Using AI to bypass reason
A weird genre of scammy posts anchored by bizarre AI images says more about humanity than technology.
Social networks are filled with AI-generated images. Billions have been created using text-to-image AI tools since 2022, many posted online. To roughly quantify:
26% of marketers use AI to create marketing images. 39% for marketing images posted on social.
71% of images shared on social media were AI-generated. In Canada, it's 77%.
AI is ubiquitous on social media. But there's one genre that fascinates me. It’s all over Facebook.
The genre uses bad AI-generated images, often incredibly bizarre, to elicit a knee-jerk emotional reaction from people apparently devoid of rational filters.
In Facebook posts, these "engagement bait" pictures are accompanied by strange, often nonsensical, and manipulative text elements, which I'll describe below.
Although they have much in common, they often fall into categories (these are my category labels) which include:
Smoldering White Jesus in the Modern World
Kitch Jesus vs. Kitch Devil
AI-generated Pareidolia
Military Vets Deserve Better
Suffering or Disabled, but “Today is My Birthday”
Some of the pictures are delightfully nonsensical, especially the Jesus ones. A sculpture of Jesus and his disciples being carried out of the ocean on a log. Jesus walking in Tokyo while texting. Jesus enjoying a single-malt scotch. And my favorite: Jesus making war with the Roman army (my Biblical education is a little rusty, but wasn’t Jesus a pacifist crucified by the Romans?).
The words
The posts often include very strange words. Posters almost always hashtag celebrity names. Many contain information about unrelated topics, like cars.
Many such posts ask, "Why don't pictures like this ever trend?"
They often have lines with nothing on each line except a period to push down the hashtag segment of the post from initial view.
The Jesus vs Devil subgenre often shows signs that say "Amen" and "Skip." The apparent idea is that if you scroll past the post without commenting, "Amen," the devil made you do it. Tens of thousands comment, “Amen.”
Many Military Vets Deserve Better posts show old soldiers in uniform holding a sign. Sometimes, they imply that men far past retirement age are being forced to serve abroad or that the US is currently engaged in large-scale war. (Around 3,400 US troops are currently deployed or stationed in active combat zones.)
According to one report, all these hyper-patriotic engagement-bait photos are posted by users outside the United States.
Some call for veterans to live tax-free or imply that student loan forgiveness is taking away from veteran benefits. I suspect these may be of Russian origin, part of a vastly larger effort to divide US society and other democracies.
One account under the name "Daisy Freeman," who posts Jesus vs Devil posts, says this in every post:
Close your eyes 70% and see magic
Today's my graduation
May 2024 is Your Best Year
#art #artist #artgallery #painting
Nothing about this makes logical sense. Partially closing eyes doesn't do anything. It's not her graduation. May is not a year. And it's not a painting. The posts get huge engagement.
Another account specializes in bad AI images of suffering children with a birthday cake and a sign in the picture that says it's their birthday. Some kids have three eyes or three heads, a fever-dream of AI hallucination.
In one image, a girl is floating in water. A sign says, "Today Is My Birthday," and the stick that holds the sign is not stuck into the ground but into her body. She's hooked up to breathing tubes. A birthday cake with candles is floating in the water. And I wonder: What is the mindset of people who comment on an image like this with "Happy birthday!” instead of “who’s lighting that candle?”
You get the idea. These bizarre posts, anchored in bad AI, are rife on Facebook.
Sure, anything and everything exists somewhere online. But what's perplexing is how successful these posts are. Many of them get hundreds of thousands of views and tens of thousands of likes and comments.
Facebook's genius algorithms decided that I'm someone who wants to see them. For months, I blocked everyone one I saw. The more I blocked, the more new ones appeared. A third of the posts in my Facebook feed are made of such posts. Now that I’ve been researching and exploring these kinds of posts, I assume I’ll be flooded with them until the end of time.
What this says about Facebook and humanity
My first thought is to reverse-engineer the operating system here and understand the working formula, which appears to be:
Use the Appeal to Emotion fallacy to fish for people who autonomously respond to emotional appeals. Reason is irrelevant.
Find emotional but abstract concepts (like good vs. evil) and make them "real" with fake AI photos.
Juice up the emotion with a sentence implying the (non-existent) person in the photo needs a boost and the reader's input.
Throw celebrity mentions and hashtags to troll for superfans captured in a parasocial relationship with a celebrity.
Hide the celebrity references by pushing them down out of sight.
These obvious tactics effectively game Facebook's algorithms, showing that Meta is asleep at the wheel.
The fact that these obvious scam posts actually work shows that the world is not ready for fake images, fake people, fake everything.
Facebook can’t or won’t stop it. Which makes me wonder: Will the weird world of scammy, emotion-tugging fakes make it over to Zuckerberg’s metaverse?
Shameless Self-Promotion
What North Korea’s infiltration into American IT says about hiring
Apple embraced Meta’s vision (and Meta embraced Apple’s)
The rising threat of cyberattacks in the restaurant industry
Apple’s planned chatbot should have no ‘personality’
More from Elgan Media!
My Location: Marrakesh, Morocco
(Why Mike is always traveling.)
Worst of it all is how many people can't see how fake these images are, but they probably still have velvet Elvis on their walls.
Dear Mike,
I am writing to you today as a long-time reader and admirer of your work. I have always appreciated your insightful commentary on AI and technology, as well as your willingness to tackle tough political issues.
However, I believe that you may be the victim of your own success. Your posts on AI and tech, as well as your hard left political bias, are feeding the algorithm. As a result, you are increasingly being shown content that confirms your existing beliefs. This can lead to a dangerous echo chamber effect, where you are only exposed to information that reinforces your own views.
I urge you to be more mindful of the algorithm and to make an effort to seek out content that challenges your existing beliefs. This will help you to stay informed and to make more well-rounded decisions.
Thank you for your time.
Sincerely,
Benjamin Roche
YES - That response was the result of an AI prompt but I approve this message ;-)
Ben