Biological Intelligence and Learning

This is for discussions about news, politics, sports, other games, culture, philosophy etc.
France iNcog
Ninja
Posts: 13236
Joined: Mar 7, 2015

Biological Intelligence and Learning

Post by iNcog »

I'm interested in going down this rabbit hole.

It is my belief that environment and exposure to different material has an incredible influence on what you know as a human, as well as your learning capabilities. My understanding of the definition of intelligence is, quite frankly, nebulous at best.

Consider this thought experiment. If you took yourself and placed a genetic copy of yourself at 5000 BC, would you know as much as you know today? Absolutely not. The foundation of human knowledge just isn't there for you to learn and know it. Even if you were placed in that time period with all of your current memories, you would probably barely be able to expand upon the knowledge you do possess.

Now, imagine that a genetic copy of yourself is now placed even 200 years into the future. I don't believe it unlikely that not only would know more things, you would be able to learn more advanced and difficult topics more easily than today.

I believe that the environment in 200 years will yield a better understanding of learning itself. Today's practice of throwing 30 kids in front an underpaid teacher with a whiteboard is probably not the actual most efficient way to give a young mind all the tools it needs when it comes to learning things and knowing things.

Genetics may play a part in intelligence, I think that there is science that is able to back this up. However, I am uninterested in the genetic influence on intelligence nor do I care to have a discussion about the nature of intelligence itself.

What is learning? How do we learn to learn?
YouTube: https://www.youtube.com/incog_aoe
Garja wrote: ↑
20 Mar 2020, 21:46
I just hope DE is not going to implement all of the EP changes. Right now it is a big clusterfuck.
France iNcog
Ninja
Posts: 13236
Joined: Mar 7, 2015

Re: Biological Intelligence and Learning

Post by iNcog »

On a fundamental level, learning takes place when insights (information perceived by us through our senses) pave the way for a change of behavior of understanding.

I'm woefully unacquainted with neuroscience but as far as I can gather, neurons and synapses working in tandem allows us to do things like pattern recognition and other such things.

1+1 = ?

You immediately thought 2.

19+3 = ?

You immediately thought 22.

You didn't necessarily count to 22 from 19. Your brain saw this pattern that it had seen before and it triggered the appropriate response. Interestingly, I think that we approach this simple math problem differently. I know that "9+" a "number" will give me +10 - 1.

Thus,
9+3 = 13 - 1 = 12

9+27 = 37 - 1 = 36

9+8 = 18 - 1 = 17

etc.

BUT, that is the pattern that my brain will use. If you've seen enough of the same math problem, you don't even do the intermediate steps anymore, you just "know" the answer.

That's because, to my understanding, your neurons just eventually understand/know what the end result is and skips to it, thus bypassing the intermediate steps. Here is the basis of what could be considered a semblance of intelligence.

---

This topic is also interesting to think about when comparing with my pets. We have 2 cats and a big, silly, golden retriever. I've become dimly aware that my dog behaves differently depending on which circumstances. She has learned to associate that me putting on socks isn't necessarily cause for excitement. My girlfriend putting on socks is more intriguing, but still not the coolest thing. If my girlfriend and I both put on socks, she immediately goes apeshit since generally when we both get ready to leave, she will most likely be coming with us.

There is no reason to assume that she is coming with us, it's just that her brain has learned to associate certain patterns with a probable outcome.

This leads me to believe that intelligence and learning can be correlated to how many patterns you've been exposed to, how many different kinds of patterns, etc. The more patterns you go through (even in unrelated fields) the more you should be able to recognize patterns from stuff you haven't seen before. Your neurons are just kind of tuned for it at some point.

I feel stupid spouting blatantly obvious things, but frankly I'm not sure how else to further my understanding of the topic without taking it from the top.
YouTube: https://www.youtube.com/incog_aoe
Garja wrote: ↑
20 Mar 2020, 21:46
I just hope DE is not going to implement all of the EP changes. Right now it is a big clusterfuck.
User avatar
Greece BrookG
Retired Contributor
Posts: 2009
Joined: Feb 21, 2016
ESO: BrookG
Location: Thessaloniki

Re: Biological Intelligence and Learning

Post by BrookG »

This sounds correct. Details aside, humans are really good at discerning patterns, even when there are none (see pareidolea). Many pedagogical theories also are based on pattern learning. Another advantage of humans is that they have the ability to apply unrelated knowledge to solve a problem and be creative. There is a study that compares how ai learns to play some game Vs some human and it argues how humans are creative with prior knowledge.

In my field, it's common ground that exposure to tonal music in WEIRD societies mostly shapes our preference and liking to tonal music.


EDIT: On a second read, I notice you mention predictions and expectations. Predicting a future outcome is crucial for an organism. If they couldn't predict, they wouldn't be able to act fast enough to avoid a danger or so. There is a theory called ITPRA (standing for Imaginantion, Tension, Prediction, Reaction, Appraisal) that summarises expectation mechanisms and how we behave to confirmation and thwarting of expectations. Check his book Sweet Anticipation, especially its first chapters are very easy to read and not too technical.
Correlation doesn't mean causation.
http://www.tylervigen.com/spurious-correlations

"mr.brookg go buy jeans and goto the club with somppuli" - Princeofkabul, July 2018
User avatar
Netherlands Goodspeed
Retired Contributor
Posts: 13002
Joined: Feb 27, 2015

Re: Biological Intelligence and Learning

Post by Goodspeed »

Our brain's plasticity is what makes us so good at learning from our environment. It was a very important evolutionary step because it allows us to grow up in an environment that is completely different from the one where our species first evolved, and still be able to function in it. It's the reason human babies are so utterly useless for much longer than babies of other species. This is, as they say, not a bug but a feature. Babies are blank canvasses with very little "instinctual" knowledge (think of birds flying to specific places to breed, cats instinctively knowing how to hunt, etc) that can be molded to fit within the society they grow up in.

What we are born with is a strong ability to recognize patterns. By recognizing patterns and forming neural connections we gain intuition, first, and then the ability to think in multiple steps, which we need in order to solve problems that are too complex to intuit the answer. We can intuit the answer to 1 + 1, but not the answer to 29 * 14. To solve the second one, we think in multiple steps.

They are the same process in our brain, though. Our brain is ultimately just a data processor. And when you input 1 + 1 it quickly outputs 2. When you input 29 * 14, it takes a different path through the neural network and does the steps required, then finally outputs 406 (hopefully).

It's a relevant topic, as AI seems to have now reached superhuman levels of intuition, but has seemingly not yet learned to reason in multiple steps. Let's call this ability planning. I think humans learn this around age 3, maybe 4? It's hard to estimate.
Interestingly, in many aspects GPT-4 is already way beyond 4 year olds. It is able to pass master-level exams and do well at math competitions. But it's very bad at planning. Why is that?

This is speculative, but imagine what would happen if you just threw all of the text data on the internet at a human baby. Imagine it had the time to process it all and form neural connections. Would it have learned how to plan? Maybe not. Maybe that takes actual experience at problem solving. Maybe forming the connections to think in multiple steps only happens when you try and fail to solve problems on pure intuition. And maybe that is key in further developing these AI models.
No Flag RefluxSemantic
Gendarme
Posts: 5996
Joined: Jun 4, 2019

Re: Biological Intelligence and Learning

Post by RefluxSemantic »

Are we really blank canvasses? Seems like lots of stuff is preprogrammed actually.
User avatar
Netherlands Goodspeed
Retired Contributor
Posts: 13002
Joined: Feb 27, 2015

Re: Biological Intelligence and Learning

Post by Goodspeed »

The blankest, relative to other species
France iNcog
Ninja
Posts: 13236
Joined: Mar 7, 2015

Re: Biological Intelligence and Learning

Post by iNcog »

BrookG wrote: ↑
03 May 2023, 20:03
This sounds correct. Details aside, humans are really good at discerning patterns, even when there are none (see pareidolea). Many pedagogical theories also are based on pattern learning. Another advantage of humans is that they have the ability to apply unrelated knowledge to solve a problem and be creative. There is a study that compares how ai learns to play some game Vs some human and it argues how humans are creative with prior knowledge.

In my field, it's common ground that exposure to tonal music in WEIRD societies mostly shapes our preference and liking to tonal music.


EDIT: On a second read, I notice you mention predictions and expectations. Predicting a future outcome is crucial for an organism. If they couldn't predict, they wouldn't be able to act fast enough to avoid a danger or so. There is a theory called ITPRA (standing for Imaginantion, Tension, Prediction, Reaction, Appraisal) that summarises expectation mechanisms and how we behave to confirmation and thwarting of expectations. Check his book Sweet Anticipation, especially its first chapters are very easy to read and not too technical.
I think I'll try to get my hands on that book, thanks for the recommendation. :)

I think the ability to apply unrelated knowledge to solve a problem is definitely another aspect of human intelligence. I had a professor when I was studying that explained that you could either have broad knowledge of many different things, or you could be highly specialized in a domain. The specialized approach seemed to be favored by the professor and I do think that this is where you can make the most serious advancements. But I cannot help but feel that having broad knowledge can be equally beneficial. However said benefits are less tangible, I would surmise.
YouTube: https://www.youtube.com/incog_aoe
Garja wrote: ↑
20 Mar 2020, 21:46
I just hope DE is not going to implement all of the EP changes. Right now it is a big clusterfuck.
No Flag RefluxSemantic
Gendarme
Posts: 5996
Joined: Jun 4, 2019

Re: Biological Intelligence and Learning

Post by RefluxSemantic »

Goodspeed wrote: ↑
04 May 2023, 17:00
The blankest, relative to other species
Yes, but to what extend is it truly blank? Could some of the problem solving abilities be more or less hardwired?
France iNcog
Ninja
Posts: 13236
Joined: Mar 7, 2015

Re: Biological Intelligence and Learning

Post by iNcog »

RefluxSemantic wrote: ↑
04 May 2023, 15:53
Are we really blank canvasses? Seems like lots of stuff is preprogrammed actually.
I mean, if you think about it, people are highly attached to the culture in which they were brought up. We're preprogrammed for a lot of the basic stuff like food, sex, etc. but I think a good part of who a person is does just come down to their environment. I think that my parents' upbringing probably had the biggest influence on me today.

To add to this, it's quite odd for me especially because I have appreciate for the cultures of both the USA and France. There is an overlap between many of these values but there is not a strong overlap in the identity both cultures have. I've noticed, in particular, that my way of thinking is not shared by many people I have met irl. There is a very American way of thinking and a very French way of thinking. I've gone and cherry picked what I like the best from both cultures and as a result, I don't belong anywhere.

Sort of an off topic tangent, but it's clearly noticeable to me and I genuinely think it has to do with being between both countries. More importantly, it has to do with me interacting with many people whom I respect from both cultures. Doubtlessly I would think differently if I had spent time in an Asian or African country for some time.

I feel these differences in both values but also pragmatic problem solving. An approach from France seems to heavily favor rigorously defining the problem (you can't solve a physics problem without first saying "this problem is supposed to be a Galilean system") whereas Americans seem to enjoy getting straight to the point as fast as possible (ignoring some of the underlying aspects of the problem).




Sorry, I rambled. What is my point? I guess that I've noticed environment to play a key role in the development of values and pragmatism. I completely agree with you, we have a lot of preprogrammed features but our brains do seem to be particularly plastic, especially in the early years. I wish that I could provide a more rigorous analysis or study rather than my own conjectures based off of personal experience. Data analysts be distraught at me atm.
YouTube: https://www.youtube.com/incog_aoe
Garja wrote: ↑
20 Mar 2020, 21:46
I just hope DE is not going to implement all of the EP changes. Right now it is a big clusterfuck.
France iNcog
Ninja
Posts: 13236
Joined: Mar 7, 2015

Re: Biological Intelligence and Learning

Post by iNcog »

Also, China is currently churning out PhD level scientists at a rate faster than any other country. The academic ability of young Chinese student seems to be the best in the world. Say what you want about the CCP but they've figured out how to mass produce highly capable scientists and engineers. I would probably point towards their educational system but alas I am woefully unacquainted with it.
YouTube: https://www.youtube.com/incog_aoe
Garja wrote: ↑
20 Mar 2020, 21:46
I just hope DE is not going to implement all of the EP changes. Right now it is a big clusterfuck.
User avatar
Greece BrookG
Retired Contributor
Posts: 2009
Joined: Feb 21, 2016
ESO: BrookG
Location: Thessaloniki

Re: Biological Intelligence and Learning

Post by BrookG »

All that destills to the nature vs nurture debate. We aren't blank canvases, there is inherent potential inside us, but still a proper environment may be needed to unlock it.
RefluxSemantic wrote: ↑
04 May 2023, 15:53
Are we really blank canvasses? Seems like lots of stuff is preprogrammed actually.
Funnily enough, my department in Amsterdam leaned on the nature dimension too.
Correlation doesn't mean causation.
http://www.tylervigen.com/spurious-correlations

"mr.brookg go buy jeans and goto the club with somppuli" - Princeofkabul, July 2018
User avatar
Netherlands Goodspeed
Retired Contributor
Posts: 13002
Joined: Feb 27, 2015

Re: Biological Intelligence and Learning

Post by Goodspeed »

RefluxSemantic wrote: ↑
04 May 2023, 18:08
Goodspeed wrote: ↑
04 May 2023, 17:00
The blankest, relative to other species
Yes, but to what extend is it truly blank? Could some of the problem solving abilities be more or less hardwired?
We have instincts obviously so it's not truly blank. I'm not sure problem solving is in there though. Our instincts have to do with basic needs like survival, food, reproduction etc. Skills are learned. If you put a kid in a dark room for 10 years after birth (assuming it doesn't die of hunger), it won't have a clue how to do anything much less solve problems. It will just know it's hungry.

We do have genetic predispositions too, but they are often overstated imo. For one, I believe anyone could be a top 50 chess player. See: The Polgar experiment.
László Polgár (born 11 May 1946) is a Hungarian chess teacher and educational psychologist. He is the father of the famous Polgár sisters: Zsuzsa, Zsófia, and Judit, whom he raised to be chess prodigies, with Judit and Zsuzsa becoming the best and second-best female chess players in the world, respectively. Judit is widely considered to be the greatest female chess player ever as she is the only woman to have been ranked in the top 10 worldwide, while Zsuzsa became the Women's World Chess Champion.
And sure, they had good genes too probably, but we like to think top players have a lot of talent and the likelihood of 2 once in a lifetime talents being born in the same family are astronomically low. We like to think of talent as mostly nature, but I'd argue that, too, is mostly nurture.
No other woman ever came close to Judit Polgar's rating. More on that here viewtopic.php?f=315&t=23091
France iNcog
Ninja
Posts: 13236
Joined: Mar 7, 2015

Re: Biological Intelligence and Learning

Post by iNcog »

The Polgar experiment was one of the thread's inspirations. Discussing the merits and methods of this experiment is something I wanted to do but I didn't have a name for it. I just heard about it through the grapevine.
I believe that the environment in 200 years will yield a better understanding of learning itself. Today's practice of throwing 30 kids in front an underpaid teacher with a whiteboard is probably not the actual most efficient way to give a young mind all the tools it needs when it comes to learning things and knowing things.
The question is, what is the answer? Are you supposed to cram them with as much information as possible before they turn 5?
YouTube: https://www.youtube.com/incog_aoe
Garja wrote: ↑
20 Mar 2020, 21:46
I just hope DE is not going to implement all of the EP changes. Right now it is a big clusterfuck.
User avatar
Netherlands Goodspeed
Retired Contributor
Posts: 13002
Joined: Feb 27, 2015

Re: Biological Intelligence and Learning

Post by Goodspeed »

It's a very telling story and I keep coming back to it in nature vs nurture discussions.
User avatar
Netherlands Goodspeed
Retired Contributor
Posts: 13002
Joined: Feb 27, 2015

Re: Biological Intelligence and Learning

Post by Goodspeed »

iNcog wrote: ↑
04 May 2023, 21:57
I believe that the environment in 200 years will yield a better understanding of learning itself. Today's practice of throwing 30 kids in front an underpaid teacher with a whiteboard is probably not the actual most efficient way to give a young mind all the tools it needs when it comes to learning things and knowing things.
The question is, what is the answer? Are you supposed to cram them with as much information as possible before they turn 5?
30 kids in front of underpaid teachers is obviously not ideal and everyone knows it, but it's just the reality given how little attention education gets in politics. It's apparently a boring subject that voters don't care about or something. Which is weird because it's one of if not the most important thing for the government to get right. Educate your population well and the rest comes easy.

AI teachers are the (near) future. 1 on 1 teaching is significantly more effective and that will be possible soon with AI, I think.

As for cramming everything before they turn 5, no need. Humans are very good at learning all the way up to early twenties. Just increase the level at a pace they can keep up with.
France iNcog
Ninja
Posts: 13236
Joined: Mar 7, 2015

Re: Biological Intelligence and Learning

Post by iNcog »

AI teaching humans more effectively is possibly the thing about AI that I am the most excited about.

I've actually gone down rabbit holes with chatgpt about different topics to learn about stuff. I wouldn't have it teach me c++ yet, but I think it would be able to teach a simpler language for example.
YouTube: https://www.youtube.com/incog_aoe
Garja wrote: ↑
20 Mar 2020, 21:46
I just hope DE is not going to implement all of the EP changes. Right now it is a big clusterfuck.
User avatar
Netherlands Goodspeed
Retired Contributor
Posts: 13002
Joined: Feb 27, 2015

Re: Biological Intelligence and Learning

Post by Goodspeed »

You should switch to Bing chat, they use GPT-4
France iNcog
Ninja
Posts: 13236
Joined: Mar 7, 2015

Re: Biological Intelligence and Learning

Post by iNcog »

Okay I am afraid to admit this but I can't keep up with the different GPTs and where they are available
YouTube: https://www.youtube.com/incog_aoe
Garja wrote: ↑
20 Mar 2020, 21:46
I just hope DE is not going to implement all of the EP changes. Right now it is a big clusterfuck.
User avatar
Netherlands Goodspeed
Retired Contributor
Posts: 13002
Joined: Feb 27, 2015

Re: Biological Intelligence and Learning

Post by Goodspeed »

It was released pretty recently. GPT-5 will be a while probably.

Google is losing this shit so hard. MS' partnership with OpenAI was bigbrain af.
Rainbow Land callentournies
Howdah
Posts: 1676
Joined: May 6, 2021
ESO: esuck

Re: Biological Intelligence and Learning

Post by callentournies »

an AI can't uphold the contract of wills between student and teacher, because it has none. Learning is more than being exposed to the content, there has to be a genuine mutual relationship imo.

Current large language models are wrong all the time and shouldnt be seen as a template for teaching. To teach, the thing has to think critically.
If I were a petal
And plucked, or moth, plucked
From flowers or pollen froth
To wither on a young child’s
Display. Fetch
Me a ribbon, they, all dead
Things scream.
User avatar
Greece BrookG
Retired Contributor
Posts: 2009
Joined: Feb 21, 2016
ESO: BrookG
Location: Thessaloniki

Re: Biological Intelligence and Learning

Post by BrookG »

On the matter of AI teaching, I'd say it's more complicated. A child is practicing not only linguistic or maths skills, but also social and behavioural skills among others. There much more into it. And what about neurodivergent children or learning disabilities? These things are still not completely figured out and I would argue it would be to absolute to have some AI teach.
Correlation doesn't mean causation.
http://www.tylervigen.com/spurious-correlations

"mr.brookg go buy jeans and goto the club with somppuli" - Princeofkabul, July 2018
User avatar
Netherlands Goodspeed
Retired Contributor
Posts: 13002
Joined: Feb 27, 2015

Re: Biological Intelligence and Learning

Post by Goodspeed »

I don't think AI will replace teachers. There will probably still be classrooms, at least at first. But the curriculum will be fully digital and AI can help students individually. Teachers will be there mostly to make sure students stay focused
callentournies wrote: ↑
05 May 2023, 00:07
Current large language models are wrong all the time and shouldnt be seen as a template for teaching. To teach, the thing has to think critically.
So you think we are far from that, I take it? How far?
No Flag RefluxSemantic
Gendarme
Posts: 5996
Joined: Jun 4, 2019

Re: Biological Intelligence and Learning

Post by RefluxSemantic »

BrookG wrote: ↑
04 May 2023, 19:29
All that destills to the nature vs nurture debate. We aren't blank canvases, there is inherent potential inside us, but still a proper environment may be needed to unlock it.
RefluxSemantic wrote: ↑
04 May 2023, 15:53
Are we really blank canvasses? Seems like lots of stuff is preprogrammed actually.
Funnily enough, my department in Amsterdam leaned on the nature dimension too.
The real question here is, to what extend are functions of our brain preprogrammed and hardwired?

I can give many examples of brain functions that are clearly not emergent from the intrinsic complexity. With animals, many complex behaviour are identical and hardwired. We call it instinct. Are we really so different that we dont have this? I dont think so. I recall that there is serious reason to believe that personality is partially inheritable.

You might argue that in some key areas of human intelligence, it is obvious that it is not instinct driven. But I dont think that it is actually so obvious. The most interesting example is abstract thinking, which children can only do starting at around 12-14. I was teaching physics, chemistry and mathematics to children of this age group last year and it is uncanny how absent this ability can be. Some of them literally can not grasp it. This kind of thinking is crucial for serious problem solving, and it seems biologically hardwired when it starts developping.

Actually, more anecdotally and moving more towards nature vs nurture, Ive noticed that my brain operates in a very similar way to my parents and in particular my sister. You may claim that this is because we grew up in the same environment, but I dont think its that simple. If we think the functioning of a brain is an emergent property, then surely the path in education (where we are actually training our brain) should have a far bigger impact. So I should think in a similar way to a childhood friend, who has had the same teachers and classes as I did from ages 4 to 18. Yet we think very differently, and actually typically I would get the questions right that he would get wrong and vice versa.

So is it really so simple? I dont think so. I actually lean towards believing that many critical functions of the brain are more or less preprogrammed. From my perspective, the actual intelligent problem solving is not about pattern recognition (which would be inductive reasoning), but about deductive reasoning. Just to add another anecdote: I have managed to win tournament games in Aoe3 through strats that I deduced while on vacation. Thats not pattern recognition, that is logical reasoning. And the way we reason logically, may very well be (partially) hardcoded into our dna.
User avatar
Netherlands Goodspeed
Retired Contributor
Posts: 13002
Joined: Feb 27, 2015

Re: Biological Intelligence and Learning

Post by Goodspeed »

RefluxSemantic wrote: ↑
05 May 2023, 07:18
BrookG wrote: ↑
04 May 2023, 19:29
All that destills to the nature vs nurture debate. We aren't blank canvases, there is inherent potential inside us, but still a proper environment may be needed to unlock it.
RefluxSemantic wrote: ↑
04 May 2023, 15:53
Are we really blank canvasses? Seems like lots of stuff is preprogrammed actually.
Funnily enough, my department in Amsterdam leaned on the nature dimension too.
The real question here is, to what extend are functions of our brain preprogrammed and hardwired?

I can give many examples of brain functions that are clearly not emergent from the intrinsic complexity. With animals, many complex behaviour are identical and hardwired. We call it instinct. Are we really so different that we dont have this? I dont think so. I recall that there is serious reason to believe that personality is partially inheritable.
We do have instincts but also yes, we are really so different as to be significantly more malleable than other species. As for personality, most of that is nature I think, but nurture can still change it in extreme cases. For example the spectrum of intro/extroversion is considered nature, but if you put a kid in a dark room for 10 years after birth it's probably not going to come out an extrovert.
You might argue that in some key areas of human intelligence, it is obvious that it is not instinct driven. But I dont think that it is actually so obvious. The most interesting example is abstract thinking, which children can only do starting at around 12-14. I was teaching physics, chemistry and mathematics to children of this age group last year and it is uncanny how absent this ability can be. Some of them literally can not grasp it. This kind of thinking is crucial for serious problem solving, and it seems biologically hardwired when it starts developping.
Abstract thinking is a learned skill. Do you think the first homo sapiens were good at abstract thinking? It's a skill that's really not needed at all in a hunter/gatherer society, and therefore it would make little sense for it to have evolved. The difference you noticed was likely just between kids who trained abstract thinking (e.g. by playing games like chess) and the ones who didn't.
Actually, more anecdotally and moving more towards nature vs nurture, Ive noticed that my brain operates in a very similar way to my parents and in particular my sister. You may claim that this is because we grew up in the same environment, but I dont think its that simple. If we think the functioning of a brain is an emergent property, then surely the path in education (where we are actually training our brain) should have a far bigger impact. So I should think in a similar way to a childhood friend, who has had the same teachers and classes as I did from ages 4 to 18. Yet we think very differently, and actually typically I would get the questions right that he would get wrong and vice versa.
Your direct family has a much greater impact than school. That your family members' brains operate similarly sounds relatable. I think it has more to do with how much you've interacted with each other, thereby developing a sort of "meta" on how to approach problems. Something not to underestimate is how genetically different you actually are from your parents. Especially when it comes to the (very complex) brain, these differences are big enough that there is, afaik, very little correlation between a person's parents' personalities and theirs. You're supposed to be a bit more similar to your sister, but even sibling personalities aren't strongly correlated.

There have been identical twin studies, which conceptually would seem like the way to answer the nature vs nurture questions, but they are often flawed in methodology. Ideally you'd want to separate a bunch of identical twins at birth but that's not realistic.

Something that needs clarified I think is the meaning of "emergent property". I don't think it means that through sheer complexity of the brain we can do a task at birth. We still need to learn it. If we say a certain capability "emerged" at a specific complexity, we mean our brain is capable of learning the skill. It has enough neurons to be able to learn it. It doesn't mean we are born with the ability to do it.

In the context of AI, GPT-4 has theory of mind of a 9 year old or something. That turned out to be an emergent property of the GPT model at a certain number of parameters, but in order to do it the AI still had to train on an insane amount of data. It couldn't to it without learning.
No Flag RefluxSemantic
Gendarme
Posts: 5996
Joined: Jun 4, 2019

Re: Biological Intelligence and Learning

Post by RefluxSemantic »

The thing is that it is likely that some abilities of the brain are in fact not learned. They may develop after we are born, but that does not mean that it is learned.

Who is online

Users browsing this forum: No registered users and 10 guests

Which top 10 players do you wish to see listed?

All-time

Active last two weeks

Active last month

Supremacy

Treaty

Official

ESOC Patch

Treaty Patch

1v1 Elo

2v2 Elo

3v3 Elo

Power Rating

Which streams do you wish to see listed?

Twitch

Age of Empires III

Age of Empires IV