I don't understand articles that are completely about a piece of media, but don't show or link to any piece of said media. This article cannot be effective without a video.
> “The first slow-motion video of a butterfly I put through our bird-vision computer model was glowing with downwards motion even though the butterfly was moving up.
Yep, really would have loved to see that video. Why mention it and not link to it.
What I like about this is that the camouflage isn't really in the pattern itself, but in the interaction between the pattern, the wing deformation, and the predator's visual system.
Is this (in effect) "flood the zone" for certain targeted information theoretic chokepoints within the visual processing system? Similar to false cockpits on planes, you're adversarially and intentionally creating additional visual processing work?
A rabbit's high-contrast cottontail serves a similar purpose.
Predators focus on the tail but as the body turns, the tail very briefly appears to continue in the same direction or momentarily disappears. The predator is forced to constantly refocus, creating lots of opportunities for them to make a mistake.
> The team combined multiple lines of evidence. They used a high-speed camera to film real butterflies taking off at over 1,000 frames per second to capture how their patterns behave in flight, as seen through the eyes of a bird.
> To test how patterns affected real attacks, 100 volunteers acted as “predators”, trying to catch virtual butterflies on a touchscreen.
This sounds fascinating, and I'd love to see footage demonstrating the barber-pole effect. If that's not available, I'd alternatively love to see footage of volunteers trying to tap butterflies on an iPad.
This weekend while cleaning dry leaves in our backyard we discovered a large moth. We all mistook it fot another dry leaf until it started fluttering. Fabulous stuff.
Trait forms by chance, trait confers advantage, next generation has more individuals with trait. No intelligence needed to climb mount improbably, just massive parallel trial and error and selection.
I don't understand articles that are completely about a piece of media, but don't show or link to any piece of said media. This article cannot be effective without a video.
Even a 10-second slow-motion clip would probably explain the result better than half the article
> “The first slow-motion video of a butterfly I put through our bird-vision computer model was glowing with downwards motion even though the butterfly was moving up.
Yep, really would have loved to see that video. Why mention it and not link to it.
Agreed. Dr. Troscianko posted a video on YouTube. https://youtu.be/joosKZ3HygE?t=316
Also the paper is easy to read and has illustrations. https://www.nature.com/articles/s41586-026-11062-w
What I like about this is that the camouflage isn't really in the pattern itself, but in the interaction between the pattern, the wing deformation, and the predator's visual system.
Sounds like the ol' razzle dazzle
https://en.wikipedia.org/wiki/Dazzle_camouflage
Is this (in effect) "flood the zone" for certain targeted information theoretic chokepoints within the visual processing system? Similar to false cockpits on planes, you're adversarially and intentionally creating additional visual processing work?
PS, still a great musical...
A rabbit's high-contrast cottontail serves a similar purpose.
Predators focus on the tail but as the body turns, the tail very briefly appears to continue in the same direction or momentarily disappears. The predator is forced to constantly refocus, creating lots of opportunities for them to make a mistake.
> The team combined multiple lines of evidence. They used a high-speed camera to film real butterflies taking off at over 1,000 frames per second to capture how their patterns behave in flight, as seen through the eyes of a bird.
> To test how patterns affected real attacks, 100 volunteers acted as “predators”, trying to catch virtual butterflies on a touchscreen.
This sounds fascinating, and I'd love to see footage demonstrating the barber-pole effect. If that's not available, I'd alternatively love to see footage of volunteers trying to tap butterflies on an iPad.
The description is interesting enough that it almost feels incomplete without the video
Well, what the kids don’t learn in biology class, they can still learn on HN.
This weekend while cleaning dry leaves in our backyard we discovered a large moth. We all mistook it fot another dry leaf until it started fluttering. Fabulous stuff.
Quite the assertion with zero evidence nor competing theories provided or quoted
The paper provides four classes of evidence for one theory lacking any.
There's no paper cited
> “To test how patterns affected real attacks, 100 volunteers acted as “predators”, trying to catch virtual butterflies on a touchscreen.”
I’d love to play this game
This is Micheal Levin-level mind blowing. His theory can be shortened to "there's intelligence in really small discrete parts of Biological entities".
How is that different from just saying "evolutionary adaptive changes can seem clever".
Trait forms by chance, trait confers advantage, next generation has more individuals with trait. No intelligence needed to climb mount improbably, just massive parallel trial and error and selection.