Hacker News Reader

Comment by corruption | Hacker News Reader

Minute differences in genomes lead to large differences in phenotype. Ever heard of databases of SNP's or CNVs? A single base change can make a world of difference.

Actually, I'm not basing this on anecdote, as there is a ton of literature to back me up here. The first thing I searched for is quite remarkable for a single generation: http://www.springerlink.com/content/r44140388u23t768/

But, without going to the extensive literature, are you seriously claiming that the rate of learning is EXACTLY the same for every human, or that the variance is so tiny that differences are negligible? Both claims are provably false if you have a quick look at the literature.

View on HN

>Both claims are provably false if you have a quick look at the literature.

Which literature? You're doing a bit of a (probably unintentional) bait and switch here if you mean the article you submitted. The article talks about learning rates of rats. Rats don't have language. How does having language change the picture [1]? We can't answer that, and therefor we can't know how applicable information about rat learning rates is to humans.

[1] Language certainly makes us more intelligent but there's also been studies that suggest that language may also dampen certain inherent abilities. In other words, maybe some people are born with drastically more "powerful" brains but language removes their advantages effectively evening out the field.

There is way too much we don't know about all this stuff for anyone to have such firm opinions on the mater. Certainly such limiting ones of "well, you're just not smart enough to ever be able to do this".

We are both making a claim. He has provided no evidence at all, and obviously hasn't read the literature.

The claim here is that the variance in learning rate is absolutely tiny. Think about this. He's basically claiming that despite a large known population variances in IQ and MA, there is no variance in learning rate. Despite research showing ritalin increases learning rate in a large fraction of society, this increase was meaningless. Despite verbal learning test normative data showing non zero variance, the variance is still zero.

There is a large body of literature comparing learning rate in different populations in both human and animal models. Take any study on learning rate, even the ones comparing "normal" to abnormal. Go to the methods and find the variance for each group. See it's non-zero, but often so large that differences between groups is hard to see.

Now, if you accept that learning rate has a non-zero variance, and also that learning rate decreases over time (well supported in the literature as well, and no I'm not going to find papers) and that time is finite, then the only conclusion you can make is that some people have a much higher chance of doing well academically. Not only that but there will be some people who despite trying will never be able to do well.

I don't see how stating this shocking to anyone at all - in a population of some 7 billion, there are millions of people >3 standard deviations from the mean.

>He's basically claiming that despite a large known population variances in IQ and MA, there is no variance in learning rate.

If this is what he's claiming it's obviously silly. A more plausible claim would be that the rate at which one could learn may be close to equal provided an optimum teaching method is used on a case by case basis.

What I mean here is: I think all the literature you're referencing about learning rates doesn't conclusively prove that it's the students at fault. It could be the teaching method. I've experienced first hand that some people who seemed to not "get it" for some subject could in fact understand it quite well when it was presented in a way they understood (perhaps they didn't see the relevance, perhaps they didn't have the proper frame of reference to understand the normal explanation or a poor grasp of those references).

For me it's just too simple to say "everything's fine, those people are just too stupid. Bad genetics". What if it's not? Then we're throwing away a lot of potential unnecessarily. And part of the reason I suspect this to be the case (to some unknown degree) is because plenty of teaching methods that are still in use are not even the best known method (e.g. how most schools teach a second language).

It is clear that there are genetic differences in brains that cause large differences in "learning rate" - for lack of a better term. Interpret this how you want, but it's obvious to those in the field what this means (I talked to a researcher today to check my assumptions here to make sure!).

It is what it is. If my optimised (via teaching, drugs, augmentation etc) learning rate is lower than your's then so be it. So what? Why do you feel the need to make us equals, I accept your superiority in learning rate and do not care one iota.

I don't feel the need for us to be equal. If feel that the idea that we're not is often a cop out to make ourselves feel better about people who could have been our equal but aren't.

That is to say, even if you're right and there is a massive difference that no technique could ever overcome, there are a lot of people who could have done better than they actually did. Maybe their environment destroyed their chances. Maybe they had bad teachers. Who knows. I'm just not comfortable writing them off as "stupid". Education is too important to ever assume we have the best system possible. No matter how good it is.

Isn't there a contradiction here:

"I don't feel the need for us to be equal. If feel that the idea that we're not is often a cop out..."

If the evidence supports the conclusion, we must accept it regardless of our personal biases. My general claim is that each person has a range of capabilities that they are born with, and that this range is ~ normally distributed across the population. By definition these ranges do not overlap for all n. This is supported by both anecdotal, population data and experimental animal and human models to the best of my knowledge.

As for your second paragraph, well as you can see I never claimed to the contrary. I completely agree that one should find a method of learning to maximise their individual potential, and that the range of what is possible for an individual is quite large - see my claim above. In fact I am very close to someone working on this very topic - determining the optimal learning strategy for children to maximise their learning rate.

>Isn't there a contradiction here:

I don't see it as a contradiction. I don't feel that we all must be equal (as in: exactly as good at everything as each other). But I feel that deciding "hey, we're not all equal" can be used as a cop out, even if true. That is, if you strive for the possibly impossible goal of making every student a PhD student whatever it takes you probably wont achieve it but you stand a much better chance of making a breakthrough with your under achievers than you would if you just said "60% of the kids in this class have no chance, so why bother".

>If the evidence supports the conclusion, we must accept it regardless of our personal biases.

We must give it credit, but we can still believe that there is more to the story until it can be proven that there isn't.

>In fact I am very close to someone working on this very topic - determining the optimal learning strategy for children to maximize their learning rate.

Awesome. I wish you all success in this most important of endeavors. Don't take my comments personal. I don't know you so I unless you explicitly state your position I tend to address what I expect most people's to be.

We'll agree to disagree then, both on the structure of the minds of people who make true breakthroughs and your views on epistemology :)

I don't take anything personally at all, and hope you didn't either :)

I think the main point of our disagreement is around rationalism. You seem to be presenting the point as "whatever the evidence shows now must be accepted". It is my belief that most everything we "know" today is incomplete or even wrong in some fundamental way. From that point of view, if what I think is your view is being "rational" then only the irrational can ever truly further our knowledge.

Now of course when making actual decisions I would go with what the evidence shows today. That's the safest bet. And for breakthroughs, of course I would focus on the "brightest" since that is likely to have a much larger impact [1]. I just hope someone isn't excepting status quot.

As far as my views on the nature of knowledge: my view is that we don't know know enough to even speculate. We're learning more but it's slow going. Language seems to have a big effect on knowledge so personally I wish more research were directed in this area [2]. I think we would learn more faster than we do with current methods.

[1] That is, if I invest a tremendous amount of work and can only get every student to "average" then I could have potentially gained more by pushing the "brightest" students even further. Then maybe they could come up with the breakthrough. :)

[2] It is a hard question due to ethical considerations. The simplest test method is unthinkable as it would mean destroying the lives of some number of people to test the effects of lack of language. Of course it was recently pointed out in an article that appeared here that such methods aren't needed. Any time a child is born deaf to hearing parents there is a strong chance they will grow up without language. But this brings another ethical question: instead of studying them after they've grown up shouldn't we be preventing them from growing up without language? But at the very least we could find as many adults as possible who are already in this state.

I do agree that some people may have higher learning rates than others but the differences are too small to matter. I will even agree that there may be outliers, a super tiny amount of the population (although I've never met one, and some that are considered true geniuses have some serious mental deficiencies). Overall the differences when it comes to the general population are too small to be that much significant. The people getting phd degrees have in general the same degree of potential learning ability as the general population.

If there were really that big of a difference between their potential ability to learn as opposed to the general population than they would be passing this down to their kids. Since you claim that the difference is so large than with careful breeding we should be able to produce a super race of geniuses within a couple of generations. That is not how evolution works.

Improvements are infinitesimally small and I don't see why would intelligence be any different. We all evolve as a species, not as individuals. I do not get the sense that we have been getting that much smarter based on our overall written history of the past thousand years.

The reason why some choose to pursue a phd and some don't is the same as why some choose to pursue a career as a painter, musician, writer and why some just choose to be a blue collar worker. It is what they feel passionate about. Some don't feel passionate about anything.

Culture and environment also play a major role and this is really what makes a difference between people choosing to do a phd or not. If you or your peers tell you that you cannot get a phd and believe it then it becomes a self fulfilling prophecy.

There are differences in rates in learning but they are not large enough to matter. Culture and environment play a major role.

Edit: Not even Einstein could be called an outlier. He was a bright person but I hardly believe that his potential ability to learn was that much higher than yours or mine. He was even considered mediocre by many of his professors. It really was luck that he happened to be born at the right time in history to be able to make the contribution he did. Even he had to ask one of his mathematician friend for help when doing the mathematics for relativity. (you need to read his biography).

Newton considered a giant of physics was considered a pretty ordinary student. He didn't amaze anybody by showing a high potential for learning during andy of his school years nor did anybody imagine that he would come to make such huge contribution to physics.

Please read the literature. There is plenty of evidence that directly contrasts your viewpoint from the last 50 years.

And by the way, your second point is directly contrasted by the article I linked to.

I read a couple of paragraphs. Is one thing to do a study on rats then to do a study in humans. This doesn't disprove my claim.

Edit: Also, it seems that you are implying that the differences are so huge that we should have a lot of retarded people walking around us. The proof that the gaps in intelligence are small should be obvious by simply looking at your environment. There are people getting phd's from all corners of the world. China, Japan, USA, Mexico, Europe, Africa, Australia. How did those same genes that enable people to get phd's get to places so far away as China or Japan which have been isolated for so long? Even for thousands of years? The thing is that those genes have been there all along for thousands of years.

I think I already agree that there are differences but I do not agree that the differences are huge. Otherwise some of these groups of people, like china or japan who have been isolated for thousands of years, would have serious intelligence differences as the rest of us.

Another example is the Mayans, also isolated for thousands of years from Europe. Independently invented the zero and developed a Mayan calendar unequal on precision until recently. Many of their descendants have gone on to get PHD's. If there were really large differences then some groups of people would not even be able to get PHD's.

Again, the there are differences but they are tiny. Yes, there are studies saying that there are huge differences between groups of people but frankly many of these studies are questionable and sometimes they almost seem to carry a hidden agenda.

> Otherwise some of these groups of people, like china or japan who have been isolated for thousands of years, would have serious intelligence differences as the rest of us.

Have you read the literature on IQ? These differences exist, whether you will accept it or not is a different question.

Please don't cite evolution without fully understanding your claims. Evolution USUALLY works through tiny changes, but in certain situations the changes are dramatic and sweeping. If your proposed experiment was carried out, I would expect such a sweeping change. Further I would posit that a pre-mating barrier would form (in the form of contrasting mating rituals) and speciation would occur after sufficient time.

TLDR: Don't claim that Science supports your view if you don't really understand the theory.

Reply on news.ycombinator.com