Bye bye algorithms.

 

We are waiting for the new step in the AI development and research process. Many big technology bosses say that this is the end of algorithms. And the next step is self-learning AI. That system can communicate with robots and all other systems. The self-learning system can learn in two ways. It can create new models that it uses in certain situations. Or it can connect a new module into itself. The reason why advances in AI go to self-learning systems is simple. 

New algorithms are very complicated. And their training requires so much time that the self-learning models are better. The thing that makes this kind of thing very complicated is that the new AI must operate in larger areas. They must control things like street-operating robots. So they need a more effective way to learn things. The street-operating robots can use platforms that look like computer games to learn how to cross the roads, and where those robots find things like apples if they go to shop for their owner. But then those robots must face unexpected things. 

Robots can share their mission records with the entire system. And that helps to develop methods on how to operate in natural situations. Basically, the difference between a learning system and a normal system is that the learning system can create new models and then compare the original model with that new model. There are parameters that determine which way to act is better. And if the new model is better, it replaces the old one. This means that the fixed model turns into a flexible model. That model lives with its environment. 

That thing is the AGI or artificial general intelligence. That kind of AI is everywhere, and it can connect multiple different systems that seem different under one dome. The biggest difference between AI and modern algorithms is that the system can bring new data from sensors to data flow that travels in the system. The AGI is a system that might be "god-like" but if that system cannot create genetic codes like manufacture the DNA it might have no ability to control living organisms. However, the system has many ways to manipulate evolution. 

The AGI can make couples that have certain skills. The fact is that dating applications are effective for dating. And it's possible that the AGI can also make it possible to select perfect spouses. So people who are not "perfect" leave without a couple. And that means only people who are suitable, or similar can make descendants. This causes segregation and loss of diversity. 

And that is a sad thing for humans. Self-learning AI is a tool that can learn from its mistakes. It learns what to do, and what it must not do. The thing is that the self-learning AI is the new common tool that can make almost everything. The system learns like humans. And that makes it the so-called AGI. One tool fits all. The system can control things like robots.

Robots can collect data for that system. The AGI works like this. One robot sits on a chair and then the teacher teaches things for, and through that thing. Then that robot shares new things across the entire AI and network. For training that kind of system, a lot of information. And companies like Meta have that data. And AI makes it possible to create things like AI agents that sneak and observe what happens in the network. Robots can learn from other robots. When one robot makes a mistake, it scales over the network. Other robots must know that they don’t make the same mistake again. 

Use of biometric data for making things more secure.

Use of biometric data for making things more secure.

Biometric data is a very good way to identify the users of some devices. But the question is what is the dark side of that kind of data? The thing is that the biometric data can be collected from the devices, and compare it with the databases. And things like fingerprints are coming to the identify method also in the private sector.  That means that also the private corporations can start to use the recognition method, which has been reserved for governmental use, and that kind of thing is causing discussions in the public media.

Fingerprints are a really good way to make tools secure. That thing is not able to fake. And also biometric data tells is there some other person, who is handling the device, which uses biometric recognition of users. That system can compare the fingerprints of people who are using the mobile devices of some firm, and that thing tells, who is borrowing the telephones of other users. In this case, the computers of the company require biometric recognition by using the mobile telephone of the firm, if they want to use the data, what is confidential.

Biometrics is a very effective tool for making things safe. And when things like fingerprints are used to open things like the mobile telephone, there is a possibility that if the telephone has been reported to be stolen or missing the system will send that data to the authorities, which will track the mobile device and arrest that person. But there is a possibility to use the fingerprints, what are stored in the system for tracking people who have made illegal actions.

This kind of thing is a really interesting point of view when we are thinking of the possibility to make the Internet more secure. The thing is that data is ever totally free, and there is no absolute freedom in the net anyway, and that means we must understand that some kind of control is necessary. If things like mobile devices use top security like a biometric key, that means that the fingerprints of the user can send to the Internet, and they can be compiled with fingerprints, which are taken from the crime scene, without telling that thing to the person or police chief. This is the thing that people must understand.

Authorities are not always telling things, what is true, even if they are under democratic control. The thing is that the voice recognition can also be used to track people who are calling the disturbing phone calls. If we would connect this thing to profile, the authorities can focus the surveillance on the people, who are denying permission to use their biometric data to solving crimes.

This means that authorities might ask permission to use that data, and if the person denies that thing, this tells that the user of the mobile device has something to hide. Also if the person writes texts, what are supporting crimes, that thing tells that there is something to hide with those things or there are persons with suspected connections to criminals. And this makes this kind of thing very useful to target the surveillance to those people, who are more suspicious than others.

Image:

https://www.cpomagazine.com/wp-content/uploads/2019/02/landmark-ruling-on-biometric-data-gives-citizens-the-right-to-sue-tech-giants_1500.jpg