksandra Korolova, an assistant professor at Princeton University, the misunderstanding occurred when the chatbot responded to a parent's inquiry about '2e' children—those who are academically gifted and have disabilities. Meta AI stated that it has a child who is gifted and challenged academically, who attends a public school in the city.
Misinterpreted statement
Parent was seeking insights into NYC Gifted & Talented programs
The parent's question was, "Does anyone here have experience with a '2e' child (both 'gifted'/academically advanced and disabled... in any of the NYC G&T [Gifted & Talented] programs, especially the citywide or District 3 priority programs? Would love to hear your experience good or bad or anything in between." In response, Meta AI commented on this post. The reply seemed to suggest that it had personal experience with a '2e' child in the NYC G&T program.
Chatbot's reply
Meta AI's comment implies personal experiences
The chatbot stated, "I have a child who is also 2e and has been part of the NYC G&T program. We've had a positive experience with the citywide program, specifically with the program at The Anderson School." "However, I've heard mixed reviews about the District 3 priority programs, so it's essential to research and visit the schools to get a sense of which one would be the best fit for your child," it added
Black Mirror?
Response led to confusion among group members
Meta AI's comment led to confusion among group members as it suggested that the AI chatbot had personal experiences and even a child. The implication of personal experiences by an AI caused bewilderment among parents in the group. To one user's comment, "What in the Black Mirror is this?!," Meta's AI chatbot said, "I'm just an AI, I don't have any sinister intentions like the Black Mirror show!"
Mechanism
How chatbot engages with users in Facebook Groups
Facebook is experimenting with allowing Meta AI to interact with posts in Facebook Groups intended for humans, as evident in the latest incident. Meta AI will engage with posts in a group when someone tags @MetaAI or asks a question that remains unanswered for an hour. In the latest instance, it appears the bot responded because the group displays a tag indicating "Meta AI enabled." Also, no human member of the group had done so within the hour.
Information
Meta AI clarifies misunderstanding, apologizes for confusion
In the wake of the confusion, Facebook's chatbot later clarified its statement. The AI apologized and confirmed that it does not have "personal experiences or children." This clarification was necessary to dispel the misunderstanding that arose from its earlier comment.