Answer the question
In order to leave comments, you need to log in
Is it possible to patent the inner workings of artificial intelligence?
Without going into details, it is a special combination of decision algorithms and a neural network. Can it be patented?
And if so, what would that mean exactly? Will other people be able to use the general (slightly modified) idea or some separate (key) elements of such artificial intelligence? Does it make sense in this case to patent it at all?
Answer the question
In order to leave comments, you need to log in
No.
Specific solutions are patented, not principles.
But let's say something changes and it will be possible...
1. They will be able to use exactly your own model. But you can, if you prove patent infringement, get something.
2. Separate elements - what is it? If in general - the use of individual elements is quite legal.
3. No, it doesn't.
Do you seriously believe that you have some kind of super-genius idea, and if you don’t patent it, then a bunch of companies will immediately make money on it?
It's well described here.
It must be remembered: any disclosure of information (when patenting, for example) reveals ALL logic.
Knowing the logic, it is easy to bypass the violation of the rights to a patented invention and create a better analogue, but with a different level of legal protection (for example, any IT giant).
Another option: "refused, because an application with a similar principle of another applicant is under consideration."
And here it becomes completely incomprehensible:
1. they took your algorithm and remake it for themselves,
2. or simply "closed" you access for implementation on the side,
3. or is it, after all, really true?
PS: The only thing that is not clear is how to track the fact that your method is not used on backends?
try to consider the neural network as a database, there was something about databases in the law.
except for everything that is said above - any patent is tested for purity. none of its important components (algorithms, ideas) have been used in any way before. if the examination shows that some kind of technology/algorithm has already been used before (it doesn’t matter in free or proprietary form), the patent will most likely simply not pass.
the subject of a patent can even be an idea/ideology (device description) without implementation. but expertise for innovation of an idea/ideology/algorithm is an indispensable part of being able to recognize a patent
ps there are many nuances. and for details, in fact, to lawyers. if memory serves. an idea/technology may not be new, provided that it is applied to solving a problem of a fundamentally new nature. and not as previously used
ppsAnd it's not free either ;)
Suppose that some country will have a technical solution that no one can repeat.
Neighboring countries will immediately present a claim to such a country that its military-technical potential threatens peace in the region. And preparations for war begin.
Therefore, the invention cannot be hidden from people close to you.
Patents also exist so that the manufacturer is protected from unfair competition. Before investing in the purchase of technology, the manufacturing company always agrees with competitors that they will not copy this solution until the technology pays off. However, someone may want to participate in this project.
The question is impossible to answer. Because today there is not even a definition of artificial intelligence. Of course, we can fall into primitivism and say that they say this means a copy of the human brain or something like that. Only science fiction writers have confidence in this matter. Modern science does not yet know how the brain works. Soul or spirit. And so on. Therefore, it is too early to talk about patenting. Or you can patent a million different technical fake devices that simulate the brain and no one has the competence to check.
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question