Large language models and fuzzy logic.

Large language models (LLMs) are problematic for programmers. They require a new way of thinking about programming. The key element in those systems is the input mode or input port. That understands spoken language. The system requires a model that transforms spoken language into text and then drives that text to the computer. And the text must be in the form that the computer can understand and turn it into commands that it can use. The system must also turn dialects into literal language that it can use for commands.  This is the first thing that requires work. The programmer must teach every single word to the system. 

The practical solution is to turn the word into numbers. In regular computing. Every letter has a numeric code called the ASCII code. The capital A (big A) has the decimal code 65. The programmer must realize that the small "a" has a different numeric code than the capital A. The little "a"'s ASCII decimal code is 141. That's why things like passwords require precise letters and if there is a capital letter in the wrong place the password is wrong. 

So, if we want to make the system more effective. We can give a numeric value for every single word that we find in the dictionary book. We can simply take the dictionary book and then give serial numbers for those words. The word "aback" can get the number code 1 (one). That thing makes it easier to refer to those words. Every word must be programmed separately into the system. And that makes programming hard. The other thing is. If we want to use dialects we must also program those words into the LLM, 's input gate. That programming is not very complicated, but it requires a lot of work. 

Diagram: Neural network

In human brains, neurons are the event handlers. In artificial, non-organic, non-biological computer networks, or computer neural networks computers or microprocessors are those event handlers. In human brains, thousands or even millions of neurons participate in the data-handling process. Those neurons make fuzzy logic to the brain. 

The idea of fuzzy logic is that many precise logical cases can make the system mimic the fuzzy logic. Fuzzy logic is a collection of precise logical answers. 

Another thing is that we must make a system that uses fuzzy logic. Making fuzzy logic is not possible itself. But we can create a series of event handlers that make the system seem like fuzzy logic. The idea is taken from the human nervous system. When a large number of neurons participate in the thinking process that makes the system virtually fuzzy. Every single neuron uses the precise (YES/NO) logic but every single neuron has a little bit different point of view to the problem. 

So the system uses a model that looks like the grey scale. There is the white that means YES and black that means NO. And then there are "maybe cases" between those YES and NO cases. Those "maybes" are the absolute logical event handlers like neurons. When that group of event handlers gets its mission, every single event handler selects YES or NO. Then the system calculates how many YES, and how many NO solutions it has. So those event handlers give votes to the solution. 

The model is taken from quantum computers. In quantum computers, data, or information travels in strings and finally, every string has values 0 (zero) and 1 (one). You might wonder how much power that kind of system requires if every event handler must process information. Before it answers. But then we face a situation where the system must answer "maybe". Another way to say "maybe" is XNOT (or X-NOT). Or if the answer is closer to "yes" another way to say that thing is XYES (or X-YES). X means that the system waits for more data.  

The system might say. That it does not have enough information in the data matrix. That is a large group of databases or datasets. And that is the major problem with AI. If the votes on the scale of "YES to NO" are equal that means the system has a problem. If the AI controls the robot that is in the middle of the road and votes are equal that robot can just stand in the middle of the road. Another thing that we must realize is that these kinds of systems are the input gates. Data handling begins after the system gets information into it. 

https://en.wikipedia.org/wiki/ASCII 

Researchers are found math neurons in human brains.

Image 1:)

Researchers are found math neurons in human brains. That thing gives new information on how and when the neurons are starting to specify. The discovery of the neurons that are specialized in mathematics is opening new paths for neurosciences. Because there are the math neurons there should be language neurons. 

And maybe there are specialized neurons for every type of skill that people have. So there could be cooking neurons and social neurons that are controlling social activities in the neural network. 

When researchers are finding when the separation of the neurons starts. That thing makes the revolution in the training and education. Learning new skills is rewiring brains. And if that process can be controlled by researchers, that brings the new and powerful educating tools to the hands of the educational processors. 

That kind of research can serve the medical care of brain damages. There is a possibility that in the future. The person whose brain is injured. Can get the neural transplant. That thing can make by using the cloned neurons. But the problem is: How to transfer the skills, that people had to those neurons? Every single skill that a person has is stored in neurons.

Image 2:)

And that thing means that if a neuron is lost. The data includes skills that are stored in those neurons gone forever. 

That is the thing that makes brain damage so hard to recover. Researchers can clone neurons. But the problem is how to recover skills and memories that are gone to the cloned neurons. Skills are memories like all other things. 

The thing is that when we are expanding our understanding of brains and their function we are getting tools how to make more effective neural networks. And those things might consist the hybrid systems where the living neurons are connected with artificial intelligence and non-organic sensors. 

We know why our brains are so effective. Brains are made of billions of neurons and that thing makes them effective. But the thing is that the brains are starting their data handling process same time in multiple locations in neural structure. 

And that increases their effectiveness. When some neuron group is needed for some other mission, that group will store the thing in memory. And that allows that other neuron group will continue with that thing. The internal axon structure can transmit the data that is stored in memory to anywhere in the neural structure. 

The new type of neural computers, networks, and deep-learning AI can use the data that is collected from the sensors. The deep-learning systems are following the success of the robots by using certain parameters. Those parameters can be simply how far a robot can operate without causing damages or getting damaged. 

The fuzzy logic makes that kind of system "quite easy to make". At the first, the AI would operate by using a large number of actors. When some actor will make mistake the AI drops that thing away. If the AI would follow robots that system records how far a robot can travel or how long it can operate. 

Then the parameter can be simple like this: (number of actions/time unit). The best result would be stored in the memory of AI. And then the AI would use mission records to find out where the mistake is made. If a robot falls into the canyon. 

The AI would control those machines to drive farther from the canyon's edge. So the best result is the same thing that makes the AI select a certain route. And of course, if we want to connect the number of actions to the robot. That thing can be picking up the ground samples. 

https://scitechdaily.com/brains-of-cosmonauts-rewired-during-space-missions/

https://scitechdaily.com/hiddenite-a-new-ai-processor-based-on-a-cutting-edge-neural-network-theory/

https://scitechdaily.com/specific-math-neurons-identified-in-the-brain/

Image 1:)https://scitechdaily.com/brains-of-cosmonauts-rewired-during-space-missions/

Image 2:)https://scitechdaily.com/specific-math-neurons-identified-in-the-brain/

Precise logic vs. fuzzy logic.

  

 Precise logic vs. fuzzy logic.

The "barren plateaus" trap is one of the most cutting edge things in computing sciences. The precise data handling is working perfectly in a closed and very unilateral environment. But in the open environment, fuzzy logic is a more effective tool. 

And the power of fuzzy logic increases whenever the environment is turning larger and the data is turning more versatile. Fuzzy logic is in the key role for handling non-sorted data mass that is collected from the natural and chaotic environment. 

Have you imagined the situation that you go to the flat, where everything is in a certain order? You can't find anything from there, and then some other person comes in. That person finds everything that is needed less than the second. 

You cannot ever find anything faster than that person because that person knows everything in that room. The person knows every single nail in that flat. And that thing makes an impression.  But the prime question is does that person know anything else?

Does that person ever go outside? We can say that if the person would stay all the time at the same flat that person knows everything that is in the flat. There is no other information than the information about things that are necessary for finding things from a closed environment. 

That person would, of course, look outside sometimes, but the information that person gets is limited. There might be a car in the front of the house. But the person would not know why the car is at the front of the house. 

The limited environment is taking us to the "barren plateaus" trap. The person who lives in a closed environment seems to be wiser and more intelligent than a person who lives in an open environment. 

When the person who lives in an open environment comes into a strange space, that person must search for everything. The person who lives in that closed space must only remember where things are put. And then pick them from those places. 

In closed and very limited space everything can have certain and precise orders. But if the environment is very wide and the data is very versatile. The answer to the problems is fuzzy logic. 

In a closed environment, the person might make notes. And read where things are left. But if we are in an open environment making precise notes for things where we left something is impossible. Or those notes must be very large and that means they turn uncomfortable. 

If the person who is operating in an open environment wants to use precise logic. The person must read all notes all the time. For finding the answer to the question, where that person left for example the screwdriver? 

So the precise logic is not a practical thing for an open environment. The person must turn to use the different types of notes. That person can use multiple memo books where is marked places where is used certain tools. If we want to keep memos by sorting them by place. 

There is the possibility that we forget to put them there if we use the screwdriver. Or if we want to sort data by using places we must always read every single line in every notebook. And hope that if there is a marked screwdriver.  

But what if we sort those memos by using the name of the tools that make it possible to find the individual tool more easily. The most effective thing is to use different memo books for each tool. 

And the data is sorted by the name of the tool. When we want to find the screwdriver we must just take the memo book where is word "screwdriver". We can find the place. Where did we use that tool the last time? The searcher should find it on the last page. 

If there is a mark that the screwdriver is taken for use. But it is not returned which means that the screwdriver probably found that place. And the exact place where the searcher can find that tool. Will be found from the work order. This thing is called fuzzy logic. In computing, those notebooks are the tables of the databases. 

https://likeinterstellartravelingandfuturism.blogspot.com/

The problem with AI is that it cannot flex

Artificial intelligence is smart. But the cooperation of AI and others is not so good as it should be. The problem with artificial intelligence is that it doesn´t flex with regulations. One way to demonstrate this thing is to think of the driver who is always following the law. 

If there is a "stop" sign at the crossroads that thing shows that another car has the right of way. But the problem is in cases where the AI-controlled vehicle has the right to first. What if another vehicle is near the crossing point and doesn't slow or stop? If the AI cannot flex. That means it would cause an impact. 

The problem with the programming of the AI is what kind of actions involve its programming. The traffic regulations are the thing that requires that everybody follow them. But what if somebody doesn't care about rules. And doesn't care about the "stop" sign? 

When artificial intelligence operates it doesn't use imagination. So it predicts that also all other people are following the rules. But the situation is different in real life. In real life, we are facing situations where we must flex. But the question is: how to tell the machine that it should flex in some situations?

X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*XX*X*X*X

Fuzzy logic

The idea of fuzzy logic is that it makes machines like a robot working smoothly and more flexible. The system has the main action like the walker has the right to go before the vehicle. But then there are sub-determinators like the distance and speed of the oncoming vehicle when the robot should let the vehicle go before it. 

In the airfields, fuzzy logic used computers might have determinators what makes them lead the aircraft to the runway. Normally the system just follows the row and uses runways simultaneously. But in the cases of emergency, that system can let the aircraft pass the row. 

That kind of thing can make aviation safer. Even if the computers are not leading the aircraft to ground independently they can show the necessary information for air traffic controllers. That data can involve things like information on the free runway. And the system can send the rescue crew to that runway.

X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X*X

We all know that the walker should go first in all cases when the road is needed to cross. But the fact is that if the walking robot is about one meter ahead of the car and it doesn't have regulations what is the safe distance of the vehicle when the robot can cross the road. It walks just under the vehicle. 

The answer to that problem is called fuzzy logic. The fuzzy logic means that the system has no strict orders for that kind of thing. The parameter for crossing the road goes that the walker must always let go before the vehicle. But then the programming of the robot has the values like distances and speed of the incoming vehicle. Those determine the exceptions for the main rule. 

Those exceptions are making the AI-driven robot smart and flexible. The thing is that the system as the rules what to do in certain cases. But then some sub-determinators are adjusting the actions. In cases like the vehicle is near the robot. 

https://visionsoftheaiandfuture.blogspot.com/