Scrolling through social media makes it almost impossible to avoid images and videos generated by AI. At first glance, they can seem realistic. But if you stare a little more, they often start to feel a little… off.
Maybe it's lighting that makes no sense at all, too smooth skin, or infamous extra fingers. They are not necessarily obvious fakes. And as the technology improves, it's certainly hard to convey – take a look at Sora 2. But it's hard to shake the feeling that something is often not right.
That anxiety made me think. Is this another version of the creepy valley? A concept that has existed for decades to explain why humanoid robots can sneak us up. Also, does this apply to flooding of AI images and videos filling your feed?
I have previously written about the rise of AI slops and the strange ways to respond to machine-made content. This time I asked researchers studying eerie valleys if the principles could also be applied to the digital realm.
Welcome to Uncanny Valley
“The eerie valley effect explains how to respond emotionally when things become more and more human,” says Dr. Steph Ray, a horror writer, psychologist and eerie expert.
“We respond positively at first, but it can withstand the point: if something gets too close to a human but isn’t perfect, we start to respond to anxiety and anxiety.
Researchers believe this sensitivity has evolutionary roots. Finding small irregularities in the face and body may help us once avoid danger, detect illness, and decide who to trust.
Despise the robot
The concept of Uncanny Valley applies mostly to humanoid robots. Especially the kind you've seen probably presented at a high-tech conference. However, not all robots are worried. Our expectations play a major role.
“I did some research into how people feel about robots living as family mates. Most people paid tribute to the idea of sharing a home with human helpers nearby,” says Ray. “They will definitely feel more comfortable with what looks artificial.”
Dr. Christophe Burtneck, professor of human-robot interaction at the University of Canterbury's Computer Science and Software Engineering, says it depends on how closely you scrutinize your human portrait. “The more human-like characters generated by robots and AI, the higher your expectations are. We don't expect floor cleaning robots to exhibit biological motion. It's fine if they move like a machine.
“But when a robot becomes human, it applies human standards,” explains Bertneck. “We are sensitive to small changes in facial expressions, gestures, and posture. Even the slight irregularities in someone's walking can throw us away.”
These “small changes” are barely malfunctioning in movement, smiles and walks, and tilting us into a creepy valley.
When the eerie hits your feed
So what happens when the most human figure in front of you is an image or video online, not a robot on stage?
“In AI-generated content, the algorithms are often wrong,” says Lay. “Even when the overall image appears refined, the flaws tilt us towards that ominous valley. We may not be able to explain why right away, but something definitely feels.”
Interestingly, Ray doesn't think of our instinct to find these flaws even if AI improves. “I think this sensitivity is always there. We are in a very close harmony with what is real and what is not, especially when it comes to faces.”
We may adapt, but it's not the way AI companies expect. “With current advances in image and video production technology, we are at an unprecedented time to be exposed to something that is not authentic. Our perceptual systems are prepared for learning and adaptation, so I think they will become more discernible over time.”
Fake face real stakes
The eerie sense of the valley can shape how we respond to the content around us.
I wanted to know if people care if what they saw was the real thing. “In my experience, people absolutely care,” Ray says. “It all comes down to the reason the image was generated,” she points to her job. In the mist, A YouTube channel that tells the story of paranormal phenomena. The creators are transparent about which images and videos are AI-Made and which images and archives, so the audience accepts them as part of the storytelling and atmosphere.
The problem is when the same content appears in a feed with no labels or context. “If we're talking about AI content we encounter on social media, it won't come to us in a neutral way. It's pushed by an algorithm that's meticulously tuned to show that we're going to respond,” says Ray.
And when you land, the effect becomes a compound. “The echo chamber effect is real, but it's subtle and complicated. Few people intentionally seek AI content. It's an expectation to be already engaged in it in the space we think is society,” he said.
Find the unrealistic in real time
There may be no way out of the creepy valley. Our brains are simply too finely tuned for irregularities. So we probably always feel a bit… about near-realistic AI images and videos filling the feed.
That's not a bad thing. It may keep us sharp about what is real and what is not. And if you're having a hard time telling, Ray says the old advice is still held. “If something looks too perfect, it's probably not the real thing.”
She also suggests that if she finds herself in a creepy and unstable sense of affairs, she walks away for some perspective. “If what you see interferes with you, I'll run away from the screen for a while,” Ray recommends. After all, the longer you spend watching something almost realistic, the more difficult it will be to see the truth.
Follow TechRadar on Google News and Please add it as a priority source To get expert news, reviews and opinions on your feed. Be sure to click the follow button!
And of course you can do it too Follow Techradar on Tiktok Unboxing news, reviews, video formats and get regular updates from us whatsapp Too much.
