- My Forums
- Tiger Rant
- LSU Recruiting
- SEC Rant
- Saints Talk
- Pelicans Talk
- More Sports Board
- Winter Olympics
- Fantasy Sports
- Golf Board
- Soccer Board
- O-T Lounge
- Tech Board
- Home/Garden Board
- Outdoor Board
- Health/Fitness Board
- Movie/TV Board
- Book Board
- Music Board
- Political Talk
- Money Talk
- Fark Board
- Gaming Board
- Travel Board
- Food/Drink Board
- Ticket Exchange
- TD Help Board
Customize My Forums- View All Forums
- Show Left Links
- Topic Sort Options
- Trending Topics
- Recent Topics
- Active Topics
Started By
Message
re: Some thoughts on IQ
Posted on 8/3/25 at 10:22 am to 4cubbies
Posted on 8/3/25 at 10:22 am to 4cubbies
I actually read this very differently from you.
He's talking about how it's not yet proven that one can change one's potential (again, this is why I phrased it as the ceiling).
Aka if all children benefited from things that we know increase IQ, would the order not remain the same because their genetics are different.
I agree that part has not been proven, what has been proven are only better and more successful behaviors, easier processing of information.
It seems he recently published a book, reading through the ebook.
Interesting way he sees the interplay.
His final conclusion.
Again, I think his work is misunderstood by most people, much like the Bell Curve.
If we give the same advantages to everyone, aka nurture goes to 100% will some people come out on top?
Will those be the people with the best genetic advantages?
Well if everyone gets brain damaged will the order remain?
no in that case the order probably shifts to new genes that are better at redundant processing.
There are likely genes that come into play only at higher levels of education and nutrition.
But overall, it's a rhetorical question.
If the bottom person can operate significantly well mentally, does it matter?
The Bell Curve and Haier's book (He's a fan of the Bell Curve), both list a number of things society can do to bring up everyone's societal outcomes and skills.
They would not insist that a PhD is beyond people without genetic gifts, except in extreme cases.
Based on the Flynn effect alone, reducing negative outcomes will raise the bottom of what people can do.
Again, 70 is a huge outlier that likely spawned unnaturally.
He's talking about how it's not yet proven that one can change one's potential (again, this is why I phrased it as the ceiling).
Aka if all children benefited from things that we know increase IQ, would the order not remain the same because their genetics are different.
I agree that part has not been proven, what has been proven are only better and more successful behaviors, easier processing of information.
It seems he recently published a book, reading through the ebook.
Interesting way he sees the interplay.
His final conclusion.
quote:
Understanding the cascade of events
from DNA to genes to proteins to neurons/
synapses to brain development to brain cir-
cuits to brain functions to individual differ-
ences in reasoning ability (notwithstanding
many steps and processes in between these,
along with environmental interactions) may
be the most formidable challenge in science,
and one with the most profound conse-
quences. We think it will be possible to
enhance intelligence by targeting the g-fac-
tor, although for now, we do not know how
to do so. It is a goal that should be pursued
with vigor. To paraphrase President John
F. Kennedy in a speech about going to the
moon, given at Rice University on
September 12 1962, we choose this goal
“not because [it] is easy but because [it] is
hard, because that goal will serve to organize
and measure the best of our energies and
skills, because that challenge is one that we
are willing to accept, one we are unwilling to
postpone, and one which we intend to win.”
Perhaps as you finish reading this book,
you might consider this goal of intelligence
research in a new light.
Again, I think his work is misunderstood by most people, much like the Bell Curve.
If we give the same advantages to everyone, aka nurture goes to 100% will some people come out on top?
Will those be the people with the best genetic advantages?
Well if everyone gets brain damaged will the order remain?
no in that case the order probably shifts to new genes that are better at redundant processing.
There are likely genes that come into play only at higher levels of education and nutrition.
But overall, it's a rhetorical question.
If the bottom person can operate significantly well mentally, does it matter?
The Bell Curve and Haier's book (He's a fan of the Bell Curve), both list a number of things society can do to bring up everyone's societal outcomes and skills.
They would not insist that a PhD is beyond people without genetic gifts, except in extreme cases.
Based on the Flynn effect alone, reducing negative outcomes will raise the bottom of what people can do.
Again, 70 is a huge outlier that likely spawned unnaturally.
0
Posted on 8/3/25 at 10:26 am to dukkbill
quote:
The latter is where it can get ugly and intersect with ego, tribalism, and eugenics. We don’t have a single scalar that globally measures the entirety of someone’s value. People will try to game that to make themselves or their group appear to be ubermensch
I'm not sure I could more fully agree than with this statement.
Posted on 8/3/25 at 11:28 am to 4cubbies
quote:
We revisited Stankov’s report on the outcomes of an experiment carried out by R. Kvashchev in former Yugoslavia
This paper personifies the adage that figures don't lie, but liars figure.
The authors took an old study, which itself was criticised for reasons I'll touch on. Then they simply "reimagined the results" through statistical manipulation.
The original study was critiqued for practice effects.
Meaning, IQ testing is designed as a one-off.
If I give you an IQ test, then show you how to navigate the type questions you missed, when I administer a second valid but similar test to you, your score will artificially improve.
E.g., Question: "If all zoodles are zogs and all zogs are babes, are all zoodles definitely babes?"
It's an elementary question requiring basic syllogistic reasoning, but it nonetheless stumps some test takers. If I teach someone how to work through it, their ability to reason hasn't changed, but their ability to answer the question has. Essentially Kvashchev did some of that in his original work.
This post was edited on 8/3/25 at 11:29 am
Posted on 8/3/25 at 11:34 am to Narax
quote:
Which AI did you use?
Does that look like an AI answer? Seriously?
Posted on 8/3/25 at 11:40 am to NC_Tigah
quote:
Does that look like an AI answer? Seriously?
Yes, it is exactly an AI answer, It completely missed what even an easy human google search and couple minutes could have turned up, it was obviously responding to a prompt, and it obviously suffered from the prompt lacking context.
This is the exact template of an LLM.
Posted on 8/3/25 at 11:48 am to Narax
quote:
Yes, it is exactly an AI answer,
I don't know whether I should feel insulted or flattered.
quote:The prompt was the content of your recent posts. As for any "suffering" in the response, you either misread or otherwise did not understand what was written.
it was obviously responding to a prompt ... suffered from the prompt lacking context.
Posted on 8/3/25 at 12:02 pm to NC_Tigah
quote:
I don't know whether I should feel insulted or flattered.
Neither, it's just true.
quote:
The prompt was the content of your recent posts. As for any "suffering" in the response, you either misread or otherwise did not understand what was written.
My apologies that I did not write it for an AI prompt, I thought I was writing for humans who would be curious about synaptic pruning,
It seems you fed it directly in, and it called a scientific fact a hypothesis as the LLMs are apt to do when dealing with a badly designed prompt.
It's part of the underlying structure, it won't flat out call someone wrong, as it knows the human on the far end might not know enough to prompt it well.
So it repeats back to develop it's patterns.
A better prompt would have been explain synaptic pruning to me.
But lets lay it out
quote:
The hypothesis is one of weaning less-used transmission pathways while strengthening important ones in a neural network remodeling as we age. Weaning/remodeling (misleadingly referenced as "pruning") results in cleaner, more refined pathways.
The standard LLM response, this is when it starts to repeat the problem back to you, it does this for 2 reasons, it provides to the user feedback on what it thinks the question is, and it starts it's pattern of cosine similarities, if this isn't right, the rest is going to be likely to be wrong.
quote:
At a general point in which the brain initially reaches peak functional IQ, we surmise less-used residual pathways still exist, albeit at lower levels. Theoretically the presence of these pathways accounts for associated creativity of youth. As remodeling continues toward maturity though, extraneous circuits continue to gradually dissipate. Though IQ remains, creativity might wane. Postmaturity, the same continuing process can subsequently lead to a slight downward IQ variance.
This is a collage of a number of studies, none of which you have read, it also refers to a brain reaching peak functional IQ, which you have basically rejected.
I don't know why you want to deny it, most people nowadays are using AI to increase the amount of information available to them.
But in this case you badly prompted and drew wrong conclusions.
All of humanity is getting used to using AI, we need to get used to it sooner rather than later.
This post was edited on 8/3/25 at 12:05 pm
Posted on 8/3/25 at 12:17 pm to Narax
quote:
Narax
If you feel comfortable sharing, what is your line of work? Every time AI is brought up, you impress me with your knowledge (not that I’m particularly hard to impress or anything but you know a lot about the mechanisms of LLMs)
Posted on 8/3/25 at 12:18 pm to Narax
quote:Dude, you're coming across as more of a nitwit than you understand.
A better prompt would have been explain synaptic pruning to me.
Pruning is a term I personally dislike because it implies sudden and permanent pathway cessation. That is not what occurs physiologically ... at all. I have no clue WTF AI would say about that as it's simply my medical opinion regarding the "pruning" metaphor.
quote:It just gets better. You'd do far better sticking with what you know, rather than making assumptions as to what I know or have read.
This is a collage of a number of studies, none of which you have read
Posted on 8/3/25 at 12:28 pm to 4cubbies
quote:His posts here notwithstanding?
but you know a lot about the mechanisms of LLMs
Posted on 8/3/25 at 12:28 pm to dukkbill
quote:
I don’t think that excerpt will give you much insight on the debate about malleability of persistent intellectual processing capability. These findings more relate to two other phenomena:
(1) Precision of measurement; and
(2) impact of training on certain measurement tools
True. As someone who hasn’t deeply thought about IQ testing until this thread, I hadn’t considered the limitations brought up in those excerpts.
quote:
Nevertheless, I’m not sure it reads on meaningful insight into the debate of heritability or the nature of intelligence itself. The ability to obtain a short term result is much different than longitudinal
measures of attainment. Indeed, the former, at best, is a means for short term decision making. At worst it’s a pissing contest.
Agreed. It reinforces how silly comparing IQ scores between different individuals is. It also calls into question the validity of the annual standardized testing to assess the academic gains of public school students starting in third grade.
Posted on 8/3/25 at 12:30 pm to 4cubbies
quote:BINGO !
It reinforces how silly comparing IQ scores between different individuals is.
Posted on 8/3/25 at 12:30 pm to NC_Tigah
quote:
His posts here notwithstanding?
I’m not accusing you of using AI fwiw. I honestly don’t care either way tbh. But in a thread a while ago, he knew a lot LLMs. It was a thread that suggested AI is sentient.
This post was edited on 8/3/25 at 12:31 pm
Posted on 8/3/25 at 12:30 pm to 4cubbies
quote:
If you feel comfortable sharing, what is your line of work? Every time AI is brought up, you impress me with your knowledge (not that I’m particularly hard to impress or anything but you know a lot about the mechanisms of LLMs)
I've mentioned it before, but more or less I'm a software engineer and since 2019 AI/ML has taken over my field.
Most things I do are at the very least AI adjacent nowadays.
Things like Gitlab duo while not yet able to make anyone into a software developer are massive improvements in productivity.
Plus much of what I work on uses ML/AI in it's function for more complex tasking.
This all requires me to keep up with the latest research and findings in transformers, assuming I want to remain employed.
Posted on 8/3/25 at 12:31 pm to 4cubbies
quote:That is a very different issue.
It also calls into question the validity of the annual standardized testing to assess the academic gains of public school students starting in third grade.
Posted on 8/3/25 at 12:35 pm to NC_Tigah
quote:
That is a very different issue.
The meaningfulness (not the best word choice) of the tests isn’t, though.
Posted on 8/3/25 at 12:36 pm to 4cubbies
quote:Nor do I.
I honestly don’t care either way tbh
But it's no more something I'd need or likely find use for on this topic than you would in your field of interest.
Posted on 8/3/25 at 12:43 pm to NC_Tigah
quote:
Dude, you're coming across as more of a nitwit than you understand.
Pruning is a term I personally dislike because it implies sudden and permanent pathway cessation. That is not what occurs physiologically ... at all. I have no clue WTF AI would say about that as it's simply my medical opinion regarding the "pruning" metaphor.
So you just think like an NPC?
It's my opinion as a developer that this is exactly how an LLM responds.
Do you believe in a peak functional IQ?
How does that happen?
quote:
It just gets better. You'd do far better sticking with what you know, rather than making assumptions as to what I know or have read.
So do you want to put actual claims about your medical background?
Are you claiming to be a specialist in brain development?
Posted on 8/3/25 at 12:44 pm to 4cubbies
quote:In performance improvement there must be a feedback loop. What is the basis for comparative feedback if not standardized testing?
The meaningfulness (not the best word choice) of the tests isn’t, though.
Posted on 8/3/25 at 12:46 pm to 4cubbies
quote:
Agreed. It reinforces how silly comparing IQ scores between different individuals is. It also calls into question the validity of the annual standardized testing to assess the academic gains of public school students starting in third grade.
Here is where we diverge. I agree IQ as a dick measuring test is limited. I also agree that the unintended effect of the test implementations have been “teaching the test”. Nevertheless, we still need: (a) measured to gauge assurance of learning; and (b) measures that help us decide how to allocate our resources. As blunt an instrument as these tests may be, we don’t have an implementable alternative that is feasible and being proposed. Accreditation “could” be a better method, and longnitudal attainment could be a method. Nevertheless, this pushes us more toward school choice and more toward accreditation. The former is routinely denounced by similar minded people as yourself, and the latter is often derided by others that bemoan “indoctrination”. For now, it appears to be our only means of getting these metrics
Page 29 of 33
Popular
Back to top
Follow TigerDroppings for LSU Football News