The generative AI phenomenon.
When a machine learns something, it interconnects data from different sources into a new entirety. The idea is that the data is stored in the system as the bites or data units. They are like boxes. When the system stores new data it stores it in the new box. So when the AI creates a new entirety or new skill from those data bites the process looks like the system plays Tetris.
When we think about the models and accuracy of the data bite or data unit, the large number of small data units is the thing that makes the AI accurate. The reason why human memory consists of billions of data units is simple. If one of those data units is damaged or corrupted a large number of small data units guarantees. That damage in one data unit does not have a big effect on the entirety and results.
The animation above introduces how the machine learns things. Every box in this Tetris animation is a data unit. And the system connects data units into new entirety when it learns something. Data that the system uses. Can be collected from sensors and memory units from mass memories.
Because modern generative AI uses open datasets, it can make almost everything. The open dataset means that the AI uses the open Internet as the tool where it searches for information.
Generative AI turns society. It turns many things like productive work to a new level. And because the generative AI is a tool that creates things like computer programs in minutes. Programming is quite mechanical. And by using generative AI the programmer can make new and complicated tools in minutes.
The generative AI can make thousands of lines of error-free program code in a very short time. And then the programmer must just fix things like paths if the AI doesn't make those lines. That means free versions of the generative AI are tools that can be used for this purpose. The next generation of productive AI is the complex code system that can make multiple things. Generative AI can be used as a tool for searching sources.
And even if the text that the AI creates cannot used as an essay, the sources that are below the generated text can be used as sources for new texts. Generative AI can give keywords for writer's texts. When AI works in a limited sector where everything happens by following certain processes and laws, it does a good job. But the things like emotions that make us humans make it very hard to predict our behavior.
Humans can make something unexpected. But things like stars are almost fully predictable. Only things like some energy bursts can disturb the star's life. This is why we can almost predict everything that happens in Betelgeuze. And only if things like powerful FRB (Fast Radio Burst) that comes from the other side of that giant star can cause the unpredicted change in its energy production.
The generative AI-based chatbot can connected to speech-to-text applications. And that allows to give spoken commands to those systems. This kind of combination where generative AI is connected with the physical robot and speech-to-text applications makes the system even more powerful. These kinds of applications are excellent tools for military robots.
But things like EMP weapons are also developed against those robots. We must realize that this technology is under development around the world. The other thing is that we must realize that this kind of technology is not allowing only good things. And the Ukraine war has given a lesson that new tools like robot soldiers are effective actors. And that will begin a new weapon race where robot weapons and counter-weapons against robot weapons are under development.
The AI development must be open. That means we should realize that people who develop AI and robotics should dare to tell if somebody asks them to make weapons. Another thing is that authoritarian governments are not the only actors who want to use AI for negative purposes. Hackers and organized crime are also interested in generative AI and its ability to create things, like computer viruses and spy tools. But also those non-govenmental hostile actors are interested in using robots as weapons and drug carriers. Quadcopters are good tools for delivering pizzas to some places. But the same system can search for things like guns and other interesting things from homes. The quadcopters can also be used to drop drugs and even guns to prison yards.
Generative AI is the only ChatBot that can make almost everything. The generative AI is a powerful tool in programming. Those multipurpose chatbots can used as tools that can create operating systems for robots. Generative AI can used to program robots. Physical robots can connected with generative AI. And that allows them to change the programs inside those systems very fast. This means that the AI chatbot. That connection with the physical robot can make it an even more multipurpose tool than it already is. Things like cleaner robots can make many things if they are programmed to make those things.
And we must realize that AI-controlled robots are not toys. If somebody thinks that the AI-controlled robot is a good opponent in some splatter war or some action game, that person should understand that those robots are extremely dangerous in the wrong hands. Those physical game robots can turn into perfect assassination tools.
We must understand that all actors. Those who are interested in robots as soldiers are not governmental. Organized crime is the actor that can think of the robots as hitmen, who will never talk. And of course, criminals want to use pizza drones as drug carriers. The pizza drones are doing deadly work in Ukraine, where they carry grenades over the enemy positions. And that thing causes the need to re-estimate the security in buildings.
https://scitechdaily.com/beyond-human-mit-experts-explain-generative-ai-and-the-chatgpt-phenomenon/
Comments
Post a Comment