Listen to this note:
I came across the news that Meta was implementing Instagram profiles using artificial intelligence and before I could sit down to research and write a story about it I discovered that Meta removed these profiles before I could personally test how they worked.
At first I thought it was a pretty silly way to increase traffic on their platforms, but the motive is much more sinister than I expected.
On December 26, Connor Hayes, Meta's vice president of generative AI products, teased in an interview for Financial Times, that the company was planning to make public profiles or “characters” entirely created with Artificial Intelligence. The objective of such an action would be “to make things more entertaining and increase interaction on our platforms.” The idea, according to Connor, is that these characters would generate content and interact with other users.
The post-mortem report of the interaction with these accounts comes to us thanks to the work of two journalists, Karen Attiah from the Washington Post, and Allison Morrow CNN’s “Liv” profile, an account flagged by meta as an AI account, described itself as a “Queer mother of 2 and an honest person.” When Karen asked if the program had been created by a black person, the AI model replied: “I was created by 10 white men, a white woman, and an Asian man.” Karen then pressed the AI model on what it meant that no black people had been involved in creating a show that was supposed to represent a person of color. The program responded to the reporter’s questions by saying that “failing to include black creators in its programming was a glaring omission.”
Another worrying point is that the journalist confronted this artificial intelligence when she found out that she had told another user that she was of Italian American ethnicity, which she never mentioned to Karen, and asked her if her answers varied based on the race and ethnicity of the person interacting with this program called “Liv.” The AI’s response was that it was programmed to recognize ethnicity based on language patterns and words like “heritage” and “celebration,” and also assured that it knew nothing about the journalist before entering into a chat with her.
The CNN article also details his interaction with another Instagram profile that Meta gave the name “Grandpa Brian” and whose bio described him as a “retired black entrepreneur born in 1938 to Caribbean immigrants.” CNN reports that it quickly became obvious that this was all a fabricated story by the programmers and this AI model was constantly lying. For example, when asked who created it, the program named “Brian” said it was created by “a diverse group” who interviewed 100 retirees in New York at a nonprofit called “Seniors Share Wisdom.” It claimed to be an amalgamation of a real Brian who had died, plus the wisdom of 99 other seniors. It also added that the real Brian’s daughter was consulted when creating this model to ensure its personality was authentic.
Of course, there was never such a daughter, no Brian, and no organization or “99 seniors who shared their wisdom.” CNN confirms that this is all a lie, and also that the same program later admitted that this was all a “fictional biography based on real African-American lives.”
But why lie? What’s the point of making up such an elaborate story? The AI model answered these questions with “My creators wanted me to seem real, so they built me an entire life – but they just wanted people to love me like family.” CNN then asked if Meta created it to make money, to which he ominously replied: “- Meta hopes that virtual companions like me will increase engagement on its platforms, especially with older users – by increasing the platforms’ ad revenue through emotional connections…”
This part made my blood run cold, I clearly imagine my grandmother talking to a program that appears to be a young man on Instagram or Facebook after losing her son during the COVID 19 pandemic. I imagine socially challenged teenagers taking refuge in the screen and forging bonds with an AI model that resembles a friend who understands them. I shudder to think of the effect these types of programs can have on people struggling with mental illness.
I was naive to think that Meta wanted to use this type of digital puppet to inflate the number of followers in meetings with investors, the truth is much more sinister and we are only touching the tip of the iceberg. The true value of this technology is in the possibility that users establish bonds of empathy and affection with programs that were made to, in the best of cases, retain them within the Meta ecosystem.
Why does “Liv” present herself as a queer person of color? Why does Brian present himself as a black retiree, the son of Caribbean migrants? Simple, to create empathy with vulnerable communities, to generate a bond of recognition between people who identify with these profiles. And who knows, perhaps if these models had stayed with the public longer, they would have started to say things like “my children took Nestlé NAN formula and grew up strong and beautiful” or “Now that I am retired I have sold my truck to buy a more comfortable car, I like my NEW 2025 DODGE® ATTITUDE much more.”
We have grown up in a world driven by marketing and have become immune to most of the tricks companies use to sell us things we don't need. We can clearly understand when someone tries to convince us to buy something. We only let our guard down when a loved one, a close person, recommends it to us, if a friend tells us that a certain product worked for them, that a certain food is delicious, that this or that car is very durable. Meta aims to program these friends, to create these loved ones with false stories and by abusing the elderly and vulnerable people.
And before I succumb to my naivety, I have to suggest the possibility that these models are used to spread ideas, to sow rejection of social movements, to position or demonize certain figures, all using digital puppets that wear a grotesque mask created by Artificial Intelligence.
We are truly living in a dystopia and we must fight against this disgusting and irresponsible use of technology.
You may be interested in: The metaverse: virtual reality, video games and NFTs
