The Future of Influencers is AI
This month we saw the crowning of the first “Miss AI.” This has implications for creators. It also has implications for women—whether they are creators or not.
The Miss AI contest was the brainchild of an influencer platform called Fanvue, a company that uses living breathing influencers as well as the madeup kind. (I was also surprised to find out the company accepts “adult” AI content, but that’s for another discussion.)
The “pageant” was created to push creative companies into upping their AI game. It was also a cheap PR play. Companies were invited to produce female AI influencers that would be judged on beauty, tech, and social clout—meaning how many followers they racked up and what was their level of engagement. The contest attracted 1500 entries.
As I have noted here before, computer-generated influencers aren’t new. Lil Miquela has been around since 2016. Dozens, if not hundreds, of brands use AI influencers. What is new is the speed and flexibility that AI allows in creating fake people and fake copy. Creating a perfect woman no longer takes art directors and a writing team. A one-”man” band can do it all with a few keyboard clicks.
Not surprisingly, the criticism came quickly. Most of it was in the form of concerns about how this would impact women who are already expected to live up to unforgiving social standards. But, before we dig into the criticism, let’s take a look at who won.
Miss AI
Kenza Layli is a Moroccan Lifestyle influencer. She has more than 200K followers on Instagram and more than 48K followers on TikTok.
On Instagram, the AI figure feels similar to other content on the site. Most Instagram posts are static photography and inserting a person into a scene using Photoshop is an everyday occurrence. On TikTok, the posts are more of a mixed bag. Some are with other people which seems more deceptive than some of the other content. Most of the posts are fashion oriented, which makes sense because Kenza is billed as a lifestyle influencer. The videos use visual tropes from the site like throwing something in front of the camera and suddenly appearing in a new outfit.
I do recommend that you check out the site for yourself so you can see what works and what doesn’t.
What doesn’t work particularly well is Kenza’s acceptance speech. The AI is clunky in this video because she is speaking, which is not the case in most of the videos.
Kenza extolls a utopian view of AI in her acceptance speech (see quote). Almost Godlike, this technology is presented as a doorway to inclusivity. “She” doesn’t specify who is being included, though the implication seems to be women and people of color.
Not according to Dr. Kerry McInerney, a research associate at the Leverhulme Centre for the Future of Intelligence at the University of Cambridge, who told CNN, “These tools are made to replicate and scale up existing patterns in the world…They’re capturing the existing beauty norms we have which are actively sexist, actively fatphobic, actively colorist, then they’re compiling and reiterating them.”
I also found it strikingly odd that Kenza, a Morrocan, sounded like an American. Sure, some would say that is due to the limitations of the technology, but they’ve been saying that for years when it comes to racism embedded in data sets. Just as Joy Buolamwini, the author of Unmasking AI.
In looking at this, I couldn’t help thinking of the quote from Jeff Hammerbacher, who was one of the first data scientists at Facebook. He said, “The best minds of my generation are thinking about how to make people click ads.”
Especially now, when the world feels like it is spinning out of control, is this really what the “best minds” should be working on?