The AI requires new hardware. Servers that can run the AI must have a lot of power.

"Scientists have created an ultra-efficient AI accelerator that uses light-based circuits instead of traditional chips—making AI faster, greener, and ready for massive scale. Credit: SciTechDaily.com" (ScitechDaily, Scientists have created an ultra-efficient AI accelerator that uses light-based circuits instead of traditional chips—making AI faster, greener, and ready for massive scale. Credit: SciTechDaily.com)

One solution. That can help to save energy. The system recycles the heat of its microprocessors. 

That heat can expand liquids. That is used in freezers. That expanding liquid rotates the generator's wheel. That helps to save energy on those servers. So, operators must bring less energy from outside. 

Temperature makes those servers ineffective. When the temperature rises too high, it slows the computers. Temperature rises oscillation. That raises resistance. The resistance denies the signal going through. 

Or resistance destroys data in the signal. That causes the system. Must raise power to make a signal travel through 

The resistance that temperature damages data structure. That travels in them. So, the system goes in a circle where. It must raise its power to make the signal go through. 

And finally, those wires melt. 

"A Harvard-led team created the first self-contained, chip-based laser that pulses in the mid-infrared – ideal for advanced gas sensing and imaging. This innovation could power next-gen tools for climate science and healthcare. Credit: SciTechDaily.com" (ScitechDaily, Harvard Builds Laser the Size of a Chip, Bright Enough to Map Invisible Worlds) The Harvard laser chip can also transmit data into photonic systems and out from it. 

So, the answer to the problem is the photonic microchips. The system can use photons to transmit data to silicone. 

That means. The system can use a photovoltaic phenomenon when it hits the silicone layer. 

There, light forms electric impulses by shooting electrons away from their orbitals. The photonic microchips can advance AI. Those microchips are colder than the electric system. Those photonic systems can operate for a longer time with maximum power than regular microchips. The photonic microchips can use lasers. To transmit data to them. There are three versions of the photonic computers. 

1) In the simplest version.

Data travels between components in laser rays. 

The system can use flat lasers to transmit data between those components. 

2) The system. The internal structure of microchips uses photonics in data handling operations. 

But prisms and mirrors that the system uses as routers and switches. And gates are electrically operated. 

3) All photonic computers: there are laser beams that transport information. Also transmits energy to those microchips. In those systems, miniature crystals can also aim information in the right way

The difference between electric and photonic computers is that data travels in a laser beam in photonic computers. That means there is no resistance. That means photonic computers are more effective than electric computers. And those systems can be the next tool to develop physical systems that can run complex AI-based systems. 

https://scitechdaily.com/ai-at-the-speed-of-light-how-silicon-photonics-are-reinventing-hardware/

https://scitechdaily.com/harvard-builds-laser-the-size-of-a-chip-bright-enough-to-map-invisible-worlds/

Large language models (LLM) are impressive. But can the small, compact language model be more effective?

Large language models (LLM) are impressive. But can the small, compact language model be more effective? 

Large language models can search and process any kind of data. That we might need. The thing that makes LLM a little bit complicated. They must analyze the data that they get. And that requires lots of calculation power. That power is not free. The LLM requires an entire group of supercomputers. 

The problem is that those LLMs require the most powerful hardware in the world. Large supercomputers require lots of electricity. 

Those systems use as much electric power as a small city. That means the LLM servers require their own power plants. That's why. Things like miniature nuclear reactors are under research. 

The only realistic way to use LLM is to use them over the net. The private actor doesn't have money for the electric bills that those servers require. And that brings one data vulnerability in the data handling process. When some other actor operates the server, where we share data. That another actor can get access to that data. The ability to bring the language model to its own servers gives the actor full access to the data. That means that the actor who administrates their servers owns and knows everything that the server does. 

The small language models (SLM) can act as multicore AI there each core can operate independently. 

The small language model (SLM) is the answer to that problem. The SLM can be like a droplet that is separated from the LLM. That lightweight language model is the tool that a normal actor can load to the power servers that are used in the CAD program's network use. The SLM is a limited language model used for special purposes. So the LLM can remain in the common use. But in secured use, the SML can be the tool, that is in the future. 

The SLMs can create a new model for the AI core. The SLMs can make it possible to create a multi-core AI system. Those cores can remain on their servers. They can operate independently without affecting other processes. 

The multicore models are tools that can turn the next page for the AI. That system can be more economical and environmentally friendly than the monolithic core. The system can switch unnecessary cores off when the system or users don't need them. 

https://www.wired.com/story/why-researchers-are-turning-to-small-language-models/

The new way to communicate using light.

"Propagation of light pulses with intermodal dispersion in multimode fibers, MMF. Words are mapped to different frequencies (different colors), the optical pulses are stretched into different temporal dispersion curves. This allows for the identification of the pulse frequencies at the receiver, thereby enabling the reconstruction of the input words. Credit: Gao, Z., Jiang, T., Zhang, M. et al." (ScitechDaily, Scientists Just Taught Light to Transmit Meaning – And It’s Revolutionizing Communications)

There are many ways to make light-based communication systems. Normally, we think that the light flashes mean 0 and 1 in a laser communication system. But the 0 and 1 can be two different wavelengths. That means the flash of red can mean 1, and the flash of green can mean 0. This is a useful tool for binary data transmission. But there is one way to transmit data by using light that we don't usually mention. 

In that version, a single wavelength means a single word. The system must only give serial numbers for light- or radiation flashes. Then the receiver can put those flashes back into a certain order. That means red can mean "if", green can mean "and" etc. Researchers can also use radio waves in the same way as lasers. And the point of the CCD camera those people can put a radio spectral scanner. 

There are approximately 600000 words in English. Some people might think to use that kind of data transmission, the system requires 600000 wavelengths. The system can send those messages by blinking the colored lasers in the order that gives words to the receiver. If one color carries one word that makes the system immune to disturbance. 

"Experimental system. a, encoding: Each word is mapped to an optical frequency according to the encoding table. b, Schematic of the experimental setup. c, decoding: Recognize each dispersion curve as the corresponding frequency using the fingerprint dictionary, and then reconstruct the words based on the encoding table. d, implement sentiment analysis using a deep learning model. Credit: Gao, Z., Jiang, T., Zhang, M. et al." (ScitechDaily, Scientists Just Taught Light to Transmit Meaning – And It’s Revolutionizing Communications)

The problem is that. Language advances all the time. And the system requires very advanced and complicated transmitter-receiver pairs. But theoretically is possible to create a CCD camera that pixels can separate 600,000 frequencies. And, a transmitter can also have the ability to send 60000 separated frequencies. 

And that means the system must have flexibility. If the system uses the English alphabet. It needs only 52 frequencies for the letters. There are 26 capital and 26 lower cases. And other frequencies for numbers and special marks like brackets. Or the sender can use the words to send numbers. But that creates limitations in massages. 

The system can use red as "A" and the light flashes can involve data which place the certain letter takes in the words. That thing is hard to jam. 

So it's immune for the outcoming errors. But otherwise, it's easy to break. 

The system can also encode lots of words in one wavelength. There can be hundreds of words encrypted in certain colors. The system must only send the serial number of the word that the system sends. This kind of system is more complicated to break. That thing makes this thing more vulnerable to disturbance. 

https://scitechdaily.com/scientists-just-taught-light-to-transmit-meaning-and-its-revolutionizing-communications/

The new things can make quantum teleportation closer than we expect.

"Quantum teleportation has been successfully conducted over a fiber optic cable carrying Internet traffic, merging quantum and classical communications efficiently. Credit: SciTechDaily.com" (ScitechDaily, Quantum Teleportation Becomes Reality on Active Internet Cables)

The term Quantum teleportation means: information that the system transmits. Is bound to a particle. Or some other structure. Like a hovering electromagnetic field. Regular data transmission is like talking to a radiotelephone. Everybody who has a radiotelephone can hear that message. 

The model of quantum teleportation is the USB stick. We can put the USB stick in the pocket.  That is a very good and secure way to transport data. But the problem is that it's so slow. If we would be in some house. We could use the pressure tube or pressure post to send those USB sticks that can involve the code for the day that the system can use to decrypt the encrypted messages and data. 

In quantum teleportation, the observer cannot see the information. The model of quantum networks is the pressure tube post. That system is visible in warehouses and apothekes. 

Basically, the quantum teleporter just packs information into the capsule and sends it to the receiver. In short-distance systems. The system can pack things like qubits in the nano-capsule and send them to the receiver using nanotubes. 

In long-distance quantum teleportation models. That cable moves the miniature data capsule through it. The cable can act as a particle accelerator.

That system is like the human nervous system. The neurotransmitter is replaced by using a nano-capsule that carries information. 

In quantum USB the system can store data in a laser ray. That laser ray jumps between 100% reflecting mirrors. When the system arrives at the receiver the other mirror is open. And then. The receiver reads the data burst. That is stored in the laser ray. 

But the system can also use a magnetic chamber where those qubits hover in a vacuum. When the data transporter arrives at the receiver receiver pumps that data into its sensor. 

"SHIFTY PHOTONS: New experiments explore the continuous transition from photons acting like particles to behaving as waves. Credit: Image courtesy of S. Tanzilli, CNRS" (ScitechDaily, Photons Traverse Optical Obstacles as Both a Wave and Particle Simultaneously)

The ability to transform photons between particles and waveforms can make the quantum network more secure. Than before. The quantum system can pack the information into the light wave. After that, it can turn the light wave into the photon's particle form. That thing makes those systems ultra-secured. The attacker cannot read information from the photon's particle form. 

In quantum wires, the system can use internal quantum fields.  If those quantum fields of atoms are connected into one entirety. The outer layer in the quantum field structure protects that internal structure.

Making the data transportation highly secured. That quantum field can be the hollow laser beam. And if somebody tries to touch the data that travels in the quantum tube. The system can see the changes in that protecting laser beam. 

The model of quantum teleportation is the network where a regular electric phenomenon is replaced using the particles where the information is packed. When information is packed in the particle that means an outsider cannot read it. 

And if an outsider tries to read the information. That is stored in the particle that destroys information. The outsider cannot see which particle the system uses to transport information. And the idea is that the system also sends empty, but numbered particles. And if some of those control particles are missing, that means attack. 

The quantum liquids that make whirls are one thing that can make quantum teleportation possible. The system can make a whirl or tornado-type structure in the quantum liquid. And then. It makes the data transmission through that electromagnetism wormhole. 

https://scitechdaily.com/laser-storms-physicists-create-light-hurricanes-for-16x-faster-data-transmission/

https://scitechdaily.com/photons-traverse-optical-obstacles-as-both-a-wave-and-particle-simultaneously/

https://scitechdaily.com/quantum-spin-liquids-are-real-and-could-change-technology-forever/

https://scitechdaily.com/quantum-teleportation-becomes-reality-on-active-internet-cables/

Einstein is invincible. (Einstein’s General Relativity holds firm)

"Einstein’s General Relativity holds firm as DESI data confirms its predictions on cosmic scales, while also revealing new insights into neutrino masses and galaxy clustering. Credit: SciTechDaily.com" (ScitechDaily, Einstein Vindicated: Stunning Cosmic Map Confirms Gravity Theory Across Billions of Years)

Einstein's greatest idea was that gravity is the attribute in the spacetime. That means the other three fundamental interactions are things that we can describe as the attributes in spacetime. Spacetime is the Higgs field that determines the base energy level in the Universe. 

We can say that Einstein keeps his place. The Theory of Relativity. The gravity models were tested on the Universe scale. And there are no mistakes. That means gravity was similar in the young universe as it is in the modern universe. The difference in energy levels and size of the young and modern universe can explain differences in the measurements. The young universe was denser and its energy level was higher. That means the energy pushed particles away from each other stronger than in the modern universe. 

The term "visible energy" means...

<>Four fundamental forces. 

-Strong nuclear force

-Weak nuclear force

-Electromagnetism

-Gravity (or gravitation) 

E = mc2—In SI units, the energy E is measured in Joules, the mass m is measured in kilograms, and the speed of light is measured in metres per second.

<>Electromagnetic spectrum. 

-Radiowaves

-Infrared radiation

-Visible light

-X-rays 

-Gamma rays

Dark energy is the mysterious wave movement whose origin is unknown.

Dark matter means mysterious gravitational fields or gravitational effects that the source is unknown. Sometimes people believe that the Maser effect in gravitational waves can explain dark matter. Or the dark matter is some kind of hole in Higgs field. The question is not what makes the mass or black holes. The question is what puts the Higgs field moving? The moving field is the thing that makes gravity waves. 

As we see from the diagram only 4,9% of the material is visible. As we see from the electromagnetic spectrum most "visible" radiation or energy requires special equipment. So that nobody can observe those wavelengths. 

The distance between objects was shorter. And that means the energy interaction was stronger. The question is did dark energy is released in the same event as visible energy is the big question. But in a dense and. young universe the visible energy had also stronger interactions than in the modern universe. 

And that energy also pushes objects away from each other. Even if we say that dark energy is the dominant effect in the modern universe, we must realize that in the young universe, the relations between visible and dark energy are different. 

When we think that dark energy is the same as dark matter we can imagine the wave-particle duality. The wave-particle duality is the thing. That makes material. That happens when wave fields cross each other. And the Schwinger effect forms the particles. It forms the particle-antiparticle pair. The idea is that the energy beam travels through across the other quantum field. 

That forms two whirls that turn into quark-antiquark pairs. So that means the particles are one packed form of energy. Dark energy should form particles as well as visible energy. But nobody is sure that theoretical material is dark or invisible to us. If that model is true, we can say that dark energy is behind weakly interacting massive particles, WIMPs. 

The thing is. Researchers believe that the so-called Higgs field makes the mass of all particles. The idea is in that model the Higgs Boson or something like that makes the Higgs field move into the objects. The question is where that Higgs field goes when it moves to the objects. That is an interesting question. The mass and gravity are things that make the Higgs field move. At that point everything is clear. But the direction. Where that field moves is the problem. What denies that we will not turn into black holes? 

In models, the Higgs boson is the thing that makes the mass. That mass forms when Higgs boson vaporizes. And then Higgs field tries to fill that space. The problem is that. Can there be material that has no Higgs boson? And can the photon be the particle that has no Higgs boson in it? 

If something causes too fast vaporization in the Higgs boson. That means the movement in the Higgs field will stop. That means the material would lose its mass. 

In that model, cosmic inflation means. That particles biding Higgs field into themselves. That means that the Higgs field turns thinner but also there are more denser points. So others say the Higgs field accumulates in other points and then turns thinner or lower energy between those points. 

But those things are still fully theoretical things. We know that a particle is one model of energy. Gravity is the attribute in the spacetime. That means the other three fundamental interactions are things that we can describe as the attributes in spacetime. As I wrote at the beginning of this text. 

https://bigthink.com/hard-science/youre-made-of-energy-the-strange-truth-about-where-mass-comes-from/

https://bigthink.com/starts-with-a-bang/dark-matter-same-dark-energy/

https://home.cern/science/physics/higgs-boson

https://scitechdaily.com/einstein-vindicated-stunning-cosmic-map-confirms-gravity-theory-across-billions-of-years/

https://en.wikipedia.org/wiki/Dark_energy

https://en.wikipedia.org/wiki/Dark_matter

https://en.wikipedia.org/wiki/Fundamental_interaction

https://en.wikipedia.org/wiki/Higgs_boson

https://en.wikipedia.org/wiki/Mass–energy_equivalence

The name of the game is industrial espionage.

Chinese introduced their new stealth fighter.J-35A to the media. That plane is surprisingly similar to the F-35A and it's a good example of industrial espionage that dangers national security in the Pacific and entire world. 

Industrial espionage causes a very big problem in the modern world. Network-based solutions bring multiple tools to computer specialists. That they can make qualified solutions. The problem is that. The hackers can also be computer specialists. 

Another problem is that. Hackers use things like non-protected networks for attacking tools. Even if we bring the data to our own servers we would not remove the network itself. When we start administrating our servers, we just move the cloud to another place. And that is the problem. We must know people who have access to data. We must know that they have the skills and other things that they need. 

We remove one thing that costs money. But we use the same tools. That we used in cloud-based solutions. And here we can say that we just turn to administrate our own cloud. We just transport data from one server to another. 

And that creates a situation where my vacuum cleaner attacks banks or defense forces. The hacker can put a small computer program into that vacuum cleaner. And that thing opens a router or wireless router. Or the attacker. The problem with industrial espionage is this. When some data is stolen. Anybody who will use that data. And laws and law enforcement are helpless in situations where espionage happens under the control of the state. 

The laws in Western nations are not like the laws in the People's Republic of China. When we look at the People's Liberation Army, PLA's newest tools we see that they look like famous Western products.  That means that this type of thing can be stolen by using hacking. The thing in this kind of hacking is that the hacker can operate from China or some other country like North Korea and steal information without the need to come to another state. The assistant can carry the remote-operated computer to a western city like Helsinki. Then the hacker can do the job by using the VPN connection. 

Before the hacker and cyber attack boom some Western actors believed that the laws like copyright laws would protect their products. They believed that hackers would go to jail if they were tracked. And the large-scale hacking, supported by governments is a surprise. The thing is that the Chinese military doesn't cover or hide that espionage at all. Industrial spying is one of the prime elements to get access to production methods. And classified chemical compounds that are used in stealth material. 

The stealth materials are physical things. But the hackers can steal things like  AI source code. 

That is a very dangerous tool in the wrong hands. Government-supported operators can create things like the famous AI's evil twins. Those tools can boost the hacking to new levels. The problem is this. When hacking happens under government control those people must not be afraid of authorities. The thing is that when the government is really authoritarian it has many ways to motivate hackers. And one of them is that the secret police use force and fear to those people. 

Lasers and optical microchips can make computing more secure.

"In the new communication system from researchers at Chalmers University of Technology, in Sweden, a weak optical signal (red) from the spacecraft’s transmitter can be amplified noise-free when it encounters two so-called pump waves (blue and green) of different frequencies in a receiver on Earth. Thanks to the researchers’ noise-free amplifiers in the receiver, the signal is kept undisturbed and the reception on Earth becomes record-sensitive, which in turn paves the way for a more error-free and faster data transmission in space in the future. Credit: Chalmers University of Technology | Rasmus Larsson" (ScitechDaily, Silent Signals: The Breakthrough Technology Powering Faster Space Data)

Laser rays that can carry information are tools that make computing more secure. Laser rays can travel in computers inside so-called nanotubes. That protects them against outside observers. And non-controlled effects like dust that can disturb information. Laser-based communication between computers and microchips is a well-known tool. The microchip that internal architecture supports photonic data transmission is the tool, that can revolutionize binary computers. 

Binary computers remain on our tables for a long time. They are tools that can act as a medium between the quantum computer and the user. The user cannot use quantum computers directly using keyboards and the quantum computer cannot control things like screens directly without binary computer layers. 

Another thing is that those photonic computers don't disturb complicated quantum systems as much as electric computers. Binary computers are tools that control quantum computers. And that makes them important. The Quantum computer requires high-power binary computers and advanced artificial intelligence to control the qubits and other effects that can disturb qubits. 

"Large-scale optical programmable logic array for two-dimensional cellular automata. Credit: Wenkai Zhang (Huazhong University of Science and Technology), edited." (ScitechDaily, Light Takes Over: Breakthrough Optical Array Revolutionizes Computing)

"Demonstration and application of large-scale optical programmable logic array. (a) Nine-input PLA realized by combining wavelength and spatial dimensions. (b) Statement machine to infer the date (month and day) according to the provided day of the year based on 9-input PLA. (c) Oscillator (pulsar) in Conway’s Game of Life. (d) Nonisotropic evolution, in which a single-cell pattern can create the Sierpinski triangle by two-dimensional cellular automaton. Credit: W. Zhang, B. Wu, et al., doi 10.1117/1.APN.3.5.056007"(ScitechDaily, Light Takes Over: Breakthrough Optical Array Revolutionizes Computing)

The Van Allen plasma belt can be a good place for research or some military satellites. The plasma can protect the craft. If the ground-launched antisatellite, ASAT rocket travels to that satellite, the Van Allen belt can damage its homing system.  And the magnetic system can keep the plasma away from the satellite's body. In that kind of system, the magnetic field pushes plasma away from the satellite. The satellite or shuttle can collect information about plasma and they can also take images of other satellites. That plasma belt can offer a good hiding place for killer satellites. The problem is that plasma can disturb radio communication between a satellite and its ground station. 

But laser communication is a tool that things like plasma and radio bursts can't disturb. If the satellite is in the Van Allen plasma belt, it can communicate with other satellites and ground stations by using laser rays. Same way spacecraft can keep in contact with the ground station using laser communication. That laser communication makes it possible for the data flow can travel between the ground crew and spacecraft even if the spacecraft is in the ionosphere. That is very important for space drones like X-37B that operate autonomously. 

The high-speed, long-range data communication is the tool, that can help the successors of the Parker probe to travel deep into the sun's corona and research its plasma. Another place where that tool can be useful is the Jupiter missions where the craft can dive and research Jupiter's plasma belts. The probe can have two parts the plasma probe that communicates with the bigger probe using lasers. 

Coherent light offers high-speed communication between satellites and laser communication systems are less vulnerable to electronic countermeasures ECM systems. The problem is that dust and other things can disturb laser communication systems. 

https://scitechdaily.com/light-takes-over-breakthrough-optical-array-revolutionizes-computing/

https://scitechdaily.com/silent-signals-the-breakthrough-technology-powering-faster-space-data/

It's possible. That hackers can bypass the generative AI's security.

When man meets AI, the man is more intelligent. That means that the creative AI is not creative at all. The man who uses the AI is creative. The thing that we should understand is this: what we see as an example of "fuzzy logic" is a large number of precise reactions or precise logical points. 

That means that there are multiple words. Or otherways saying:  multiple triggers. That is connected with a certain operation. This allows users to use dialect words or literally while using the AI.  

The dialect words and literary words are both connected to certain operatios. That gives the user freedom in the form of language. What the person uses to command the system. 

In this model, users can use both, dialects and literary words. That makes the AI more flexible than if the user must use precise literal words. 

We can say many ways something that launches the action. The words, connected with certain operations are called logical points. Developers can connect each action with multiple logic points. Which makes the system seem like it uses fuzzy logic. 

That means that when we give a code or order encoded into hex-decimal or ASCII mode is not precise same as the orders that we write in regular text mode. This is the thing that might bypass the protection of generative AI. If somebody wants to use the generative AI as the tool, that makes it possible to write malicious code very effectively the person must just bypass the safety mechanisms of those generative AI. One version can be that the commands are given in the form of Hex or ASCII mode. 

Another thing that can cause the ability to bypass the security of artificial intelligence is to give orders as small pieces. The hacker- or otherwise the malware developer can use multiple AIs to create the code. And then that person can connect the results. The problem is this. The code writer must have deep knowledge of coding. And the other thing is this. Hackers can use AI to develop the base code as legal developers. 

They can freely use AI to develop legal software like chat programs and firewall software. The spyware that hackers use to steal information is the chat program or firewall that is modified to send data without the user's permission. So hackers can make legal applications, and then change the code so that it runs backward and steals information. One version is to create a tool that allows as an example teachers to follow what students do with computers. 

Normally those programs tell the user. That they are operating. Their use is limited to classrooms while teachers teach things like programming.  There can be some red frames and text. That tells the observation program is in use. Or the program requires user acceptance. Hackers can remove those manifests and then they can observe the targeted computers. 

https://cybersecuritynews.com/encoding-technique-jailbreaks-chatgpt-4o/

When requirements grow, programs grow.

The main problem with computing is when developers make something. There is needed new solutions almost immediately. Those new solutions make the application interesting. And many times. Those new solutions mean new abilities for applications. New skills require new libraries. And that thing cumulates the size of the program. 

The new libraries require more space and more effective computing. The thing is that AI is going to more detailed program code. When we think about the first AI "chatbots" the simple programs that asked people some questions, and then the program gave some pre-programmed answers, we must realize that those programs looked intelligent. 

One example of those programs is a program that asks: 

Are you a boy or a girl?

A: Boy

What's your name"?

A: John

How old are you?

A: 23

Where do you live?

Bristol

Then the program said: 

Please to meet you: 

You are a boy, whose name is John. You are 23 years old, and you live in Bristol. And I'm your servant. 

Those things are quite easy to program by using C, Java, or C++. But that program puts answers to the right points in the programmed text. And the form of the text doesn't require us to notice things like "she" or "he".  That simple code is an example of pseudo-intelligence. One "else" command makes it possible to create an answer that if the answer is something else, the computer answers that "you must answer "boy" or "girl". That makes those programs look intelligent. If the programmer has time, it's possible to add all British first names and all the world's city names to the list of allowed worlds. And that gives an appearance of intelligence. 

The thing is that when we think about the modern AI that can control things like cars, we must realize that in the middle is a large language model, LLM. The LLM itself does nothing. But it transfers the data to the part of the code. That controls the self-driving cars. That LLM stays at a tolerable size. But the libraries that the system needs grow. That makes it hard to control data in the system. 

The network-based architecture means that the system can have multiple, limited AIs or LLMs that are connected into the entirety. That thing makes the system more flexible. Because the system has many separate operating cores. It allows the developers to create those cells as modules. Every module is an independently operating part of the structure. 

Above: Mesh network. In network-based systems, every computer runs its own LLM and modules connected to it. So every computer involves certain action series. And, each computer is responsible for its part of the system's skills. 

The structure might look like some kind of mesh topology. That means that there is a central LLM. That is called the specialized LLM to make the mission. The system operates like a mesh topology. When as an example a robot car drives a robot to the shop it sends a mission to the next operator. 

That operator is the robot that goes into the shop. The mesh model means that the system must not bother the central actor all the time. That means that if the central processor or LLM is busy or the system is out of connection the sub-system can continue its mission without the central actor's support or control.  

When a user gives orders to the central LLM. That can transfer the mission to the sub-LLM that has data to complete the mission. The networked structure makes it possible that one mission doesn't keep the entire system busy. When one module operates with some mission. Other parts of the system can perform other duties. 

In modern neural networks, the middle of the system is the LLM. The LLM is connected with sub-LLM systems. The system looks like a neural network. When an owner calls, the car to the door. The central LLM gives a mission to the LLM that controls the car. The car's AI system can interact with surveillance cameras and the surveillance system tells if there are some dangers in that area. 

Or if there are no surveillance cameras the owner can send a drone that looks if there is something behind corners. The drone can hang over the vehicle and send the image to the computer where the car's corners are going. The car requires a 3D camera system and radar. That it can drive safely. 

And then it requires large-size databases that tell what the car should do in all situations that it can face on the road. There are lots of things that the system must avoid. When we think about the models of the complex systems we must create the systems using the cell- or neural-based networked architecture. For these reasons, I told you before.