Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> We now know that bees can count, recognise human faces and learn how to use tools.

I was curious what that means in this context and found this research (co-authored by Prof Chittka mentioned in the article): https://journals.plos.org/plosbiology/article?id=10.1371/jou...

Apparently only a small minority of bumble bees can figure out how to pull a string to access a reward, but then other bees adopt the behavior by mimicry. IMO I think we're doomed to move the goalposts on intelligence for a while, like with the statistical abilities of LLMs to manipulate language and insects' ability to use tools. Moravec's paradox keeps rearing it's ugly head as more and more complex systems turn out to be relatively easy compared to basic cognition (the system that keeps them flying and identifies threats, flowers, etc.).

It'd take a lot more to convince me that bumble bees are conscious just because of the their brains' simplicity compared to humans or other animals that appear more intelligent like pigs, corvids, octopuses, etc. I'm not categorically against such a possibility, but I think the bar for recognizing intelligence in general has been set too low.

This sidesteps the main problem anyway: What is consciousness? I don't think we're any closer to rigorously defining that anymore than intelligence.



> Apparently only a small minority of bumble bees can figure out how to pull a string to access a reward, but then other bees adopt the behavior by mimicry.

Humans are the same, try teaching basic math equations to kids. When I was in school, most of us learned math by memorizing steps to solve the problem instead of understanding it. Then on the test if we found something we couldn't recognize, we wouldn't be able to solve it. There were usually only a couple of kids in the class who really practiced and understood the principles. I believe most humans usually learn patterns and rarely have genuine insight or grasp of the subject.


Even as an adult I don't really know how anything works, lol. I just copy whatever's on Stack or GPT, get paid in magical numbers, and exchange those numbers for food like my mom did. At least a bumblebee knows how to make flowers. When we want to grow something, we end up copying them...


> I think the bar for recognizing intelligence in general has been set too low.

I think that the problem is not being too low, but being not very clearly defined, and by assuming that you can just be quantified by numbers; I think that intelligence (and consciousness) cannot be properly quantified by numbers.

> What is consciousness? I don't think we're any closer to rigorously defining that anymore than intelligence.

It is a valid point. You will have to define such things clearly in order to know them scientifically, properly.


> IMO I think we're doomed to move the goalposts on intelligence for a while

I'm old enough to remember back when we thought language was the defining element of intelligence. Dogs can't talk - dogs aren't intelligent.

Now LLMs can talk, and we've shifted intelligence to mean animal intelligence - being able to predict the motion of a falling ball, wanting to protect your children, etc.

Some people truly cannot tolerate the idea that our intelligence/existence isn't magical, so they'll desperately move it again and again and again... forever. Watch.


I still feel like animal intelligence is completely downplayed.

Dog's can learn new environments and tasks. The simple act of recognizing what a door is and how they work takes quite a bit of intelligence.

As far as I can see, AI still can't make a robotic dog.


There is some conflation between consciousness and intelligence going on. Consciousness is subjective experiences. Intelligence is cognitive ability. There isn’t a necessary link between the two. We can say LLMs are intelligent but not conscious. We could say a lizard is conscious but not that intelligent.


"Apparently only a small minority of bumble bees can figure out how to pull a string to access a reward, but then other bees adopt the behavior by mimicry."

Replace bumble bees with human beigns and "pull a string" with an engineering task and you'll find we are not so different. /sarcasm, but not too much, average human being is quite dumb, tbh.


> average human being is quite dumb

Compared to what? Average human beings are geniuses compared to animals.


Well, I think that depends upon what is being measured. Is this general intelligence, or is this mastery of a single thing?


And compared to what manner of assessment of that genius? Animals had the Earth to themselves for millennia. Humans for a very short period. You judge!


That's ok. Moving the goalposts (in good faith) is part of the proceed of understanding the question we're trying to answer. Some times you don't realize you don't understand the question until you start thinking about it.



If a rigorous definition of consciousness existed, how would you be convinced that it was correct?


Same way I am convinced of relativity without having recognized experiencing the effects. Deduction and consistent observation. That's just me. Getting people to believe in things, is not a hard problem.


Consciousness is the inflection upon the potential of existential being. Literally life technology animating the quantum sieve within our neurons. Consciousness isn't a complex deck of cards, it is the singularity inside the quantum sieve.

Intelligence is the mitigation of uncertainty. If it does not mitigate uncertainty it is not intelligence. All that other stuff about more or faster or sophisticated is something else. obviously we're describing a scalar domain. Your expectations overload and out leverage the simple truths.

This may be the moment to evaluate our questions.


Even an LLM couldn’t generate this.


It's an interesting perspective. I understand that the language choices they made were strong and thus not for everyone, but I don't see a reason to be rude about it. If it didn't make any sense to you it's probably because you didn't find it worth your time to parse it out, or similarly that there was unfamiliar terminology you didn't judge it worth your time to look up. That's fine, but that's a choice you made, no sense blaming them for it.


Information is the removal of uncertainty, if it does not remove uncertainty it is not information.

Uncertainty is the soft synonym for entropy.

Entropy is the distribution of potential over negative potential.

It is not that Truth is impossible to find, it is that undeceiving the self is difficult.

Undeceive yourself of your own words if you will find something you have not anticipated.


What about if I think I know something and then realize I don't? Didn't I receive information that increased my uncertainty?


When you do a math or logic problem you would intuitively use parentheses, allowing the problem to be split into scopes. Use parentheses in your mind and then regard where uncertainty is increased and where it is decreased.

In nearly every circumstance, new information will generate MORE questions. To the untrained mind this may sound like more uncertainty, however in actuality you did not know what questions to ask until some preliminary uncertainty was resolved.

For one scope, uncertainty is resolved, in doing so new domains are revealed.

Uncertainty (or information for that matter) isn't a lump sum that grows or shrinks. One question is resolved revealing more separate and further uncertainties, which are very specific and reduced from chaotic infinity by some resolve . In this case, realizing you didn't actually know that "something" to begin with.

The tendency to jumble these delicate distinctions into a single glob of unknowns obfuscates otherwise simple (however numerous) truths. And by "truth" I mean, anything consistent with existential reality.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: