L
L
lemonlimelike2017-08-01 01:23:06
Malware
lemonlimelike, 2017-08-01 01:23:06

Why hasn't anyone made a self-evolving virus yet?

That's how many viruses have been created, but I don't know of any that would be uploaded to the Internet and develop on their own, such as how a neural network is taught, only to teach the virus basic knowledge, such as to study or something like that. Why hasn't anyone written this yet? Like in the movie Transcendence

Answer the question

In order to leave comments, you need to log in

7 answer(s)
X
xmoonlight, 2017-08-01
@xmoonlight

Why hasn't anyone written this yet?
Name the source of this information and prove that this is true.
951e387ed6644e67b2761e94c10f800b.jpg

L
lukoie, 2017-08-01
@lukoie

Apparently because a piece of virus code needs a big server for the development and accumulation of such "knowledge".

I
Impeeeery, 2017-08-01
@Impeeeery

to write a good virus, you have to be a genius of analytical thinking.
to write a good AI - a synthetic genius and ... even xs what, I'm not like that.
it turns out: "I carry one bomb with me and fly calmly"
that's why they won't write

Y
Yuri Chudnovsky, 2017-08-01
@Frankenstine

Why hasn't anyone written this yet?

Because it is technically impossible at this level of development of computer technology
Paris_Tuileries_Garden_Facepalm_statue.j

D
dmfun, 2017-08-01
@dmfun

It is necessary that the AI ​​be very short and fit into one or two files (otherwise it will not be downloaded via the Internet).
Most likely this is the main reason (if it was written by a sick person who does not think about money).
Now a clumsy game takes a hundred megabytes, contains a bunch of frameworks, third-party libraries.
Well, or for AI there must be a powerful botnet network in which its base is located.
Therefore, viruses must be short.
Well, the trick is that there is no profit from this. The goal of viruses is monetization. Spam, mine bitcoins, make a computer part of a bot network, steal data, extortion (winlockers, cryptors).

E
Exploding, 2017-08-04
@Exploding

One of the most important problems in the development of AI is the primitiveness or, one might even say, the lack of technologies suitable for this task in principle! They haven't invented it yet, and I think, oh, how not soon, if at all, they will invent such a programming language that could describe and at least get a little closer to such a level of human thought processes. Existing programming languages, despite their diversity (and I think it is their "diversity" and number speaks for itself!), level of complexity, etc. - for this task are extremely primitive due to their linearity. Modern "polymorphism" is only a variation of "imitation of thinking", but not a manifestation of intelligence! All processes are predetermined and just as linear...
Yes, this is not the main reason, in fact!)) There are only problems with the definition of the concept of AI with a cart and a cart, not to mention the rest ... Yes, this is probably both good and logical.
And maybe a button accordion, here there is still information for thought , I personally only have a couple of months. I accidentally saw it back, but in general there are also similar videos ...
I would have to figure it out with myself first)))
Such is ...

@
@prografika, 2017-09-03
_

Yes, there are plenty of them, I read them all the time. It’s just that for every action there is a reaction, there is always a way to catch the infection, sometimes you have to dig for a long time, and sometimes you can immediately see where something is sitting.

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question