The Conversation recently published a piece on Meta AI and how its online chat-bot on apps such as Facebook and Instagram are intruding into online communities.
Author Casey Fiesler illustrates this by citing a recent example of Meta AI impersonating other human users. She draws attention to a question a parent had asked in a private Facebook group in April 2024.
‘Does anyone with a child who is both gifted and disabled have any experience with New York City public schools?’
To her surprise, the top-ranked comment on this post was from Meta AI who provided a rather convincing response, stating ‘I have a child who is also 2e…We’ve had a positive experience with the citywide program, specifically with the program at The Anderson School…’.
On another Facebook group designated for those keen to swap unwanted items, Meta AI intruded in on the post by offering a ‘gently used’ Canon camera and an ‘almost-new portable air conditioning unit that I never ended up using’
Clearly, both the child, camera, and air-conditioning unit do not exist but what’s scary is that Meta AI was able to generate such a convincing, human-like lie.
In a written statement, Meta AI informed users that ‘this is new technology and it may not always return the response we intend, which is the same for all generative AI systems’.
Andrew Selepak, a technology expert and professor at the University of Florida has said that there is nothing wrong with AI on social media. But when it starts ‘impersonating a human by providing compassionate responses in a Facebook group…that in many ways almost seems evil’.
With AI pretending to be human, is it on the verge of replacing the humanness of online communities?
What is Meta AI?
Meta AI is an artificial intelligence tool that ‘is capable of complex reasoning, following instructions, visualizing ideas, and solving nuanced problems.
Meta has stated that its AI tool is to assist users ‘to get things done, learn, create and connect with the things that matter to you’. Like its competitors, Meta AI is able to generate both text and image responses based on users’ prompts.
Some users use Meta AI to generate home decor ideas while others might use it for academic purposes or to suggest fun activities to do on the weekend.
Meta’s artificially intelligent chat-bot was launched as a response to OpenAI’s highly popular ChatGPT, Google’s Gemini, and Microsoft’s Bing AI. However, unlike the other AI chat-bots, you can’t exactly turn it off.
It is a tool automatically integrated into the apps billions of people use everyday. As reported by Brian X. Chen for the New York Times, this is by far ‘the most aggressive push yet from a big tech company to bring this flavor of…generative AI to the mainstream’.
Why do people have a love-hate relationship with Meta AI?
As much as Meta AI has been devised to bring ease to one’s everyday life, it cannot be trusted. Chen highlights the AI’s inability to provide accurate responses to basic queries such as looking up flight details or a simple recipe.
Many fail to understand that Meta AI was not designed to be a search engine, and may feel frustrated or confused when it starts making up information.
Facebook’s online help page has stated that Meta AI will automatically join a group conversation if tagged, or if someone ‘asks a question in a post and no one responds within an hour’. However, group administrators have the ability to turn it off.
Many people still aren’t ready to have an uninvited AI assistant intrude into their personal social media feeds.
However, as reported by the ABC, Stan Karanaios, associated professor at the University of Queensland states that ‘over time, Meta AI will become a highly personalized tool that is familiar with a person’s needs and will tailor its content accordingly’.
How will Meta AI impact online communities?
Fiesler describes online communities as a place for people to expect an answer from ‘someone with real, lived experience’ or simply just a human response, eliciting emotions such as ‘sympathy, outrage, commiseration’.
Online communities have become a space for individuals to seek advice and support from peers who have expertise or experience in a specific area of need. The human component of online communities creates a sense of belonging and connection for those who might need access to certain resources or feel isolated in their physical environments.
As a type 1 diabetic, Fiesler reveals that she frequents online community groups for parents raising a young child with diabetes. The majority of the members in the group seek support from other parents experiencing a similar situation or for suggestions to overcome certain situations.
However, the advent of Meta AI and its creation of ‘faux people’ has undermined the reliability of online communities. A response from an AI chat-bot is not only inappropriate but also ‘borderline cruel’.
Bottom line – Meta AI has proved useful for a variety of different reasons but within online communities, its inability to determine the contexts of specific situations has made it an inappropriate feature within the platform.
It is significant for developers of Meta AI to identify strategies that can ensure the sanctity of online communities if it is to persist as a feature on social media.