Artificial Intelligence 2020 - is there progress? - page 18

 
Maksim Neimerik:

I'll distract you a little from the controversy:) In previous comments somewhere in this thread there was a discussion about the 'Boltzmannian brain'... I read a bit and came across this suggestion:"According to the infinite monkey theorem, an abstract monkey that hits the typewriter keys indefinitely will end up typing a given text at some point / © Wikipedia/New York Zoological Society".

I'm really interested in this, so I just sketched out a script for randomly typing nth number of lines with nth number of characters... The result hints that this is a very, very implausible assumption:) What do you think about it?

you've got a monkey emulator at your typewriter :-)

The letters are arranged so that they are easy to type - figuratively, the most used toward the centre, the rare ones around the edges. So, typewriter partially takes into account statistical features and monkey will type two or three comprehensible words quickly enough.

But your program does not ;-)

 
Maksim Neimerik:

I'll distract you a little from the controversy:) In previous comments somewhere in this thread there was some talk about "Boltzmann's brain"... I read a bit and came across this suggestion:"According to the infinite monkey theorem, an abstract monkey that hits the typewriter keys indefinitely will end up typing a given text at some point / © Wikipedia/New York Zoological Society".

I'm really interested in this, so I just sketched out a script for randomly typing nth number of lines with nth number of characters... The result hints that this is a very, very implausible assumption:) Who thinks about it?

I will say in general: the monkey really has a better chance of typing "War and Peace" because the time required for its evolution into a human is much less than the time required for a similar script, which never evolves to work in an AI. So by typing, the monkey is likely to get progressively smarter and there is a chance that in a few million years it will repeat the classic.))

"Chattering" of parameters (letters) as a driving creative force developing conceptually complex systems (works) should not be considered at all, and no optimization with GA will help. Only reason can create "ideologically based" objects/concepts. imho.

ZS. in terms of probability, a monkey can type a novel the first time.))

 
Valeriy Yastremskiy:

Or don't pay attention right away. Only later, after the next action. This method works for me sometimes positively.

Yeah, said Valeri, who's the one who originally started the fire. :)

Here is another property of non-artificial intelligence. Contradictions within one discussion. Here too, it is worth pondering whether such contradictions are normal or the non-artificial intellect does not belong to itself. It is simply allowed to think so. :)

 
Реter Konow:

Let me generalise: a monkey really has a better chance of typing War and Peace because the time required for it to evolve into a human is much less than the time required for a similar script to work, which never evolves behind the work in AI. Therefore, a typing monkey is likely to get progressively smarter and there is a chance that in a few million years it will repeat the classic.))

"Chattering" of parameters (letters) as a driving creative force developing conceptually complex systems (works) should not be considered at all, and no optimization with GA will help. Only reason can create "ideologically based" objects/concepts. imho.

ZS. in terms of probability, a monkey can type a novel the first time.))

why would it get smart in the absence of feedback ? in the absence of feedback, learning is impossible.

 
Реter Konow:


ZS. in terms of probability, a monkey can type a novel the first time.))

Can with the lowest possible probability. And the number is too small, and if the reverse is true, too big to write it. And the problem is memory storage devices))))))

In our lifetime it won't print for sure with 100 per cent probability.

 
Maxim Kuznetsov:

Why would she get smart in the absence of feedback? In the absence of feedback, learning is impossible.

Well, right. Only the feedback (I mean target function) can appear in the process of fruitless work, and then her overshooting of the letters will start to be filtered by it, which will increase its speed considerably. Gradually, some conditionally-reflexive connections with certain combinations of symbols will mature, and further on - her learning will increase.))

 
Реter Konow:

Well, right. Only the feedback (I mean the target function) may appear in the process of fruitless labour and then her overshooting of letters will start to be filtered by her, which will significantly increase its speed. Gradually, some conditionally-reflexive connections with certain combinations of symbols will mature and then its learning will increase).

If the monkey in the process wises up by itself, it will stop aimlessly banging on the buttons.... -)

 
Реter Konow:

Well, right. Only the feedback (I mean the target function) may appear in the process of fruitless labour and then her overshooting of letters will start to be filtered by her, which will significantly increase its speed. Gradually, some conditionally-reflexive connections with certain combinations of symbols will mature and further on - her learning will increase.))

You're talking more about the evolution of the monkey than the probability of typing... As far as I understood that article, it was about unconscious typing...

 
Vitaliy Maznev:

Well yes, said Valery, who himself initially added fuel to the fire. :)

Here's another property of non-artificial intelligence. Contradictions within the same discussion. Here too, it is worth pondering whether such contradictions are normal or the non-artificial intellect does not belong to itself. It is simply allowed to think so. :)

Apophenia is peculiar and not alien to man, in addition the day of the medic)))) and in the evening will be even worse))))

So that there is no contradiction, there must be a definition of the terms, concepts and words and a clear understanding of them.

Here (in a developing field) it is unlikely, but communication is more pleasant than....))))

And regulation would at least prohibit communication in a not very operational state in order to speed up the result))))

 
Maxim Kuznetsov:

if the monkey gets wise to himself in the process, he will stop banging aimlessly on the buttons.... goodbye great novel :-)

Exactly!)))