Eyes Do Not Work the Way They Appear To
The article discusses a visual phenomenon where the way our eyes actually process light (from a mix of bright and dim areas) differs from how we perceive an image, using a folded-card optical illusion as an example. It highlights the discrepancy between physical eye function and visual perception, suggesting our eyes behave more like high-end cameras than simple light-measuring devices.
Background
- This is an essay by Ishan, a writer and technologist who explores perception, cognition, and technology on his blog.
- The piece argues that human vision is not a passive, camera-like recording of external reality, but an active, predictive process where the brain constructs what we "see" using prior knowledge and expectations.
- It draws on modern neuroscience, especially "predictive processing" theory — the idea that the brain continuously generates models of the world and updates them based on sensory prediction errors.
- The essay challenges the common-sense intuition ("eyes are windows to the world") and uses examples like optical illusions, blind spots, saccades (rapid eye movements), and change blindness to show how much of vision is hallucination corrected by input.
- Key reference: the work of neuroscientist Anil Seth, who argues that perception is a "controlled hallucination."
- This matters because it undermines naive realism — the idea that we perceive reality as it is — and has implications for AI, virtual reality, epistemology, and how we understand consciousness itself.