Monthly Archives: December 2012
How would you give a description of what “difficult” is? Wouldn’t you say that it is something we cannot easily do or comprehend? Something that is unfamiliar, that we have not exposed ourselves with, that needs skill and non-ordinary effort? Something complex?
Noesis Theory can consolidate all those definitions/explanations into one that has to do with the workings of the brain and one that… is actually quite simple 🙂
Difficult is any external stimuli that is out-of-context and that is governed by a level of complexity significantly larger than the one represented by the current state of our brain neurons that are tasked with “handling” these incoming signals. In other words, whenever the complexity of our environment is greater than the internal complexity we have managed to build inside our brains, in order to comprehend what is going on (in order to be able to predict the incoming signals and thus turn them into in-context), this is what we call difficult.
Even simpler “if complexity outside the brain is greater than complexity inside the brain => difficult“. Simple, isnt’ it? 🙂
The truthfulness of the above statement can be easily deduced by thinking of the methodologies we’re using in eliminating difficult: we explain, we train, we experiment, we repeat… What is the purpose of this? It is an attempt to build inner neural complexity! Somebody can explain you a set of rules of how this stimuli occurs (explain)…. can showcase you specific occurence and the patterns it follows (train)… or if no rules are available can let you experiment or try again and again until you stumble upon the pattern by chance or method. In all of these cases, the end target is the same: you are trying to understand the rules of how this external system occurs and construct an internal representation inside your neural networks of the same type of complexity, in order to be able to predict at any time the potential expression of the external system. In this way you have transformed the out-of-context to in-context, you have PA links ready to send signals from input directly to output and you can handle this stimuli without thinking (i.e. without creating Driving Pockets and stealing Battery power).
One additional thing to note is how specifically this translates into signal traversal: since one of the basic rules that we have discussed in Noesis Theory is that “similar incoming signals traverse by default to neurons that are in the save “vicinity” (few hops needed to go from one to the other) this means that “easy” would be something that for minor alterations in input would require small alterations in the output needed!!
And reversing this (de morgan style), we have an even more specific (but possibly not of full coverage) definition of difficulty!
Difficult is a set of stimuli that for small alterations in the input require significant variation in the output.
So if you were tasked to translating Chinese and you were a European, you would need from an input of a set of vowels and consonants that are the same in both cases, produce as a translation a completely different word, depending on the pronunciation used! I.e. for a small variation in input (different pronunciation of the same syllables) you need to produce a significantly different output… This is difficult! And you need training and effort to construct a system of neurons inside your brain to match this complexity and be able to distinguish between these small input differences.
Do you want another example? If you’re playing with a Pro tennis player, he can quite easily insert some significant spinning into his throws towards you. But the exact amount of spin (and thus the final direction that the ball will travel to) is not easy to distinguish because you have very little input to judge (a slight variation in the trajectory of the ball until it bounces on the court, as well as possibly a slight adjustment in the hand positioning of your opponent when hitting the ball) thus it is quite difficult for you to adjust significantly your output to cope for the different responses you would need to give, depending on the spin of your opponent’s throw. And the fact that you have very little time to react increases the difficult even more because more intense action might be needed, i.e. more adverse output for these small input variations.
Isn’t difficult simple now?
I have noticed two distinct categories of ability in humans: specialized and generic. They certainly can be found in different variations of quantity in various humans, but it is easier to note (and thus study) them in the “extraordinary” amount; so we’ll now refer to this quantity type.
The first type of ability, the specialized one, is more common and more focused. And I have come to the conclusion that it must closely relate with the amount of neurons & neural synapses that are available in the specific area of the brain that handles the corresponding type of activity. For example, if somebody has a lot of neurons in the area of the brain that handles body movement, he can become quite easily a very able dancer.
Side-note: I don’t know much (actually almost nothing) about brain physiology so I don’t know if it’s the amount of neurons or the amount of synapses or something else, but I am certain that a structural element of the brain is more abundant in a specific location/area of the brain and this is the differentiating factor that drives a person towards excellence in a specialized field of human expression. This exact factor allows for more “complexity” to be stored/categorized/sorted/handled in this area of the brain and thus what is judged as difficult for others is easy for this specific person (I’ll have to also write an article about what difficulty is and it will become clearer then).
In the same manner, somebody with “a lot of neurons” (or that other structural thing that I don’t know) in the visual cortex is very capable of remembering many visual details and all the different faces that he has seen, whereas I for example cannot store anything but the very basic details about images & scenes of the past.
There are many places where a person might happen to have a greater concentration of neurons (if any) in his brain and thus give him an advantage over his peer on a specific set of skills. That’s why we can say that some people have a high IQ, or a high EQ, or high social intelligence (SQ?), or are very able with speech, or have a vivid memory, or have excellent motor skills. Now, it doesn’t mean that by having this extra concentration he will become extraordinary in this field; as we said there is a significant variation in the quantity of “extra” neurons somebody can have, but also in the environment that can nurture such growth with relevant stimuli or hinder growth by not applying the proper incentives or opportunities. But what is certain is that if somebody has this extra quantity, it is certainly within his potential to become great in this area, whether it materializes or not.
Finding the brain area with this structural advantage should be a major goal of the professional orientation courses for kids… After all that’s why we send kids to various activities in early age and try to assess in what they’re good at! Do they have good motor abilities and are suited for sports?
Finally, I would tend not to discriminate between these types of specialized abilities. First and foremost because nobody had a choice as to where to allocate his/her extra neurons, second because having any type of specialized ability is uncommon with regards to the rest of the populace (many don’t have such specialized abilities at all!) and third because most extraordinary special talents can prove to be of use.
On the other hand, what is quite special and rare is the “generic” type of extraordinary ability. I have seen very, very few people with such a type of ability. It’s the type that you give them a task, however unfamiliar to them, and with time they master it and jump ahead of the pack. And of course it also has to do with the driving forces of these individuals and the environment which has we said can enhance or reduce this expression, but the key differentiating factor is that they can be very good at anything; and this is truly special. They can be good at math, but then you expose them to something very unrelated like understanding and dealing with your emotions and soon they grow to have a very good EQ. They might have started as totally unsocial, but then you bring them to the real world you show them the importance of good social skills and soon they can manage people and grow to a high social aptitude… and the list goes on. Of course one might argue that this type of ability, the generic kind, has much lesser chance of achieving true greatness in a field (e.g. a person with general ability will never be like Will Hunting). But I would really prefer if we had more of Leonardo Da Vinci in this world than of Einstein. In any case this is just a personal preference, making predictions as to what of the two would benefit more humanity if you had to choose…
Finally, what I wanted to point out regarding generic ability is (in order to bind it also with Noesis Theory) that I believe that the source of such a type of ability is predominantly not attributed to the amount of neurons/synapses. Yes, it can play a role, but I believe the main differentiator of this type of people is the way they build feedback. Something in the feedback mechanism is supercharged in their brain and they can learn any type of knowledge faster and more efficiently than their fellow humans. I’m not in a position to postulate on the exact mechanism, but if I had to search, feedback would be my first target.
One big misconception that is well justified but dates back to Asimov’s days (or even earlier, I’m not that knowledgeable in the history of science fiction) is that robots do not have emotions. It seems pretty reasonable if you come to think of it. In the mind of everybody they’re made of algorithms, chips & metal, and a very structured language that leaves no room for emotion and inspiration. The essence of the human psyche cannot be fitted inside this pile of metal. It would be inappropriate and even degrading for the human nature.
In other words, according to public opinion, it’s ok if we manage to replicate our intelligence and rationality into another kind of existence (after all, it will be our creation, an expression, a copy of our mind), but only if it had no feelings, no true emotions. We would still have an edge over it. And we know from countless creations of literature and cinema how much we value our emotions and how useful they can be in overcoming difficulties and managing to surmount circumstances where pure, simple logic may be lacking. Emotion for us humans, even for hard-core science fiction fans or AI researchers, is a temple that should not be decimated. It is the ultimate boundary between what is true and what is artificial; a differentiating factor that sooths our mind when it would feel threatened by the AI manifestations that traverse and trespass into our realm of humanity.
Well, guess what: we were wrong! And it was not easy for us to figure it out… The road we had taken thus far when trying to create AI clearly indicated towards the inexistence of emotion. But if you look at Noesis Theory and you also believe that this is the correct way towards “true” AI (yes, still this is just a belief that has to be proven), you will see that emotion is everywhere. It is an integral part of the algorithm. And its name is… “Driving Pocket”.
Driving Pockets are the emotional responses of the Noesis algorithm. Think of it: what are the 3 key elements of an emotion?
- It must have something to do with me (affinity of the incoming signal)
- It must be different from what I expected it to be (out-of-context)
- I should not have an easy way to alter it and make it the way I would like it to be (because in that case, the out-of-context experience would diminish and I would stop dealing with it)
The same 3 factors are also applicable to a Driving Pocket: it is produced by out-of-context stimuli (factor 2), it is produced by “agitating” a Driving Force to promise us some pleasure/discomfort (factor 1) and we have no easy way to turn it off, i.e. we have not created some P-A links to spring into action and resolve this out-of-context experience (factor 3).
The conclusion: Driving Pockets are the “emotions” of a brain that implements the Noesis Theory model. This means that any Artificial Intelligence that would be built with NT would automatically have emotions as an integral part of its operations. Even more specifically: it would not even be able to operate normally without experiencing emotions! And as the babies are full of emotions and as they grow up they learn to control their emotions, the same would hold true for this AI. At first it would be full of Driving Pockets (emotions) and with feedback and linked learning it would gradually transition into a state with less emotions, just like grown-up humans. But the emotions would always continue to be part of its normal functioning; it would always be its incentive to act, just like the same is true for humans.
I don’t know if this frightens you, but it is what the future holds and it is the only known way to create a sentient being. So whether we like it or not, true AI and emotions will go hand in hand, and we’ll have to find another edge for us.