Uncanny Valley (2015)

The Default Valley: Why Photorealism is a Neurological Minefield and Why We Want It Anyway

Two Ways to Break a Brain

There are two distinct ways immersive technology destabilises the mind and they are not the same thing.

The first is overwhelming and abstract. Lawnmower Man. Early Tilt Brush sessions that turned psychedelic. The original 90s VR wave — Virtuality machines, VRML — that tried too hard to be real and failed, but was disorienting in a different way regardless. Your brain knows it isn't real. It just doesn't know what to do with the signal. The limbic system floods anyway.

The second is about photorealism. About almost-right. The video above is about this kind — and it's the one that has been following me.

The Breach: When "Almost" Becomes "Total"

The Oculus wave of 2015 started blocky too — stylised worlds, abstract spaces. But the direction of travel was clear. When a simulation becomes sufficiently high-fidelity, it stops being a symbol and starts being a signal. Your brain is a prediction engine. Your visual cortex doesn't wait — it sends a message directly to your nervous system: This is real.

But the rest of your body — your inner ear, your skin, your sense of gravity — knows it isn't. This is Baudrillard's argument made visceral: the copy doesn't just represent the original, it replaces it, until there is no original left to return to. For some people that gap is a minor glitch. For others it's a place the brain cannot settle — and they don't always know they're in it.

3D Gaussian Splatting. Neural rendering. AI-generated environments. The photorealism that was always coming is nearly here. The valley is about to get very crowded.

The OASIS Was Not a Warning to My Students

Here is the tension I keep coming back to. Post-COVID, when I asked students what they actually wanted from education, a significant number described something close to the OASIS — a fully immersive shared space, persistent, social, embodied. Not escapism. Presence. They had spent years in flattened, disembodied learning environments and they were hungry for somewhere that felt real.

The same technology I am treating as a neurological minefield was the thing they were reaching for as a lifeline. Immersion as connection. The simulation not as a threat but as the closest available version of being somewhere together.

So it is not simply dangerous. The question is more uncomfortable: for whom, under what conditions, and do we understand the difference well enough to design accordingly?

The Unanswered Question

In 2015 I posted this film on the last day of the year. Something to provoke thought rather than answer it. A dark thread to carry into a new one. Ten years later the simulation isn't theoretical — it's arriving, and AI has dragged the valley out of the headset and into the everyday. Generated faces, generated voices, synthetic memories of events that never happened.

We are conducting a massive, uncontrolled experiment on human susceptibility — sometimes for harm, sometimes for genuine benefit — and we haven't built the frameworks to tell which is which, or the lifeboats for those who get lost in the process.

Who is already struggling to recognise reality? Who is being held together by a simulation that feels more real than their waking life? What are we doing to both of them?

Technology evolves and society evolves with it — that's always been true. We have work to do to ensure we make good decisions about how we get there.