Comment by tombert | Hacker News Reader

Comment by tombert | Hacker News Reader

Parent Full thread

I am human (at least as far as I am aware), though I guess with the advent of LLMs that's becoming less evident.

It seems obvious to me, though I might be redefining "smartness" in a very human-centric way.

It seems pretty clear to me that humans are the smartest in terms of analytical ability; you hear things in the media all the time like "crows are smarter than 7 year old humans", but I think that really depends on how you measure it. I am pretty sure I could teach a seven year old child that four quarters, ten dimes, twenty nickels, one hundred pennies, five dimes and ten nickels are all equal to the same amount. I don't even think that a seven year old human child would struggle with that terribly; I'm pretty sure I knew that five one-dollar bills was the same thing as a five dollar bill at that point. I don't think any other animal is capable of making those equivalencies though; I am pretty sure that a crow could not be taught this, no matter how hard you tried.

Now, it's entirely possible that assigning weight to this is demonstrating bias towards human concepts; this stuff might be obvious to humans because humans invented it, and thus comparing human and non-human ability in terms of human creations might be an entirely malformed premise. I know nothing about neurology or philosophy (or really anything outside cartoon trivia), so this is just kind of "gut feelings" on my end, which I will admit that that's not very firm ground to stand on.

> are all equal to the same amount

This may be modern schooling at work. I've heard that one motivating factor for the ancient egyptian's strange-to-us mathematical style ( https://en.wikipedia.org/wiki/Egyptian_fraction ) is that when they divvied stuff up, they needed to do it in a way that was more obviously fair than we do.

Example: we have four workers on our gang and need to pay them with 3L of beer and 3 loaves of bread at the end of the day. The beer is easy; just pour equal levels in three similar containers. For us, the loaves are easy too: give 3 workers 1 * 3/4 loaf and the last one 3 * 1/4 loaf. For the ancient egyptians (who might complain the last one either got smaller portions or more things) it'd be important to give everyone obviously the same portion: 1 * (1/2 + 1/4) loaf.

Oh this is by no means limited to ancient times, try giving any two young children the same amount of food, but cut into different sized pieces - smaller cuts result in more pieces than the larger cuts... Guaranteed meltdown. :)

That's a good time as any to learn that might makes right.

The simple solution- have one cut the food in half, but the other gets to choose their half first from the two.

There is a pretty profound communication gap that you seem to be discounting.

A closer analogy might be trying to teach a 7 year old who speaks a different language. Or even deeper, a 7 year old child who only knows sign language and you do not.

The communication gap makes it very hard to figure out how much is a “intelligence” difference versus just being unable to even communicate the contents of the test.

Sure, but even then isn't that making an assumption that animals have the same capacity for communication that we do?

Let me try and explain; I don't speak Spanish, but if I had teach a 7 year old who only knew Spanish, presumably I, as a human, could learn Spanish and then teach the 7 year old whatever I needed to.

I'm not sure there's an equivalent "crow" language that is even possible for anyone to learn, at least in the same way that humans have language. I know that animals "communicate" in some senses with mating calls and the like, but I don't think they have anything even approaching "grammar" and "syntax".

But I suppose if you're going on a deeper level, if you had some way of directly communicating with a crow (like beaming information directly into their brain already translated in a form that they could understand), I still don't know that you could teach them that money equivalence analogy I used.

Again, I'm speaking out of my ass here, I don't know anything about this stuff, just spitballing.

No. I am pointing out how being unable to communicate makes it very hard to determine intelligence. It is hard to determine if another human is smart without being able to speak the same language or explain the procedure and contents of the test.

Given that there is no known means to effectively communicate with animals, we are restricted to tests that do not demand shared communication to identify intelligence. These tests must be benchmarked on mutually unintelligible humans.

For instance, a grade school Chinese exam might be a good way of gauging the intelligence of Chinese children, but is a terrible way of gauging the intelligence of French children. Such a test is lousy in certain domains. We must be cognizant of such testing methodology limits in the design of animal intelligence tests.

tl;dr A test that would fail animals and speakers of a different language are not good tests. An animal intelligence test needs to at least pass all humans of comparable intelligence regardless of language to reach the bare minimum of acceptability.

Raven’s progressive matrices, obviously intended to test the intelligence of birds.

“four quarters, ten dimes, twenty nickels, one hundred pennies, five dimes and ten nickels are all equal to the same amount” - only a human could come up with something so unintuitive.

The crows can't believe our foolishness - a single silver dollar is shinier than even a hundred pennies!

Is this more of a limitation of our ability to communicate with the animals? If we could talk intelligibly with a crow, maybe we could teach it? Conversely, if our communication with a 7-year-old child were limited to the level we achieve with crows, could we teach the child and know for sure that they understand it?

> could we teach the child and know for sure that they understand it?

Yes, children aren't born knowing languages so you teach the child the language and then you can see it understands it. You don't need to share a language to teach language. We haven't been able to teach any animal language at even 4 year level, so animals at least doesn't have the same level of language intelligence as humans do.

You are going beyond the constraints, “if our communication with a 7-year-old child were limited to the level we achieve with crows.”

We know we are able to teach children language because we have shared modes of communication through which we can observe their learning; speaking, writing, hand gestures.

> You do not need to share a language to teach a language

True, but I would argue that you need to share a mutually intelligible mode of communication.

How would you teach (and know that they have understood it correctly) the concept of monetary denominations to a 7-year old child if you were limited to communicating on the level you can with a crow? I have no idea, and would probably give up. But this does not necessarily tell us anything about the intelligence or capacity for learning language in the child, it tells us only of the gap in communication that we cannot bridge.

> so animals at least doesn’t have the same level of language intelligence as humans do.

This does not necessarily follow. What follows is that humans are not able to teach animals human language on the same level as they are able to teach humans.

We’ve not been able to reach any animal language at a four year old level that they can speak but I think we have taught animals language at a level that they are able to understand? Though I guess we can’t necessarily prove that understanding.

Reply on news.ycombinator.com