Skip to content

Five Minutes of Dread · Episode 28 · 9 min · English audio · 9 September 2026

Unsettling in Five: The Chicken Nugget That Wasn't

How AI-generated food photos trigger the uncanny valley and leave us queasy. Ordinary sights, deeply wrong.

From the show

Five Minutes of Dread

Something unsettling each week in under five minutes — no gore, just the kind of dread that lingers after the audio ends.

Stories for Adults · 33 episodes · Latest 28 October 2026 · English audio

Make a show like thisWith the Lissin app for iPhone

What this episode covers

An unsettling short about the uncanny valley, starting with a Hollywood food stall whose AI-generated pictures of chicken nuggets went viral because they looked almost, but not quite, like food. Japanese roboticist Masahiro Mori coined the term in 1970: as a machine becomes more human-like we like it more, until, close to human, our affinity plunges into revulsion. The episode sets out three explanations: an instinct to avoid corpses and disease, a brain unable to categorise what it sees, and mismatched cues. It notes the effect applies to sound too, and ends on face filters and AI voices that are climbing out of the valley.

Key points

  • A Hollywood food stall's menu went viral for AI-generated pictures of chicken nuggets that did not look like food.
  • The term "uncanny valley" was coined in 1970 by Japanese roboticist Masahiro Mori.
  • Mori's graph shows affinity rising as a machine looks more human, then plunging when it gets very close to human.
  • One theory is perceptual mismatch: cues such as skin without pores or eyes that blink wrongly do not add up.
  • The episode cites a 2026 study that found the most human-like synthesized voices were rated more likable and less eerie.

Sources for this episode

Play this episode

9 min of audio, free in your browser — no account, no app.

Transcript

1,081 words · Script-based transcript; wording may differ from the recording.

In Hollywood, a food stall's menu went viral for its pictures of chicken nuggets. Last week, we talked about the unsettling feeling that comes from ordinary habits gone wrong. This is about an ordinary sight, gone wrong in a way your brain can’t quite name. The menu was displayed on a glowing screen at a night market. It showed pictures of food. But people stopped. They stared. They felt a queasy, visceral sense of wrongness. One person said, “Those don’t look like chicken nuggets. Those don’t look like typical food that we eat.” The images were generated by artificial intelligence.

The pasta looked like plastic. The chicken had a rubbery sheen. Another person in Berlin, looking at similar images, said, “There is something wrong with it. It looks too geometric. I imagine that it’s like if you take acid and food stops looking like food.” The food was almost real. And that is PRECISELY what made it so repulsive. This feeling has a name. It’s called the uncanny valley. The term was coined in 1970 by a Japanese roboticist named Masahiro Mori. He was thinking about prosthetics and robots. He proposed an idea, a graph.

On one axis, you have how much you like something, your affinity for it. On the other axis, you have how human-like it is. As a machine becomes more human-like — say, from an industrial arm to a toy robot — our affinity for it goes up. It gets cuter, more relatable. But then, as it gets very close to human, something strange happens. The line on the graph suddenly plunges. It drops into a deep valley of revulsion. Discomfort. Dread. This is the uncanny valley. It’s the place where things are almost human… but not quite. Mori used the example of a prosthetic hand.

He wrote, “Once we realize that the hand that looked real at first sight is actually artificial, we experience an eerie sensation.” It’s not the fake-ness that bothers us. A simple hook for a hand is not uncanny. It’s the FAILED attempt at perfection. The almost-real hand is far more disturbing than the obviously mechanical one. The question is… WHY. Why do our brains react with such primal fear to something that is almost human? There are a few theories. One is about survival. An evolutionary alarm bell. Think about what a nearly-human but flawed figure might represent to our ancient instincts.

It might look like a corpse. The waxy skin, the vacant stare, the unnatural stillness. Our brains are hardwired to feel revulsion toward the dead, to stay away, to avoid disease. Or, the subtle imperfections might signal sickness. A subtle facial asymmetry, a jerky movement, an odd skin tone. These are all biological red flags. The uncanny figure isn't a robot… it's a walking illness. Your discomfort is an ancient command: get away. Another theory is about mental conflict. Your brain is a categorization machine. It needs to put everything in a box.

Human. Animal. Object. Tool. When it encounters something that blurs these lines — a robot that looks like a person, a doll that seems to breathe — it can't file it away. It’s a walking paradox. It breaks the rules of your mental model of the world. This creates a kind of cognitive static, a deep and unsettling error message in your perception. The thing in front of you SHOULD NOT exist. And the discomfort you feel is your brain struggling to process a contradiction. Then there’s the idea of perceptual mismatch. This theory says the problem isn't the category, but the details.

The cues don’t add up. The skin looks real, but it has no pores. The eyes look human, but they don’t blink with the right rhythm. The voice has perfect grammar, but the emotional tone is flat. One part of the figure screams “human,” while another part whispers “fake.” Your brain detects this inconsistency, this lie. And it reacts with suspicion and disgust. It’s like hearing a beautiful symphony played with one instrument horribly out of tune. The perfection of the rest only makes the flaw more jarring. But it isn’t just about what you see.

The uncanny valley is something you can hear. Horror game designers know this. They use eerie, almost-human sounds to create fear. A footstep that’s just a little too regular. A breath that doesn’t quite hitch. A voice that is ALMOST a whisper. And this is where the valley gets complicated. A 2026 study looked at synthesized voices. You would expect that as a computer-generated voice gets more human-like, it would get more eerie. But the study found the opposite. The most human-like synthesized voices were rated as more likable, and LESS eerie.

The researchers found that the key factors were pleasantness and trustworthiness. A voice that sounds warm and consistent, even if artificial, can climb out of the valley on its own. The problem, again, is mismatch. If a perfectly human-sounding voice comes from a stiff, unmoving robot, the conflict is immense. The uncanny feeling isn’t just about realism. It’s about coherence. Everything has to be real in the same way, at the same time. Any inconsistency, any single false note, and the entire illusion shatters. And you are left staring into the valley.

This isn’t just about robots in a lab anymore. Or AI-generated food on a menu. The uncanny valley is slowly seeping into your daily life. It’s the video filter that smooths your skin just a little too much, erasing the pores and lines that make a face look alive. It's the AI-generated profile picture on a social network, with eyes that don't quite focus. It’s the automated customer service agent whose voice is perfectly modulated, perfectly polite, and utterly empty. For decades, the challenge was making the artificial look real.

But that is a problem being solved. The images, the voices… they are getting better. They are climbing out of the valley’s depths. And the eerie feeling of revulsion is being replaced by something else. A quiet uncertainty. The question is no longer, “Is this real?” The question is becoming, “Does it matter?” The dread no longer comes from the machine that looks like a person. The dread comes from the moment you look in a mirror, or at your own face on a screen, and for a fraction of a second, you see the lag. The artifact. The slight imperfection that makes you wonder if you are the original… or the copy.

Made with AI. AI may assist this episode’s research, and it writes and voices the episode. Neither the transcript nor this page is an editorial fact check. To report an error, email hi@lissin.com with the episode URL and timestamp. See our editorial standards.

Make your own show like this

Choose a topic you want to hear about, and Lissin makes the show around that choice.

Get Lissin for iPhone

More to listen to

Collections with this show: Scary stories to listen to

About Five Minutes of Dread

Unsettling AI-generated stories each week, with no gore, just dread that lingers after the audio ends.

All 33 episodes · More stories for adults shows