What intelligence truly is (?)

[fluffy]

In basic terms of perception and actions we have in the middle reasoning or thinking.

So pretty much as people we need to understand things we don't know yet to get things done.

Understanding requires knowing many things at once and trying them out to see what would work.

Usually knowing more can help solve problems faster.

In a nonlinear fashion having things in place is about parallel processing. Cause and effect is not just about trial and error but there are multiple considerations at play we take account of where one part effect the others in complex ways.

Attention and memory and correction of errors in reasoning had at its core situational awareness. Looking at what is and understanding the consequences of anything leading to the next condition. In fact it is a good way to organize thought.

People with more short term memory can use it to think up more possibilities and view points that could work and then decide the best approach. Reasoning about a given task would mean we think about how the entire system would change and if that's good or bad. We would want the best outcome so pondering what would need to happen keeping in mind the mistakes that could happen and contingent responses.

The larger picture in the meta sense is about purpose of application. Why am I doing this and what good is it for. Do my actions relate to my ultimate goals that these smaller sub goals accumulate to.

It is possible to create really detailed and complex systems out of thoughts and objects. What then is intelligence is having better ways to see the interrelationships. More or less.

 

[SignalCobra]

What stood out to me is the idea that intelligence isn't just about what you know, but about how well you see the connections between things. In real life, we often look for direct causes and effects, but most situations are more like a web than a line. A small change in one part can trigger a chain of reactions somewhere completely unexpected. If you don't see the whole picture, your solution might seem logical at first but lead to bad outcomes later.

Another thing I was thinking about is that when we try to reason through something, we aren't just processing data. We're building a model of the world. And the more accurate or adaptable that model is, the better chance we have of actually making a difference. So in a way, thinking isn't just a tool to solve problems. It's a way of being more conscious and deliberate in how we live.

 

[Hadoblado]

@fluffy
I think at this point you should consider making an ultimate intelligence thread and funnel all new thoughts on that subject into it. You make a few specific topics you've returned to many times over the years. Intelligence is one of them. These posts are welcome, but when it's the same people talking about the same things it's best for clarity and for following the conversation if they occur in the one place. This feels like a singular long term project of yours, but the way you communicate it is diffuse and short-term. You keep burying both the building blocks and the replies by making new threads.

 

[fluffy]

Because of the current interface I have I cannot collect all the threads but I will keep everything to this one thread for now and not make any others @Hadoblado

-

I was thinking about how we use our intelligence in so many ways. But the core way we use it is by looping around that allows us to be intentional with our thoughts.

Like how I understand that I exist and want to show people things or make something but in order to do so I must have a sense of self and piece things together in chunks.

Understanding what I want to do depends more on my maturity than the span of my working memory because as long as I add things together I can make big things it just takes longer to memorize all the details which consumes mental energy. Dedication is more important then. Having to do with sustainable willpower and I believe willpower is a mechanical process.

Bodily awareness helps regulate willpower.

So you can focus a long time to get things done.

Self reference in a way to correct errors and looking at everything in your design to work.

Then you don't need as much working memory to be good at the goal you have because you know the interelationships by specialization.

Maybe it's like being good at art. As long as you practice you can get eyesight to see many details efficiency. It depends on the goal after all. The most effective way.

It's the input and output.

Maturity builds into the middle part. So you have skills that improve your whole life. But it still happens at a relative pace. If someone has the same goal and works hard and is as intelligent as you, you both get to the same place at the same time. That means you don't choose a goal because you're better at it but because you preferred it to anything else. Others will be better. There's always someone better but you want to do it. That's the point.

As long as growth happens in the middle you are better than 5 years ago. And I believe I understand that middle parts functioning. How the operations happen. Why growth exists. And this allows me to improve my designs. Creativity is trying new things all the time. Intelligence is just the breath at which you can create depth. Yet you can get a snow ball effect at some point. This creates endless amounts of ideas to work with.

 

[fluffy]

Mathematically it's impossible to exceed an IQ of 200 - the curve would either become unsymmetrical or you'd get negative IQ points. The curve depends on the sample you put into the gaussian formula.

IQ is relativistic meaning only one person can be at the top and not exceed the structure of the curve. The curves has nothing to do with brain function but with the position each person is next to. It's not the concentration of people by intelligence but the variance between peoples intelligence.

Today about 100,000 people have an IQ of 160 - this is the estimate for Einstein.

 

[fluffy]

Spidermans IQ is sometimes said to be 260

But any IQ above or near 200 is suspect.

Here is what you do to find a persons real IQ

Real IQ = (√(260/100))•100 = 161 for spiderman

Isaac Asimov said his IQ was 300

(√3)•100 = 173

And the same for William James Sidis

Marilyn Vos Savant's IQ of 228 would be 150

 

[fluffy]

IQ was calculated as mental age in the past but this has the same problem going above 200

Instead we can calculate mental age this way:

Mental Age = (IQ/15)•2

137 would be age 18.2 and 100 be 13.333

13.333 is the usual age of Jean Piaget's formal operations stage of development.

A person who is iq 137 reaches this stage at about age 9 and a half.

But beyond a menat age of 26 (195 IQ) we would not have any way to tell what this means. A person at age 10 would be age 20 mentally but iq 200 would be double the development speed in intellectual feats.

 

[fluffy]

I made a chart for mental age and movies I saw as a kid.

The top movie has a pseudo IQ of 201 and the bottom has a pseudo IQ of 60

I put these numbers in a 16 by 16 correlation matrix. I adjusted for the age I saw the movie and the normalized IQ to find g loadings.

201 became 189 and 60 became 94 respectfully.

These numbers were still way out there.

I multiplied the age I saw the movie by 1.37 then divide the movie age ratio by it.

After that I adjusted my age to 13.33

Example I saw The Land Before Time at age 6

Age 6•1.37 = 8.22
(pseudo IQ 163.25/15)•2 = 21.76
21.76/8.22 = 2.64

(ratio IQ √2.64)•100 = 162.7

(162.7/15)•2 = 21.69

21.69/13.33 = 162.7

162 is the IQ one needs to understand The Land Before Time at age 13.33 giving a mental age of 21

This movie was the most difficult to understand for me. At 6 years old.

 

[fluffy]

Executive Control

This could very well be the most important part of intelligence.

If you want to accomplish a goal you need to get everything into position, tracking multiple contents simultaneously.

First you remember everything you are doing then you monitor if you are making progress finally you switch between tasks.

Pattern recognition is more in the back brain and is important as well as you need to fit together the components. Yet it is tracking large numbers of them that is really difficult for people to do.

I was tested to be better than 75% of people in executive function but there are those better than 99% or even 99.9999%

Another term for it is metacognition.

Where your self monitoring is higher. You make less mistakes have greater self awareness and can pay attention to more factors involved. It is not the same as measurements on IQ tests because patterns are the main features on them. Metacognition as critical thinking has no correlation but it's still part of intelligence. The structure most associated with it is the anterior cingulate cortex (ACC) working memory as pattern recognition involves the dosal lateral prefrontal cortex and the partial lobe connection.

Furthermore one IQ subsection not based on patterns corresponds to the temporal partial junction has its own set of working memory called quantitative reasoning. This is best thought of as figuring out rules by mental manipulation (what goes where and why) because of the amount they represent. It can figure out what amount should existing be placed in relation to each other. It tells you if too little or to much exists in different places.

Depending on how the frontal lobe is wired up self monitoring just is not part of the IQ construct. A person can have high pattern recognition but low in there ability to know when they made mistakes in life. High pattern recognition can compensate in a limited way as a person can recognize correctly if a pattern exists but then they will be impulsive and not check if this particular pattern is in cohesiveness with other parts needed. They won't double or triple check. And on top of this they may not understand that back ups must be in place in case of failures. Metacognition not only looks for mistakes but levels of design structure at all angles.

High Executive function then can make plans with whatever situation exists. The more it can store in memory the better. It's careful to decide what to do if something is unknown or if anything unplanned for happens. Because it knows manny areas of the subject come into place. The brain must put it all together. If not it might not work even if one can think on ones feet plans must be thoroughly thought-out ahead of time.

 

[fluffy]

Rethinking brain mechanisms in light of evolution with Paul Cisek


The brain is not an input output decoder. It is not a Cartesian dualism with mind separate from body. This is why a.i. currently doesn't work.

The brain is a nested control loop keeping the body in homeostasis.

Dopamine was the first system to evolve. In high concentration of food you don't do anything but in scarce food you move around more. Eyes help avoid shadows (predators)

In the primate brain you can think ahead. You can weight options for delayed gratification and future orientation. A long term goal over short term benefits. Looking ahead a couple moves. Get the far away apple instead of the closer Berry.

Most systems in the brain are in-built pre planned behavior (instinctual) learning in mammals comes with somatasized touch. Digging grooming bring food to the mouth. Remembering that high concentration of good or bad things exist in 3D environments. Primates grasp arms long distance.

Motion is used to keep body temperature within bounds. The hypothalamus is the top of the hierarchy of control.

The cerebellum evolved to correct muscle movement errors and is useful in land animals when they first appeared.

The human brain is divided into four or more networks that help with motion not just the motor cortex. We build motor commands atop preliminary motions. Humans can teach offspring behaviors not genetically built in.

Thinking was not mentioned in the video. They don't have a model for human cognition yet because of experimental limitations and ethical consideration. I didn't recall how hippocampus, amygdala or basal ganglion functions.

I believe we can conclude that thinking is a controlled process like the others. It might be self reference for understanding consequences with feedback memory.

@Hadoblado
@EndogenousRebel

I would like your help in understanding these processes. Both of you have studied this subject more than I have in the biology.

 

[fluffy]

I thought more about IQ tests.

The number is a relative term.

Yet it has to mean something since the top one or 0.1% does exist.

I don't know what people do at these levels because so many skills can be developed.

It's just that they develop faster at the top.

My guess is that memory plays a big role.

Not exactly memorization by rote muscle memory. But trying different things in mind.

I never figured out the rubix cube.

I suppose if you take the shape of the pattern in orientation and sequence you can solve it but that requires flipping it around in many ways to see the right way. It requires to much of my memory to see which ones exist in space.

When you do an IQ test a quarter of the test is about shape patterns. Fitting them together and manipulating them. This is an excellent way of determining your proclivity to do technology whenever a technology requires multiple parts handled together. What you can do with them.

Of course there is also the abstract part. Visual people can put things together but if a math problem happens the rules may become obscure to them. A variety of ability exists. And we all learn rules in different ways under different circumstances. It's not one way for everyone.

I sometimes have a hard time doing projects by myself. I get really abstract but cannot make anything concrete. There are those people who can make things by themselves and I wish I could do that. I have so many ideas that just don't get made.

 

[fluffy]

It occurred to me that some people have greater metacognition than others.

Maybe as we think about intelligence it needs a better delineation between peoples ability to solve problems vs know when something's up not really part of the obvious circumstance.

Pattern recognition and executive function are the main functions of intelligence. These allow problem solving.

Yet metacognition is more about the reason behind why we solve problems.

One time Facebook said my IQ was 155 but I cannot solve problems harder than 125 on IQ tests. Working memory is limited. But maybe I have high metacognition and Facebook made a mistake understanding that.

Here is my understanding of what people can do for example at different levels of working memory.

75 operate a microwave oven instructions for a tv dinner

100 fix a car, understand calculus

125 write a college paper or book on math, make simple computer programs, reverse engineer complex spatial designs.

150 make a sophisticated video game by themselves in highschool - utilize quantum mechanics in solving statistics problems. Understand brain structures enough to replicate them.

175 understand physics as to designing fusion generators. Make a.i. systems that solve moderate problems by themselves. Look into the physics of UFOs. Create neurointerface with the brain to read the mind. Plan ahead what purpose their technology will be used.

 

[Hadoblado]

Facebook didn't make a mistake. Those tests deliberately inflate scores to improve the chance they're shared. Facebook is not a psychometric expert, it's a platform for competing over attention in exchange for ad revenue.

 

[fluffy]

Facebook didn't make a mistake. Those tests deliberately inflate scores to improve the chance they're shared. Facebook is not a psychometric expert, it's a platform for competing over attention in exchange for ad revenue.

It wasn't a test.

Facebook does network statistics and a.i. related data collection analysis.

 

[Hadoblado]

Oh. I've never heard of that. My bad. I do remain skeptical it could give valid IQ estimates, but it doesn't fit the systematic over-estimation I described.

 

[fluffy]

Oh. I've never heard of that. My bad. I do remain skeptical it could give valid IQ estimates, but it doesn't fit the systematic over-estimation I described.

In 2012 they found out that people who liked curly fries were more likely to be over 115 in iq because a bunch of college students began liking it. Cambridge analytical download a bunch of user data in 2016 where they said enough data points can predict your personality better than your wife and family and friends. That is also when Facebook was doing work in a.i. to stop hates speech on its platform and picture neural networks could tell what was in photos. Advertisment also needs accurate data to find who to sell to because patterns can tell what you will buy by what you post.

In fact Facebook has such huge computational capabilities that today they have a profile on everyone on the planet. They make predictions constantly.

When I was on Facebook in 2016 I had friends who were university professors and other groups and I created user group for quantum physics and religion. Back then if you asked the a.i. to analyze your entire profile it give accurate answers more truthfully but Facebook saw the election fraud and gave away less information to keep secrete it's projects from the public. Kids being bullied accelerated research when Zuckerberg went to Congress and parents confronted him over how many committed suicide because of the platform. He said it take five years (this was in 2017). But what I saw their experiments were bigger and in the open before Trump won.

 

[fluffy]

For some people it is easy to rotate things in the fourth dimension. I think this is required for a.i. - it's a way of finding combinations faster that would work. But you'd need to integrate it with a memory system. The way brain networks are set up is recurrent (sending the signals back into themselves) so the base structure is 4D to begin with. It's how we navigate the world. Working memory is a layered back and forth in the forest of that dimension. Some would know how to see multiple paths in the forest at the same time from the higher perspective. Like a 4D being would.

 

[fluffy]

Thought about it in my sleep.

The reason objects get closer together is because the universe exists in the space of the fourth dimension. But there is no gravity particle like the graviton. No, it's just a higher relationship between objects that when they move through it they come closer together. That is why electricity force and gravity force do not interact. It could be that we can use angular momentum in new ways to counter the higgs field otherwise we cannot have antigravity.

-

Intelligence as I was thinking uses surfaces to control the body and thoughts.

Thinking is a good example of what intelligence is. You can do actions trial and error but with thought you do so in the head and remember what you got wrong and right. This requires you visualize what fits together with the surfaces because seeing shape is 2D because if they fit they fit.

That is why the body is necessary to be intelligent because fitting things together you just need a 2D skin surface and why the brain is made of surfaced neocortex.

It is the thinking or planning ahead that make humans different. We make things for a reason. To get things done. So we need to understand consequences and metacognition allows us to complete the tasks required. The connections going back into themselves to learn and plan and think as a special kind of attention mechanism on a surface.

 

[Hadoblado]

What do you mean by the fourth dimension exactly? Why is it easy for someone to rotate objects in the fourth dimension mentally but then this person is not considered a fourth dimensional being?

 

[fluffy]

What do you mean by the fourth dimension exactly? Why is it easy for someone to rotate objects in the fourth dimension mentally but then this person is not considered a fourth dimensional being?

I think it's a math ability to do so.

We have three dimensional bodies but the brain can imagine limited spaces above that dimension (perhaps as a network structure that can rotate in that space). We cannot be four dimensional beings because atoms are not making the body density higher than what we already are. Our Space does not have the directions we'd need to move in with our atoms as a fourth dimension like hyper cubes or hyper spheres, but some people can imagine what those objects do.

 

[fluffy]

A rubix cube has 27 cubes inside it. 3^3

A four dimensional rubix cube would be 3^4 or 81 parts.

You can twist a rubix cube 12 different directions because of the axis points. 12 = 3 • 2 • 2

Id imagine you can move a hyper rubix cube in more axis points including twisting but also shifting trough the fourth plane.

4 • 2 • 3 = 24

24 could be the numbers of ways you came move one step in the hyper directions.

I am not sure? I never studied axis in Four space.

 

[fluffy]

The frontal lobes can be used for self reflection and thought. To do so requires a top down mechanism.

The dorsal lateral cortex coordinates the top down processing. With the partial lobe as spatial recalibration.

The ventral lateral selects and shifts objectives. The ventral medial looks for opportunities. The orbital frontal weight options. The dorsal medial inhibits incorrect action. And the anterior cingulate routes signals for consequences learning when errors are detected.

First you orient yourself (DLPC parietal) then you do an action. If a new action is better, that's to say the ventral medial detects a better option the DLPC inhibits the dorsal medial which stops inhibiting the ventral from shifting. The OFC must agree as it tracts both options. Then the ACC detects if this works and wires up the consequences. Monitoring progress.

When thinking there is an area that sends signals backwards to imagine new scenario. This is a shifting where you send new data into working memory and if it's a better idea you remember it. Compare and contrast.

The top down mechanism must keep track of everything. Separating tasks into subtasks and making imagination separate from reality of the current concrete goals.

Learning is a build up set of successful and failed goals at the smallest task to the largest. Mental building blocks where attention to what's happening gives weight to figuring it out.

 

[fluffy]

The brain is folded up so these two dimensional representations do not exactly show how people operate in the 3D world.

It really would not show you how to make an agent that could play video games.

Tho there is some mathematical relationship you can model in people it only works in crowds. Looking at higher resolution is computationally expensive.

Signals intelligence where you look for reflections in where they come from, the source and how change is normal or abnormal behavior. Then a model is built.

Cog science then can be seen as part of this but expect experimental results to follow certain trajectory in chaos theory. Small rounding errors can make large effects in weather models so medium size simulations can become unpredictable fast.

 

[fluffy]

Cognitive Control (Executive Function)

I made a small model of it.

Knowing when a new task becomes more important than what one is currently doing is paramount. This requires evaluation of what is done and monitoring if another value exists. So taking two thought at once. Then implementing them where past experience tells you what has worked and if new strategies need to be tried out. Thinking about what could be done before doing it.