> Over three months, 66 chimps tried and failed to solve the puzzle. Then the Dutch-led team of researchers trained two demonstrator chimpanzees to show the others how it was done. After two months, 14 "naive" chimps had mastered it.
So 14 out of 64 chimps learned a relatively simple trick after two months of "cumulative culture" exposure. Just over a 20% success rate?
> Thornton said the research again showed how "people habitually overestimate their abilities relative to those of other animals".
People learn how to tile floors, install dishwashers, change their cars oil filter, build websites etc. all from mediocre youtube videos. I don't think I'm overestimating humans "abilities relative to those of other animals" ...
Also, the human scientists figured out ape intelligence to the point where they can design puzzles that are too hard for the apes to solve, and they figured out how to use that to further study the apes. Underestimate the apes? These scientists seem to be absolutely dunking on the poor little guys. They should go pick on orcas or some other actually clever species.
The bee results are suss. Bees use pheromones to mark things so there may be no socially learned behavior beyond seeking food at a marked source which is built in. There's also the small sample size which the article mentions. It is still cool that they learned the blue/red trick in the first place either way.
Oh, it's for sure not a done deal. But IIRC bees are known to have at least some capacity for memory, in terms of seeking out known food sources that go beyond just following a pheromone marker. I don't know if the mechanism behind this is well understood, but going by the blue/red trick it seems like they can to some extent 'learn' to recognize various types of flowers and seek them out by memory. Whether this type of information is transferable between bees in ways /other/ than a pheromone marker that guides other bees to the same immediate sensory experience of "blue flower = good nectar source", I don't know. But it's intriguing!
Also worth noting that bees can transfer navigational information to eachother by 'waggle dancing', so they have multiple possible vectors for information sharing going for them.
I am a dog owner, my gf is a dog owner. My dog does this very particular thing when he wants attention and my gfs dog have started doing the same after some time.
Humans may be good at learning from others and pick up stuff fast but that don't automatically mean we're unique in our ability to do so. This idea that humans are so special feels very odd to me.
Social transmission of behavior in macaques has been documented since the 1950s. See the potato washing studies of Imanishi et al (ref. #9 in the main paper).
Humans create and pass along culture, but we also have the capacity for reasoning about ethics, justice, etc. When we say animals are also very intelligent etc. with a couple of examples, it is worth contextualizing the edges or shortcomings of the intelligence on the gradient.
Zooming out and looking down on the intelligence spectrum, we have to be humble and concede that we, too, while far superior to other living creatures we have encountered so far, probably also lack some dimensions that even higher intelligence would unlock.
Does this necessarily mean "learning" in the sense that humans use that word?
Approximately 15 millions years ago, leafcutter ants completed the 30 million year process of domesticating another species [1] (something also done by ambrosia beetles and termites, not just humans) - the Leucoagaricus gongylophorus fungus cultivated by the adults is used to feed the ant larvae, and the adult ants feed on leaf sap. The fungus needs the ants to stay alive, and the larvae need the fungus to stay alive, so mutualism is obligatory. All with 250,000 neurons!
Whether it was the fungus that domesticated the ant or the ant that domesticated the fungus, can passing on this skill also be considered a learned activity?
> Then the Dutch-led team of researchers trained two demonstrator chimpanzees to show the others how it was done.
> After two months, 14 "naive" chimps had mastered it.
I am not sure the study is showing human like “cumulative culture”. It is showing that if humans train them, they can show others, which is different from humans learning a thing on their own, showing it to others, and then someone else using that as a foundation to discover something new building on that.
Basically, the study is showing that a chimp can teach other chimps something humans trained it to do.
I don’t think that is a novel or controversial finding. If you grow up around dogs, you know that one trained dog will help the other dogs learn.
"cumulative culture" is also called "the ratchet effect" ... when you learn something from someone else and then improve on it and pass on the improvements, over and over and over across generations.
This is just learning, there was no ratchet effect.
How cumulative it is, is a fair question to ask. But I don't think animals passing on "culture" has anything to do with humans. Look at lion prides, where one pride has learned how to hunt elephants while another buffalo, and the young ones pick it up from the older generation. But put one member of one of those prides in the other's spot and they'll be clueless until they learn it from the others.
I agree completely. I think “passing on culture” in terms of training younger members to do some skill has been known for millennia. Dogs do it. Lions do it, I am sure all social animals do it to some extent.
Haven’t octopuses already shown this to be the case? Learning from each other (even though they are pretty not-so-social creatures). The claim that bees are the first invertebrates to do social skill passing seems weird.
"an ability previously thought to be unique to humans".
This sentence.
I’ve read and heard it so many times now, that according to Bayesian statistics, I should correct my assumptions and assume that, in the end, we will find out that there is not a single thing that’s unique to humans.
The general sentiment is spot on - animals are amazing, complicated, sophisticated, and we will continue to be "humbled" over and over again, but maybe you've gone too far.
A few candidate "unique to humans" characteristics:
- surgical replacement of defective organs. I wouldn't be surprised if some animals have processes for amputations - especially insects - but I'd be pretty surprised if there are any kidney transplants going on.
- written language systems for durable information passing, phonetic alphabet systems. As far as I know, a phonetic alphabet was invented only once among humans, which makes it a rarity even for us.
- haircuts. This one is super plausibly wrong, but I don't think that any animals do this, and its a good example of something that they could do but just don't.
> As far as I know, a phonetic alphabet was invented only once among humans, which makes it a rarity even for us.
A bit of an aside, but I'm a bit puzzeled by this statement. There are many phonetic alphabets invented through history of course. Are you saying they all have the same root or inspiration?
> A bit of an aside, but I'm a bit puzzeled by this statement. There are many phonetic alphabets invented through history of course. Are you saying they all have the same root or inspiration?
I think you're thinking "alphabet", which indeed, has been independently invented multiple times. The most widely used alphabet is the one derived from the Phoenician script. Other alphabets that were independently invented include the Brahmic scripts and the Hangul script.
Wear clothes - I know that hermit crabs wear shells
Make music - I know that birds sing, and a quick Google search shows whales and seals seem to make "music" as well
Theorize - Depending on what you mean of course (but I agree with you about e.g. scientific hypotheses to be tested), but as an example apparently birds "theorize about the minds of others", again by Google search
Writing - Definitely a stretch here (and so I agree with you) but animals do leave chemical markings for others
> Wear clothes - I know that hermit crabs wear shells
You show me one outfit, I raise you an entire textile, clothing, and fashion industry allowing us to show our features in displays of status all the way to enabling us to walk in space.
Both degree and kind. While hermit crabs and other animals who wear garments do very little manufacturing (they use something as is or do some “minor” fashioning), the techniques and kinds of clothes we make are of a totally different sophistication. Our clothes allow us to enter ecosystems and hostile environments, for example. Our clothing is both functional and cultural. Etc.
> Were we less human when we couldnt yet replace defective organs?
Over time, humans have demonstrated a vastly more impactful ability to manipulate our environment and ourselves than any other living organism we have encountered.
There aren’t many fundamental unique differences between an 8086 and a M3 Max.
But the differences in scale are so vast that it opens massively different capabilities.
At some point, quantitative differences are so large they become qualitative differences.
Although both can hold matrices in memory and do math operations, the M3 Max can hold so much more and run math operations so fast that LLM inference becomes possible making it seem “intelligent” on a whole different level than an 8086 even though they are at some fundamental level very similar.
You're implying that bees are less intelligent than humans.
However they can sense electromagnetic fields and can sense whether a flower they're interested have pollen or not without looking at it with their eyes.
Elephants and whales can communicate over vast distances via sound. Bats can see without eyes. Salmons and pigeons can find the point they have born without even trying. A dog can smell history of a place, plus get much more information from a single smell.
Humans can do none of these things without tools.
Also, in electronics, there are accelerators which are much simpler in transistor count and architecture, but which can do much more than a more complex counterparts. GROQ inference cards and FPGAs come into my mind.
So neither capability, nor capacity in numbers is a valid measure for intelligence or capabilities in practice.
Just because a bee has less neurons than a chimp doesn't mean it can't have some kind of comparable intelligence when you compare the things they can accomplish.
Oh, also crows understand and exploit physical phenomena and can manipulate things with tools to get what they want.
> You're implying that bees are less intelligent than humans.
Intelligent is not a synonym for “have amazing capabilities” but rather the ability to process and adapt to new information and transform their environment.
No other life form comes even close to humans in that regard. Our abilities (for better or worse) are godlike compared to other animals. There is a reason this epoch is called the “Anthropocene”.
This isn't so correct. Humans in isolation aren't so vastly superior from other animals, if you observed primitive humans and your typical troop of chimps it wouldn't be the slam dunk you're singing about. The truth is we're just enough better to develop and use written communication, and the majority of human progress beyond that has been trial/error and imitation with successes preserved through writing. Transmission of the written word enabled humanity to become a collective intelligence of sorts, individually we aren't that smart.
Our communication and experimentation was sufficiently far ahead of other advanced mammals to cause mass extinction events long before writing came into play. (It's debatable whether some fairly advanced civilizations even had "writing" in the sense of general symbolic communication)
"Just enough better" is doing a lot of work when that leads to space travel and computers and inventing reasons to feel bad about our phenomenally efficient expansion and predation, whilst members of other species prove extremely limited at acquiring human knowledge even when we're expressly trying to share it with them.
I think the argument in this thread is more between people thinking animals have 'zero' ability, it is just instinct, or some other mechanism different than what is taking place in a human. And the group that is saying, animals are using the same mechanisms as a human, just humans are scaled up. Abilities between Humans and animals is a sliding scale.
They are wanting to argue that Humans have a singular ability that does NOT exist anywhere else, kind of a 'divine' argument. And animals are doing something different.
"comes even close" is indicating, it's a scale, might be a big scale, but it is the same scale.
>You're implying that bees are less intelligent than humans.
>However they can sense electromagnetic fields and can sense whether a flower they're interested have pollen or not without looking at it with their eyes.
Which, yes, like your other examples, is an impressive feat, but not an impressive feat of intelligence. Perceiving one or another part of the spectrum isn't intelligence. Neither is hardwired responses. In contrast:
>Oh, also crows understand and exploit physical phenomena and can manipulate things with tools to get what they want.
That would be a feat of intelligence. But you're lumping all impressive feats into "intelligence", regardless of what's responsible for them.
> You're implying that bees are less intelligent than humans.
While bees thrive in their ecology by using intelligent behaviors like dancing, counting, navigating, remembering colors and scents, and making decisions using a process that resembles cost-benefit analysis, the sheer scale and complexity of human intelligence is unparalleled in teh animal kingdom. Our neocortex facilitates advanced cognitive functions, including our capacity for reflection, conceptual thought, and technological innovation generally puts humans at a higher level of cognitive complexity.
It’s a common trope that “we are no different from animals,” but this is not true. Of course we are.
Far more interesting than the fact that animals “can” ape behavior is how they do - and how that differs from humans.
Check out the work of Richard Byrne [1]. Our best theories suggest that the way primates share behavior is through literal parsing and replication of sequences. A chimp is able to watch a fellow chimp complete a complex maneuver to open a nut, and can replicate that sequence. But the sequence can contain odd moves that obviously have no effect, and those moves too will be replicated. This signals a lack of understanding. Primates use an advanced copycat mechanism, which makes innovation very slow.
The way humans learn and transfer behavior is vastly different. We watch another complete a task and develop an explanation for how it works. Our reproduction isn’t us following a literal sequence, but using our understanding/mental model to solve the task. For example, if I watch a complex sequence to open a tricky nut, I’ll have developed a model for the physics of the nut, weak points, ways to get leverage, etc. I might memorize and repeat your sequence, but I might also get creative with my own method, knowingly or not.
And, of course, we can do this one-shot.
It’s the difference between a parrot mirroring words and a person retelling a story they heard.
We are super different than animals, and denying that fact will hide important secrets about knowledge and creativity.
The "extremely complex" problem solving in this study, for chimps, was putting a ball in a draw, and then closing it. Would you have assumed that training a chimp to do so would be impossible before this study? Or that it would be impossible for other chimps to ape him, so to speak? In other words, is this study actually shifting your assumptions in any way? Or is it just the arguably grandstanding language being used that may be swaying you?
So for instance I think it's pretty well known that monkeys in Bali have taught themselves how to steal from tourists and then exchange the goods for treats. [1] And that was entirely self-learned/taught. That, to me, seems somewhat more impressive than putting a ball in a drawer, yet didn't really leave me with any awe beyond what one would normally have when interacting with monkeys. And such things leaves me decidedly unimpressed with chimps and a drawer, and even more cynical about the language used to describe it.
My assumption is that difference is not that something is missing but just in complexity of things.
It is strange that people put other mammals in same category with birds for example. Thinking that humans are separate kingdom. I'm closer to pig than pig is to chicken.
I believe the one truly unique thing humans can do is imitate complex actions with high fidelity. When we evolved this ability it lead to an explosion of behavioral complexity simply because imitated behavior satisfies all the criteria for evolution (s/o Dawkins).
We’re not unique in our ability to imitate, but we are unique (for now) in our ability to imitate well enough that imitated behaviors are stable enough to evolve.
There are of course some Australian birds that famously spread fire, so partial credit IMO.
Also, figuring out how to make fire is pretty hard. I’d argue it isn’t something “humans do” in general. It is one of the earliest examples of something that somebody figured out, (or maybe it was figured out independently in a couple different places) but mostly it is a taught skill.
Teaching and long distance running are our special abilities. In both cases other animals might do the thing, but we’re much better at it than they are.
IIRC that's more like a Australian media stunt from one single source, never verified or replicated.
There's this one guy trying to explain why there's wild fire and it's not any human's fault, which inadvertently caused a great debate "animals can utilize fire".
> I agree about fire, but I'm pretty sure you can count bees making honey as cooking.
Cooking is to apply heat to transform raw ingredients into a different, often edible form. Bees make honey by collecting nectar from flowers, storing it in their honey stomach, and then regurgitating it into honeycomb cells. During this process they add enzymes to the nectar and fan their wings to reduce the water content, transforming it into honey. Bees transform raw material, but not by heating it.
Eagles can 'see' better than humans. No one is is up in arms about how wonderfully unique human eye-sight is, how Eagles are using 'just instincts'. Humans are divine in how complex their sight is, that it could never occur in other animals or machines. Totally impossible that we would ever have a machine that can 'capture images'.
But there is laundry list of features that some animal does better.
Humans on whole do everything 'ok'. Humans have a lot of features, and for each feature they aren't the 'best', but on aggregate they do them all together on average better. They are one animal that combine total features better than any single animal.
Guess my point, is when it comes to 'mind' people are still thinking humans are divinely 'exceptional'. But really, it is one more 'feature' that the human animal does best, but that does not mean other animals don't have the same feature, but scaled down.
Seems like people complaining that 'bees' aren't intelligent are just trying to keep humans propped up on a pedestal. When really, it is one more feature that is scaled.
Does it effectively keep the humans on a pedestal to consider our traits as fundamentally different?
I think it is all on a continuum, humans are animals. But putting us on that continuum makes all the animals look like absolute garbage, intelligence-wise. If we want to put humans in the game, we’re going to be absolutely dunking on, like, every animal in terms of ability to abstract and teach concepts.
Technically humans are dunking on all other animals. We're killing them all rather handily.
I'm arguing against the group that think humans aren't animals, that they are imbued with some extra metaphysical secrete sauce that we are incapable of understanding.
This group usually come out strong against any of these studies that hint at some continuum that includes humans and animals on same scale. Even if humans are on far end of the scale.
Didn't Kubrick prophesy this? I'm thinking of the first act of the movie 2001: A Space Odyssey where one advanced primate learns how to use the femur of a tapir as a tool, then teaches the other primates in his tribe how to use bones as a weapon and how to walk upright (though we are not shown the teaching or knowledge transmission process).
At some point, can’t we just admit to ourselves that we operate from a human-centric worldview and that clouds pretty much all of our thinking on what makes humans special (if anything)?
As a species we definitely have some narcissistic tendencies.
> At some point, can’t we just admit to ourselves that we operate from a human-centric worldview and that clouds pretty much all of our thinking on what makes humans special (if anything)?
I think this is true. The evidence being our ruthless exploitation of our fellow animals without any regard to their interests, suffering, etc.
while I'm very far from subscribing to the idea, yet there is for example that scientifically looking article arguing for the genetic instead of the cultural transfer in humans of the skills of making of those stone handaxes. Would some future scientists argue for the genetic transfer of python programming skills in the humans of the 21st century...
I can't remember where I read it, but I remember a similar study with rats. Previous generations were shocked when interacting with something specific. Current generations seemingly already knew not to interact with that object. I think they even expounded to study the effects on rats born within an entirely different locale too.
as I don't believe in the genetic transfer, the only other way it seems to be is that the older rats would be telling youngsters a tale of "a shiny prongy thing which gives a painful scare on touch", and such possibility seems very fascinating for me.
And dogs. And cats. And beavers. And most ever mammal that raises young. What is more interesting and more telling imho is to look at what happens to animals that are not by their animal parents. We all know that dogs not raised around other dogs don't develop doggy social skills. Geese won't learn to migrate without other geese. Teaching and learning is more common in the animal community than most researchers want to admit.
I don't see how this was thought to be unique to humans? For instance, it seems to be common knowledge that big cats have to teach their kids to hunt by showing, for instance.
Edit: Going to the source papers clarifies a little, the hypothesis they're challenging relates specifically to social learning for "the knowledge of the production of a particular trait" and not learning in terms of recreating existing behaviors.
This would probably also be why pets and birds would be excluded. While they can learn and teach each other various things, most of those things are not relating to production and can be difficult to separate from behaviors that should be considered to already be part of the species' repertoire. Bird song or whale song may be a sort of learned "culture" but it's so endemic to their species that it's treated as distinct from what's being discussed.
I am still pretty intrigued by the whole topic even if it's a spectrum.
From what I've read, beavers build dams instinctually if they hear running water. This applies to beavers raised in captivity that were separated from their parents. It's both an unusual behavior and one that seems (on the surface) like it would take plenty of reasoning.
My cat didn't knew how to jump over the chain link fence until he saw a neighbor's cat do it.
But the point of this study is about extremely difficult problems that an animal couldn't solve by itself, unlike my fence that the cat would have probably solved by brute force given a certain time.
Many animals are able to hear outside of our range of hearing.
I tried to capture in when I noticed it happening on 2015, but you can't easily hear the sound, some of it was audible to me, but not with the camera I had (I had to grab the closest thing I could find).
If you watch at 0:50, you'll see in a few seconds the crow is doing a whole-body movement. That's it making the sound, which I could feel against my skin. If you crank it, you can hear the weird, audible part of it.
Then, before they really went to work again... one of the crows noticed me. And you can see at the end, the other gets really pissed that it didn't notice me.
Basically, he walks to where he can see in my house, spies me, and is all "YOU BASTARD!"
Part of an email I sent to a friend at the time:
--
This is the first time I've noticed this behaviour.
Crows can make sounds, and hear sounds, outside of our range of hearing.
And, it was raining. Soaking the topsoil.
It is also grub season. Soon, June bugs will appear, but right now they
are in grub form, but pupating. I'm not sure if the added density helps
in detection or not.
You'll likely have to turn this up. I had to grab the nearest device I
had, so the resolution sucks too.
Watch the larger crow, as it walks around, sounding for grubs. It did
this over a large portion of my lawn.. systematically.
So 14 out of 64 chimps learned a relatively simple trick after two months of "cumulative culture" exposure. Just over a 20% success rate?
> Thornton said the research again showed how "people habitually overestimate their abilities relative to those of other animals".
People learn how to tile floors, install dishwashers, change their cars oil filter, build websites etc. all from mediocre youtube videos. I don't think I'm overestimating humans "abilities relative to those of other animals" ...