 |
| Image: Artist's impression of the programmer during his work |
The "open source" AI-development project means that the developers use free software for their development process. The database connections and programming codes can get from the internet. And that means there is nobody who controls that development process. The world is full of artificial intelligence programs.
Some of them are "simple" discussion programs. And the other programs can control robots. But the fact is that there can be millions of developers in the world. They have their own private "free time" application development in progress. The thing is that every programmer in the world uses their personally-owned computers in that kind of project.
This means their own companies or workplaces wound not have a chance to make anything for the things. What those persons do in their part-time. Programming tools like editors are free and there is lots of literature about programming and development.
Artificial intelligence is a problematic tool. Its missions are hard to predict. And that means especially when artificial intelligence can collect data from the internet straight from the homepages that mean that system might have unpredictable abilities.
But when we are thinking about artificial intelligence from the point of view of the military. The unpredictable actions are increasing the power of that system. The unpredictable actions guarantee the success of the operations. But when we are thinking about the classic questions like can AI make crimes?
We must say that AI doesn't make crimes for money like human criminals. There are always some people behind the criminal AI. The AI is an effective tool in the cases like hacking. The hacking algorithm is quite easy to make. And the hacker who wants to create the automatic, AI-based hacking algorithm must only record the commands that this person uses in hacking operations.
So the problem with the predictions of the abilities of the AI. Is that they are made for the cases where the algorithm is running on the binary computer. Quantum computers are a million times more powerful than the fastest possible binary computers. And the number and capacity of those computers are increasing all the time.
When we are thinking about the possibility to make a digital copy of human brains by creating 200 billion databases. The only thing that denies us from making that thing is time. There is needed a lot of time for making that number of databases. In the time of binary computers, that kind of system seemed science fiction.
But in the time of quantum computers, the quantum system can run that kind of complex database structure. And the ability to make new databases and establish connections between those databases automatically makes it possible that the system can learn spontaneously. And that kind of system is almost like humans.
The key element in artificial intelligence is the internet. That allows making the systems in many positions at the same time. The net makes it possible to control large project teams which are working for a common purpose.
Artificial intelligence requires complicated programming code. And the internet is the best possible platform for that kind of R&D operation. In that case, there is the possibility that somebody steals the code. When we are thinking about programming code anybody can learn to program. The code itself is free.
Computer programming languages are standards, and programming instructions are free to read from the net. And that means there is the possibility that somebody has the undercover "open source" artificial intelligence project.
https://www.sciencealert.com/calculations-suggest-it-ll-be-impossible-to-control-a-super-intelligent-ai
Comments
Post a Comment