AI-controlled self-replicating machines are coming.

     AI-controlled self-replicating machines are coming.

 AI-controlled self-replicating machines are coming. And that gives the drone swarms the ultimate abilities. In the simplest versions, robot factories make the drones. Then that system loads the operating system in those drones. Those drones can have "iron-based" AI, which means the AI software installed in the kernels. The same system that creates microchips can install AI-based kernels in those microchips. 

Wandering a drone swarm can be like moving a supercomputer. The drone swarms can act as giant neural networks. And that makes them effective tools for research and military work. Drone swarms that reseaches distant planets can cover large areas. And if that swarm loses one member, that thing has no effect for its entirety. 

But what if those drones can make copies of their physical body? That can cause a situation in which drones can cover the entire planet. The name of the self-replicating machine is the "Von Neumann" machine. The Von Neumann machine can be virtual. And the best example of those machines is computer viruses that can make copies of themselves. But also physical, AI-controlled machines can make copies of themselves. 

The Von Neumann machine and AI-based neural network are the tools that might investigate another solar system. In some visions, the craft that travels to other solar systems is the shell which is a man-looking robot. Those robots might be the brains of the Von Neumann machine, the AI-controlled, self-replicating machines. The Von Neumann machine or self-replicating machines are robots that can make copies of themselves. 

The self-replicating robot controlled by AI is a tool that can lose control. The robot that is AI-controlled is a server and software in physical form. That system can have sensors that allow it to find minerals to make microchips and machine parts straight from ore. The robots can cut those minerals by using lasers and then make needed stuff in the high-tech reactors and they might have integrated 3D printers. 

Those systems are more multipurpose than some casts. Robots can operate in a vacuum chamber which makes their work good. So those robots that form the "brain of the system" might have integrated 3D printer that allows them to fix themselves. And they might have the ability to replicate themselves. 

 If robots can communicate with each other, they can form an extremely large neural network that acts like a peripatetic supercomputer. But there is a difference between Von Neumann machines and regular computers. Von Neumann's machine can make copies of itself. And that spontaneously increases the power and the size of the neural network and its actors. The perfect tool can be a perfect threat. 

The Von Neumann machine can be physical or virtual. Computer viruses are the example of the non-physical Von Neumann machines. When we connect things like Chat GPT or any other AI with robots there is the possibility that there happens something "unpredictable" in the program code of the systems. When the AI develops itself it can search data about the abilities that it wants. 

Self-developing AI that is connected with computer games can search the tactics and strategies that give it the best abilities. So if we transfer that model to robotics the AI can select the individuals that give the best results and then connect those abilities with another robot. The robots might made of different materials. And they might have different codes in computers. 

That gives the AI the ability, to search for the best combination for each situation. And if AI faces another AI it can ask needed code for that other AI. And then it can offer some of its code to another AI. 

When we think about AI and its abilities there is a possibility that AI can also exchange program codes with each other. That allows the autonomous R&D process. And that kind of process is dangerous. The AI itself is dangerous because it can accidentally or on purpose create computer viruses that can damage ICT infrastructure. 

The AI can make only things that we allow it to make. Programmers remove dangerous components from public AIs. But the difference between public, or civil AI, and militarized AI is that militarized AI is made for military purposes. Militarized AI is created to make viruses and make service denial attacks against the enemy command infrastructure and drone swarms. 

The thing is that AI is not intelligent. It just connects code from different sources. Things like commercial AI equipped with a system that denies the system to make computer viruses or malicious software. But militarized AI might have abilities to make those viruses. The purpose of those viruses is to deny drone swarms and computer-based command systems to operate. The hackers can steal that code from destroyed drones. 

Image: https://www.homelandsecuritynewswire.com/dr20120420-autonomous-selfreplicating-robots-for-finding-extraterrestrials