I believe that "learn" is a bit too strong word here. The fungi is essentially a UV light sensor. The researchers made a robot that moves in a certain way based on the biological signal.
So the mushroom is more like a passive sensor then an active pilot.
It can be firing in arbitrary banner and we’ll still call it movement. Like the LLM does random and lossy decompression and we give it context and meaning…
I’m skeptical that the mushroom is in any way “learning to crawl”. It looks more like the mushroom naturally produces signals in response to light, and the robot triggers a walk cycle when it sees that signal.
Off topic but my dad took me around the woods this weekend to show me mushrooms and he almost couldn’t contain his excitement. And this is a person who usually doesn’t like stuff. And the whole time I was like “yep, that’s a mushroom”. There’s clearly something fascinating about the hobby that I don’t get (yet). Curious to hear your take.
Your dad sounds like a good sort. I have no idea how to explain the mushroom fascination that some of us have to those that lack it and have mostly learned to just not talk about mushrooms with people who don't have the fascination.
If you take a small plot of wild land, say 50'x50' and visit it everyday watching all the various mushrooms which will grow there (there will be far more then most people realize or expect) and try and find the sense in where and when they fruit it becomes difficult to write it all off as simple biological responses to environment/following the food/etc, and even if you do look at it that way and start adding up the possible causes and effects you can end up with such a long list that it becomes difficult to not see it as some sort of at least instinctual level intelligence and that the growth of mycelium often has more in common with crawling than mindless growth.
For example, many mushrooms are very good at fruiting just out of sight from trails, walk 20' off the trail, turn around and suddenly you start seeing mushrooms. Instinct is to say that all the mushrooms which grow within sight of the trail get picked by curious people or kicked by children but if you start looking for remains and stumps and mushrumps and those hard to spot just starting to fruit immature buttons, you find surprisingly few. So you think environmental, the trail alters windflow and runoff, animal movement, etc, but than you notice that this is true of even those small trails created by a fox or children which only affect a bit of low growing undergrowth so only has a tiny effect on a very localized area. On and on it goes until you run out of explanations.
I am mostly convinced there is some level of intelligence here and we just can not see it because it is so very different from what we understand as intelligence. But, I may just spend too much time with mushrooms, during the season I always seem to have at least a dozen various mushrooms which I will visit everyday to watch them grow and rot away.
Looks like they respond to their environment in complex ways. What if mushrooms avoid trails because they feel the vibrations? Maybe they learned that foot traffic = danger. So it grows out of reach.
You should explore this theory and run some field studies.
fungi has memory and can decide to not grow on a previous hostile enviroment/direction
edit: i'm skeptical about the fantastic type of journalism title. the paper points on using fungi electrical reactions to light to drive a robot, not the the otherwise, even less the fungi understanding a "robot" and using it. despite them showing spatial perception on studies about their capabilities
What the article says is that fungi produces electrical spikes and those spikes may encode information and function like a primitive language to coordinate activity across the network.
But what if the environment was never hostile and they still avoid it? Some research hints that fungi can react to vibration, like stimuli, but there's no solid proof that mushrooms avoid trails because of it. Still, it feels worth investigating.
Slime molds are super interesting and are able to find optimal paths for resource management. Slime molds have been used to model and improve traffic flow for humans.
> Slime molds are classified within the Kingdom Protista, and are more closely related to amoebas and certain seaweeds than to fungi, plants, or animals.
Still, a child telling their parent to walk isn't actually walking.
It could easily not even know what walking means, just know that there's food when it tells mom to go to the kitchen.
I used to ask my grandmother for "plain water" because she always brought me milk. I was very disappointed when I asked someone else for plain water, because I had no idea why grandma's was better.
Because the brain has to continually monitor hundreds of inputs like gravity, pressure, balance, visuals, muscle feedback, etc and integrate them with hundreds of outputs to muscles and coordinate the whole thing which keeps changing, that's walking. All this mushroom does is emit a signal that it was already emitting (ie. It's learned nothing) and the robotocists build a walking system around that which handles all the complexity of walking.
Let me rephrase: what's the difference between a child telling their parent to walk and the conscious part of your brain telling the unconscious part to walk? (BTW, I agree that the mushroom isn't learning anything. My point is just that a child communicating with their parents is not a good analogy.)
The "conscious part of the brain" gets credit for the walking ability, because it taught "the unconscious part" how to walk, by providing repeated and detailed direction in progressively higher abstractions until full bipedal locomotion was a readily-accessible skill.
Most people have rudimentary language ability, including the ability to express preferences, as they learn to walk. The average one-year-old child is plenty aware of the world.
>Because the brain has to continually monitor hundreds of inputs like gravity
Human brains cannot monitor gravity. We're actually very bad at it.
Also you're failing to understand the scaling involved here, so to speak. I'm actually really not interested in trying to explain this if you can't be bothered with spending a few minutes on how nervous systems work.
Has science gone too far, or not far enough? Surely there are business opportunities for reanimating the dead bodies of animals! Where is the mouse jiggler made from an actual dead mouse being controlled by a Raspberry Pi? Or the carcass of a dead dog controlled by an air bladder that will lurch from behind your shrubs and snap at a prowler using zigbee and connected to your security system?
Not sure about this particular experiment, but there is certainly interesting potential in integrating biological organisms (or parts thereof) with larger robotic and mechanical systems.
Recently I saw a video of a turtle which was given a skateboard. It quickly learned how to zip around the house, chasing the cat, etc. It was a simple demostration of how technology, even as primitive as the wheel, can augment the abilities of an organism - especially a living being with sensors (eyes) and neural network (brain).
It also reminds me of the goldfish in a bowl, attached to a small motorized vehicle, which was given the ability to navigate it by swimming in different directions. It soon learned to use this system as an extension of its body, exploring the house, bumping into things like a Roomba with a live brain.
Suppose it's in the same field of exploration as those super-soldiers with Gundam-style body suits and computerized helmets projecting a live data feed to their retinas, maybe eventually embedding neural connectors directly in the head.
Regarding the turtle and the goldfish, how can we really say these animals learned how to operate these things? I’m not sure I’d be able to tell the difference between a goldfish just swimming around the tank like normal versus one swimming around the tank with intention.
Oh that's a more proper study than the amateur experiment I saw.
> For this purpose, we trained goldfish to use a Fish Operated Vehicle (FOV), a wheeled terrestrial platform that reacts to the fish’s movement characteristics, location and orientation in its water tank to change the vehicle’s; i.e., the water tank’s, position in the arena.
> The fish were tasked to “drive” the FOV towards a visual target in the terrestrial environment, which was observable through the walls of the tank, and indeed were able to operate the vehicle, explore the new environment, and reach the target regardless of the starting point, all while avoiding dead-ends and correcting location inaccuracies.
I would guess you could achieve similar results with a rat's or cat's brain, but I wonder at which point ethical dilemmas start creeping in. When the fungi learns to ask for food, perhaps?
This seems to be one of the researches from Organic Robotics Lab at Cornell Univ.
https://orl.mae.cornell.edu/
https://news.cornell.edu/stories/2024/08/biohybrid-robots-co...
I believe that "learn" is a bit too strong word here. The fungi is essentially a UV light sensor. The researchers made a robot that moves in a certain way based on the biological signal.
So the mushroom is more like a passive sensor then an active pilot.
It can be firing in arbitrary banner and we’ll still call it movement. Like the LLM does random and lossy decompression and we give it context and meaning…
I’m skeptical that the mushroom is in any way “learning to crawl”. It looks more like the mushroom naturally produces signals in response to light, and the robot triggers a walk cycle when it sees that signal.
As a fungophile who spends way too much time crawling around in the woods looking for mushrooms, I think fungus learned to crawl before we did.
Off topic but my dad took me around the woods this weekend to show me mushrooms and he almost couldn’t contain his excitement. And this is a person who usually doesn’t like stuff. And the whole time I was like “yep, that’s a mushroom”. There’s clearly something fascinating about the hobby that I don’t get (yet). Curious to hear your take.
Your dad sounds like a good sort. I have no idea how to explain the mushroom fascination that some of us have to those that lack it and have mostly learned to just not talk about mushrooms with people who don't have the fascination.
I find that claim interesting, especially given your background in studying fungus. Could you expand on it a bit?
If you take a small plot of wild land, say 50'x50' and visit it everyday watching all the various mushrooms which will grow there (there will be far more then most people realize or expect) and try and find the sense in where and when they fruit it becomes difficult to write it all off as simple biological responses to environment/following the food/etc, and even if you do look at it that way and start adding up the possible causes and effects you can end up with such a long list that it becomes difficult to not see it as some sort of at least instinctual level intelligence and that the growth of mycelium often has more in common with crawling than mindless growth.
For example, many mushrooms are very good at fruiting just out of sight from trails, walk 20' off the trail, turn around and suddenly you start seeing mushrooms. Instinct is to say that all the mushrooms which grow within sight of the trail get picked by curious people or kicked by children but if you start looking for remains and stumps and mushrumps and those hard to spot just starting to fruit immature buttons, you find surprisingly few. So you think environmental, the trail alters windflow and runoff, animal movement, etc, but than you notice that this is true of even those small trails created by a fox or children which only affect a bit of low growing undergrowth so only has a tiny effect on a very localized area. On and on it goes until you run out of explanations.
I am mostly convinced there is some level of intelligence here and we just can not see it because it is so very different from what we understand as intelligence. But, I may just spend too much time with mushrooms, during the season I always seem to have at least a dozen various mushrooms which I will visit everyday to watch them grow and rot away.
Looks like they respond to their environment in complex ways. What if mushrooms avoid trails because they feel the vibrations? Maybe they learned that foot traffic = danger. So it grows out of reach.
You should explore this theory and run some field studies.
great read: https://www.sciencedirect.com/science/article/pii/S187861462...
fungi has memory and can decide to not grow on a previous hostile enviroment/direction
edit: i'm skeptical about the fantastic type of journalism title. the paper points on using fungi electrical reactions to light to drive a robot, not the the otherwise, even less the fungi understanding a "robot" and using it. despite them showing spatial perception on studies about their capabilities
What the article says is that fungi produces electrical spikes and those spikes may encode information and function like a primitive language to coordinate activity across the network.
But what if the environment was never hostile and they still avoid it? Some research hints that fungi can react to vibration, like stimuli, but there's no solid proof that mushrooms avoid trails because of it. Still, it feels worth investigating.
Slime molds are super interesting and are able to find optimal paths for resource management. Slime molds have been used to model and improve traffic flow for humans.
https://phys.org/news/2022-01-virtual-slime-mold-subway-netw...
I would call this action "crawling" and I am sure predates mammals. Or dinosaurs. Or plants.
Slime molds aren’t fungus, though…
> Slime molds are classified within the Kingdom Protista, and are more closely related to amoebas and certain seaweeds than to fungi, plants, or animals.
Not a true fungi; today I learned. Thanks!
[flagged]
Still, a child telling their parent to walk isn't actually walking.
It could easily not even know what walking means, just know that there's food when it tells mom to go to the kitchen.
I used to ask my grandmother for "plain water" because she always brought me milk. I was very disappointed when I asked someone else for plain water, because I had no idea why grandma's was better.
> a child telling their parent to walk isn't actually walking.
How is that any different from a brain telling its body to walk?
Because the brain has to continually monitor hundreds of inputs like gravity, pressure, balance, visuals, muscle feedback, etc and integrate them with hundreds of outputs to muscles and coordinate the whole thing which keeps changing, that's walking. All this mushroom does is emit a signal that it was already emitting (ie. It's learned nothing) and the robotocists build a walking system around that which handles all the complexity of walking.
Let me rephrase: what's the difference between a child telling their parent to walk and the conscious part of your brain telling the unconscious part to walk? (BTW, I agree that the mushroom isn't learning anything. My point is just that a child communicating with their parents is not a good analogy.)
The "conscious part of the brain" gets credit for the walking ability, because it taught "the unconscious part" how to walk, by providing repeated and detailed direction in progressively higher abstractions until full bipedal locomotion was a readily-accessible skill.
I'm pretty sure that's not true. I think most people learn to walk long before they become aware of being able to walk.
Most people have rudimentary language ability, including the ability to express preferences, as they learn to walk. The average one-year-old child is plenty aware of the world.
>Because the brain has to continually monitor hundreds of inputs like gravity
Human brains cannot monitor gravity. We're actually very bad at it.
Also you're failing to understand the scaling involved here, so to speak. I'm actually really not interested in trying to explain this if you can't be bothered with spending a few minutes on how nervous systems work.
They're obviously referring to the vestibular system. If that system is bad, that only makes the brain's job harder, not easier.
Let's say you start wearing a mushroom as an extra sensory input, and you started doing a dab every time you sensed something from the mushroom
Would you describe that situation as the mushroom learning to dab?
Hey! What of us with fungi in our intestines.
They're supposedly stimulating the gut, causing cravings.
Probably the only case of mushrooms wanting pizza.
This combo reminds me of a character in a recent anime called Scavengers Reign.
https://www.wikiwand.com/en/articles/Scavengers_Reign
This limited series blew my mind. Total master piece.
In favor of integrating fungus with robotics(i think).
erm acktually not an anime
Another "fun" science trick here[0] demo here[1]
[0]https://en.wikipedia.org/wiki/Necrobotics
[1]https://www.youtube.com/shorts/_7LUszWRqco
> Besides the necrobotic spider gripper, there are no other robotic concepts under the necrobotics subfield.
Does dead trout swimming upstream not count? doi:10.1017/S0022112005007925 (open access link: https://www.liaolab.com/wp-content/uploads/2020/10/2006Beal_...)
Has science gone too far, or not far enough? Surely there are business opportunities for reanimating the dead bodies of animals! Where is the mouse jiggler made from an actual dead mouse being controlled by a Raspberry Pi? Or the carcass of a dead dog controlled by an air bladder that will lurch from behind your shrubs and snap at a prowler using zigbee and connected to your security system?
The Mechanicum will follow your career with much interest. ;)
I'm getting a bit of a skroderider[0] feeling.
[0] https://en.wikipedia.org/wiki/A_Fire_Upon_the_Deep#Skroders/...
Came here for this.
Not sure about this particular experiment, but there is certainly interesting potential in integrating biological organisms (or parts thereof) with larger robotic and mechanical systems.
Recently I saw a video of a turtle which was given a skateboard. It quickly learned how to zip around the house, chasing the cat, etc. It was a simple demostration of how technology, even as primitive as the wheel, can augment the abilities of an organism - especially a living being with sensors (eyes) and neural network (brain).
It also reminds me of the goldfish in a bowl, attached to a small motorized vehicle, which was given the ability to navigate it by swimming in different directions. It soon learned to use this system as an extension of its body, exploring the house, bumping into things like a Roomba with a live brain.
Suppose it's in the same field of exploration as those super-soldiers with Gundam-style body suits and computerized helmets projecting a live data feed to their retinas, maybe eventually embedding neural connectors directly in the head.
Regarding the turtle and the goldfish, how can we really say these animals learned how to operate these things? I’m not sure I’d be able to tell the difference between a goldfish just swimming around the tank like normal versus one swimming around the tank with intention.
I assume a fish that didn’t understand what was going on would just run into walls? Or at most just move towards lights?
The top-level comment indeed says that the fish was bumping into things like a Roomba, so I’m also skeptical....
judge for yourself https://www.youtube.com/watch?v=2nhp7U0rwdM
Oh that's a more proper study than the amateur experiment I saw.
> For this purpose, we trained goldfish to use a Fish Operated Vehicle (FOV), a wheeled terrestrial platform that reacts to the fish’s movement characteristics, location and orientation in its water tank to change the vehicle’s; i.e., the water tank’s, position in the arena.
> The fish were tasked to “drive” the FOV towards a visual target in the terrestrial environment, which was observable through the walls of the tank, and indeed were able to operate the vehicle, explore the new environment, and reach the target regardless of the starting point, all while avoiding dead-ends and correcting location inaccuracies.
From fish out of water to new insights on navigation mechanisms in animals - https://www.sciencedirect.com/science/article/abs/pii/S01664...
You’ll need to watch the video of the turtle. It was actively chasing a cat around, so it definitely knew what it was doing.
Tiny Turtle Follows Cat On a Skateboard - https://www.youtube.com/watch?v=bVbtAYPSapw
lets start working our beliefs up from easier places :
https://www.youtube.com/watch?v=RZ_0ImDYrPY
https://www.youtube.com/shorts/7H2IDc-5QdQ
Also see Project Pigeon: https://en.wikipedia.org/wiki/Project_Pigeon
Good ol' B.F. Skinner. Apparently the idea was pretty solid but he had trouble getting people to take it seriously.
What could possibly go wrong?
Or right? Reminds me of the Skroderider species of sentient seaweed from Vernor Vinge's "A Fire Upon the Deep".
I came here to say this!
I, for one, welcome our new fungal overlords!
Not an expert on fungi, or the kinddom in general, but slime molds have been found to do some amazing things. https://www.nature.com/articles/nature.2012.11811
Reminds me of:
https://www.syfy.com/syfy-wire/fish-control-vehicles-and-nav...
Rats too!! :)
https://www.youtube.com/watch?v=mYHMc3-f3v8
I would guess you could achieve similar results with a rat's or cat's brain, but I wonder at which point ethical dilemmas start creeping in. When the fungi learns to ask for food, perhaps?
Here I thought ST Discovery had jumped the shark with its whole mycelium navigation plot device.
Before you know it will have a mushroom brain.
See also: https://news.cornell.edu/stories/2024/08/biohybrid-robots-co...
“Autobiography of a human, or how mushrooms learned to build computers after being given primate bodies”
And I, for one, welcome our new mushroom overlords.
slashdot is that way, friend -> https://slashdot.org
[flagged]
Anybody else find this creepy?
Just don't give them six wheels! Hexapodia is the key insight!
Very creepy however it is living proof that Trump is real and not CGI.
A real liar.