> To arrive at the heart of their new findings the researchers created an immersive virtual reality environment to test the fly’s behavior via visual stimulation and coupled the displayed imagery with an infra-red laser as an averse heat stimulus. The near 360-degree panoramic arena allowed Drosophila to flap their wings freely while remaining tethered, and with the virtual reality constantly updating based on their wing movement (analyzed in real-time using high-speed machine-vision cameras) it gave the flies the illusion of flying freely in the world. This gave researchers the ability to train and test flies for conditioning tasks by allowing the insect to orient away from an image associated with the negative heat stimulus and towards a second image not associated with heat.
This is seriously beyond what I would hace thought feasible. Literally a virtual reality rig for flies that tracks their wing movement to guide virtual movement? And yet we barely have a similar set up for human locomotion? What gjves?
Compared with humans, flies don't have many processing neurons between their effectors (wings) and their eyes. It's why they can take off so quickly when you try to swat them. Their simple but effective processing allows them to be more easily manipulated then humans. Also, the flies aren't going to say "Meh this VR has fallen into the uncanny valley" or similar.
> It's why they can take off so quickly when you try to swat them.
It's also why defeating them is simple. They fly up and away from whatever is coming at them. But if you come at them on at least two sides, they simply fly up. If you sweep your hands toward them and then move up as you bring your hands together, you'll nail the fly almost every time.
That's a valid suggestion, but unfortunately wouldn't hide the substance. I think I have separate revulsion to insect contact and insect weight/mass. I'd prefer certain bugs to not touch me (e.g. ladybugs are fine), but I _really_ would prefer them to feel weightless. Something about a bug's weight being perceptible is intensely revolting to me, as if they're meant to be weightless wisps
The new part here is not the fly glued into a VR rig. Researchers have been doing the VR fly thing for well over a decade! Dickenson was doing this at Caltech at least in 2008.
> more sophisticated cognitive abilities than previously known
No. More sophisticated cognitive abilities than previously modelled
using atemporal neural nets.
Anyone who has spent an afternoon being outwitted by a fly in their
study knows that even a common housefly can be cunning beyond belief.
It's not just that their reaction times out-pace humans. They seem to
remember, anticipate, plan and even deceive.
Those faculties seem inexplicable by a bottom-up network of 100,000
neurons using naive connectionist models we currently understand in
artificial networks. That's because time is missing from the
models. Biological networks are spiking, and (last I researched it)
can operate at up to 1kHz, and using integration, differentiation and
non-linear feedback can work like analogue computers (Hodgkin and
Izhikevich models). Instead of sending simple activation thresholds
around the network that's like sending network packets that can
perform remote procedure calls.
Both in terms of the jaw-dropping technical setup (full VR rig AND real-time brain imaging for untethered flies), and the results (proof of otherwise unknowable cognitive abilities) with interesting philosophical implications about the differentiation we humans perform when categorizing animals (for example, is killing a fly morally different than killing a horse/cow?)
I have witnessed countless of wasp die because they don't know how to go through an open window it's such an ordeal to catch the damn things and get them out that I stopped bothering.
A drop of sugar syrup on a plate near where they're landing will usually get and hold their attention - just a drop, too much can become a glue trap - and slow movements while you take the plate outside will typically not disturb them. Then you just need to bring the plate back in later, once they're done with it and have gone about their occasions.
Solitary wasps tend to be a little twitchier than social ones, but I've had this method work as well with both sorts. On the other hand, I understand wasps well enough not to be frightened of them and they seem to like me pretty well too, so I suppose I can't vouch for good results in all circumstances.
Can we appreciate the fact that humanity has come to a point where dystopian shit like this has become the norm? Oh but of course we can! We do!
But should we really? I mean, come on. This is nuts. There's already farmers putting VR goggles onto cows. Because they give better milk. It's progress alright, but it's nuts.
Completely bonkers, if you ask my british personality fragment. It really isn't that far from putting VR onto everything and everyone simply because people have stopped appreciating actual reality, with actual nature.
This research is being done to understand cognition. To connect to the biochemistry and biophysics. To improve our models and begin making bigger leaps.
Even simple systems like drosophila are wildly complicated.
A hundred years of this type of investigation and maybe one day we'll be able to record memories. Transfer consciousness. End death.
This isn't just "fly VR". This is the curiosity of being human. And yet another slow, methodical step towards solving it all.
Willing acceptance of our eventual fate of turning into ash while being born in a generation that has all the capacity for increasing lifespan is what's dystopian.
Thinking intelligent life should continue in its current state of being mostly meat flesh bags until the sun boils our ancestors is dystopian.
Status quo, not trying to be special, not trying to reach for the stars with our limited time. That's not just dystopian, it's Kafkaesque. We may as well be in cockroach bodies waiting for the universe to crush us, because thats what not looking up from the dirt gets us.
Curiosity, wonder, and unbounded dreams are among the best qualities of human life. They transcend our weak and dying bodies and our mortal lifetimes.
There are worse things than dying. If I trusted humans not to immediately do all of them upon achieving the capability to do them, my opinion here might more resemble yours.
I don't see where learning that fruit flies are a lot smarter than most people thought makes us appreciate nature less.
If you want a dystopian take, why not this one instead? Climate change is bad enough already if you just assume the billions of nonhuman animals killed, displaced, and degraded by it are meat robots. How much worse to know instead that they are, consistent with their capabilities, no less learning, thinking, and feeling beings than we are?
39 comments
[ 3.2 ms ] story [ 92.6 ms ] threadThat's wild!
It's also why defeating them is simple. They fly up and away from whatever is coming at them. But if you come at them on at least two sides, they simply fly up. If you sweep your hands toward them and then move up as you bring your hands together, you'll nail the fly almost every time.
https://en.wikipedia.org/wiki/Fly-killing_device#Flyswatter
The fly gun looks like fun too.
https://authors.library.caltech.edu/90960/
> more sophisticated cognitive abilities than previously known
No. More sophisticated cognitive abilities than previously modelled using atemporal neural nets.
Anyone who has spent an afternoon being outwitted by a fly in their study knows that even a common housefly can be cunning beyond belief. It's not just that their reaction times out-pace humans. They seem to remember, anticipate, plan and even deceive.
Those faculties seem inexplicable by a bottom-up network of 100,000 neurons using naive connectionist models we currently understand in artificial networks. That's because time is missing from the models. Biological networks are spiking, and (last I researched it) can operate at up to 1kHz, and using integration, differentiation and non-linear feedback can work like analogue computers (Hodgkin and Izhikevich models). Instead of sending simple activation thresholds around the network that's like sending network packets that can perform remote procedure calls.
And yes, they love bananas.
Both in terms of the jaw-dropping technical setup (full VR rig AND real-time brain imaging for untethered flies), and the results (proof of otherwise unknowable cognitive abilities) with interesting philosophical implications about the differentiation we humans perform when categorizing animals (for example, is killing a fly morally different than killing a horse/cow?)
Its more about glass being transparent and all than intelligence I would imagine.
We never ever keep misreading situations and making the same mistakes over and over.
Funny -- but also first time I've heard of it!
Solitary wasps tend to be a little twitchier than social ones, but I've had this method work as well with both sorts. On the other hand, I understand wasps well enough not to be frightened of them and they seem to like me pretty well too, so I suppose I can't vouch for good results in all circumstances.
We put a fly into virtual reality. A fly.
Can we appreciate the fact that humanity has come to a point where dystopian shit like this has become the norm? Oh but of course we can! We do!
But should we really? I mean, come on. This is nuts. There's already farmers putting VR goggles onto cows. Because they give better milk. It's progress alright, but it's nuts.
Completely bonkers, if you ask my british personality fragment. It really isn't that far from putting VR onto everything and everyone simply because people have stopped appreciating actual reality, with actual nature.
A fly!
In virtual reality!
It's unironically amazing!
We're fucked. Seriously. lol
This research is being done to understand cognition. To connect to the biochemistry and biophysics. To improve our models and begin making bigger leaps.
Even simple systems like drosophila are wildly complicated.
A hundred years of this type of investigation and maybe one day we'll be able to record memories. Transfer consciousness. End death.
This isn't just "fly VR". This is the curiosity of being human. And yet another slow, methodical step towards solving it all.
Thinking intelligent life should continue in its current state of being mostly meat flesh bags until the sun boils our ancestors is dystopian.
Status quo, not trying to be special, not trying to reach for the stars with our limited time. That's not just dystopian, it's Kafkaesque. We may as well be in cockroach bodies waiting for the universe to crush us, because thats what not looking up from the dirt gets us.
Curiosity, wonder, and unbounded dreams are among the best qualities of human life. They transcend our weak and dying bodies and our mortal lifetimes.
If you want a dystopian take, why not this one instead? Climate change is bad enough already if you just assume the billions of nonhuman animals killed, displaced, and degraded by it are meat robots. How much worse to know instead that they are, consistent with their capabilities, no less learning, thinking, and feeling beings than we are?