How algorithmic vision is changing the way we see the world, or, what does happiness really look like?
This video shows us something that we cannot see without technology. It’s actually not a recording, but an animation based on scientific studies of the brain, first published at Art of the Cell, a medical animation company, but since reposted many places, often with the following text:
This is what happiness really looks like: Molecules of the protein myosin drag a ball of endorphins along an active filament into the inner part of the brain’s parietal cortex, which produces feelings of happiness.
Part of the reason this image appeals to us is the anthropomorphic strut of that myosin — and the words that go with the image: “This is what happiness really looks like.”
What do we mean by that? What happiness really looks like? Is there a single, true way of “seeing” happiness?
Consider the first photographs of an unborn child, popularised in the 1960s by Lennart Nilsson’s still popular book, A Child is Born. The photo below is a slightly updated version of one published in Life Magazine in April 1965 (This issue was digitized by Google Books so you can look at the cover here, and scroll through to page 54 for the whole story). To me, the child looks like a traveler in space, the specks like stars against the black of space.
When I first looked at the image, I thought of photos of the Earth taken from space. I assumed that the image of the foetus was referencing the image of Earth. But the iconic Blue Marble photo of Earth wasn’t taken until 1972.
Do these photographs show us what an unborn child, or the Earth, really look like?
In 1878 Eadweard Muybridge took a series of photographs of a horse, galloping, and for the first time clarified the question of whether or not a horse actually has all its feet off the ground at any time during galloping.
As you can see from the second and third images, it does. But not in the way most painters had seen galloping horses before. Théodore Géricault’s painting shows how people saw a galloping horse before photography proved our eyes wrong: with legs out, flying over the ground.
OK, so Muybridge showed us what a galloping horse really looks like, correcting our misconceptions, right? Maybe not. The sculptor Auguste Rodin disagreed.
‘No. (..) It is art that tells the truth and photography that lies. For in reality time does not stand still, and if the artist manages to give the impression that a gesture is being executed over several seconds, their work is certainly much less conventional than the scientific image in which time is abruptly suspended. …’ (quoted by Paul Virilio in The Vision Machine, 1994, pp. 1–2)
Looking back at Géricault’s galloping horses after more than a century of fast photography, it is hard to see them as more truthful than Muybridge’s photographs. But Rodin’s point is broader than that. Art speaks a different truth than technology. And perhaps from a human perspective, the truth of art, or of our flawed human perception, is more important.
Muybridge’s photographs offer us one kind of truth. But other kinds of photography and other kinds of technology offer us other truths.
Technology always distorts what we see. When Parmigianino painted this self-portrait in 1524, he used a convex mirror, and that technological choice caused a distortion in the image, making his hand appear larger than it “really” was.
The distortion of the foetus in Lennart Nilsson’s photograph is caused by the opposite effect due to the fisheye lens that was used to photograph inside the small dark space of the uterus.
Leafing (well, scrolling digitally) through the issue of Life where Lennart Nilsson’s photos were published, I notice the spherical shape is repeated. First, on the page immediately after the section on the in-utero photographs, there is an ad for a car, where a fisheye view of what can be seen from the back seat of the car is shown, bright blue on a black background.
Then on page 83, which is a full page ad for Hughes and Comsat, showing a new satellite that will enable live trans-Atlantic telecasts and phonecalls. A globe is shown in the ad. Not a photograph of the Earth itself, because no photograph of the whole Earth yet existed. There seems to be a desire, though, for photographs of spheres floating in space.
The spherical image in the Ford ad was clearly taken with a fisheye lens. Fisheye lenses weren’t mass-produced for photography until the early 1960s — so just before this issue of Life was published. Nilsson used fisheye and wide-angle lenses both for his photography inside the body and for other photographs. And he even presented images actually taken outside of the body — like that of a foetus taken from the womb of a woman who was killed in a traffic accident — as though they were taken with wide-angle lenses. (Here’s a recent academic article about Nilsson’s Fish-Eyes.)
These images would not have existed without the fisheye lens. Perhaps we also wouldn’t have wanted to see these spherical images without that particular technology?
We see the world around us through technology. We use machines to understand the world. And we trust our technology. Yes, this is what happiness really looks like. This is what a foetus, or Earth, or a horse galloping really looks like.
But what else could it look like, if we saw it through different technology? And what happens as we increasingly, today, use machines and cameras and algorithms to sense our surroundings?
What happens to our understanding of the world when we no longer primarily rely on human perception but use machines and algorithms to sense our surroundings?
Will machines change the way we see?
I’m teaching a graduate seminar on these questions this semester. If you’re interested, check it out. If you are a graduate student at the University of Bergen, you can email studieveileder@lle.uib.no to sign up for the course.
If you enjoyed this story, please click on the heart so that others will see it on Medium!
