|
|
kei9
from Argentina on 2023-12-22 02:41 [#02631564]
Points: 424 Status: Lurker
|
|
the problem is not about mathematical technique or complexity that is in place to evaluate functions. the problem is we cant even begin to understand what is the function (or set of functions) for the intuition that can formulate meaningful axioms or good functions. just as we cant synthesize pain or balance we cant synthesize intuition (No one can do this because no one knows how it is done. You can simulate the behaviour of a subject after feeling pain but you cant emulate pain itself. Just as you can make a robot that walks like a human but you cant make it have proprioception, or an intuitive feeling of gravity).
take newtonian gravity for example. No matter how good you know the system (matter) there is no description of gravity in any part of the system. To come up with that explanation a leap of imagination (induction) is needed to figure out theres something you cant see that its explaining the behavior. This is the kind of intuition you cant simulate. Regardless of how accurate or not newtonian gravity is, it is meaningful. The construction of meaning is another thing machine learning cant grasp at all. So you see the mind is not as simple as you fist thought.
in principle, this all could be boiled down to probability.but that would tell you nothing about what is going on in the mind when it comes up with a good induction. just as you could give 1 million monkeys a typewriter each and in an unlimited time frame maybe one will write goethes faust letter by letter, but that wouldnt make that monkey goethe.
so you cant synthesize induction, you can simulate its results (in principle). Just as you cant synthesize pain (these things happen in the mind and no one knows exactly how).
|
|
recycle
from Where is Phobiazero (Lincoln) (United States) on 2023-12-25 22:43 [#02631623]
Points: 39954 Status: Regular
|
|
1 point
|
|
Wolfslice
from Bay Area, CA (United States) on 2023-12-26 04:04 [#02631626]
Points: 4857 Status: Regular
|
|
Didn't Einstein just flat out write an equation for gravity that has not been diNewton. With regards to general relativity? Who needs newton.
I never believed this one million monkeys thing at all.
You'd have one million retarded scripts of monkey jibberish
|
|
Wolfslice
from Bay Area, CA (United States) on 2023-12-26 04:05 [#02631627]
Points: 4857 Status: Regular
|
|
Di Newton = disproven. What is auto correct doing!
Making me look like a monkey
|
|
RussellDust
on 2023-12-26 09:04 [#02631630]
Points: 16047 Status: Regular | Followup to Wolfslice: #02631626
|
|
“ I never believed this one million monkeys thing at all.
You'd have one million retarded scripts of monkey jibberish”
Ouch…. Worst post you’ve ever made.
|
|
RussellDust
on 2023-12-26 17:33 [#02631644]
Points: 16047 Status: Regular
|
|
Maybe it was tongue in cheek on your part sorry
|
|
Wolfslice
from Bay Area, CA (United States) on 2023-12-27 00:09 [#02631655]
Points: 4857 Status: Regular
|
|
Nah I'm actually serious.
If a trillion monkeys wrote for a trillion years, I believe none would actually pelt out the completed works of shakespere of gothe, or any other cogent work.
It would be a trillion years of monkey gibberish. You'd get a similar passage here and there, by chance. Never a complete novel.
Now if the monkeys were having sex and evolving over that period of time im sure they'd eventually become smart enough to write something similar, but the metaphor is about static conicidence and probabability.
|
|
Wolfslice
from Bay Area, CA (United States) on 2023-12-27 00:32 [#02631659]
Points: 4857 Status: Regular
|
|
I also don't really buy the quantum theories that there's a reality for every possibility.
Like, there's a reality where Leonardo DiCaprio, remaining, in this reality, a successful actor, married a dwarf covered in 3rd degree burns. Or one where the united states duly elects a cat for president.
There's something wrong with the theory itself. At a point, the odds are zero. I think the odds are zero with the monkeys. Forget a trillion years. I'd bet no monkey writes the complete work of shakespere with *infinite* time.
|
|
RussellDust
on 2023-12-27 07:46 [#02631669]
Points: 16047 Status: Regular
|
|
I know it as just a way to explain infinity and infinite odds. Of course a trillion years might have just given you a word by accident. ‘Surplus’ or something.
|
|
RussellDust
on 2023-12-27 07:47 [#02631670]
Points: 16047 Status: Regular
|
|
I’m surprised you didn’t mention the monkeys would die.
|
|
kei9
from Argentina on 2024-01-08 01:50 [#02631842]
Points: 424 Status: Lurker
|
|
hey guys sorry for the hiatus
this rant is something I worte about the limits of AI
the newton theory is quoted as an example of a theory that can make very accurate predictions
its axioms cannot be deduced from the objects it acts upon as it is not any of the objects in the system but an induction of how the objects in the system interact with one another. its ultimate nature of what explains this interactions is forever beyond our grasp. the theory itself only exists as a meaningful explanation because it is instrumental for a subject as an abstraction of its intuition of time and space.
this intuition of time and space is meaningful for a subject as a way of attending to the necessities of a body which AIs are oblivious about. there cant be no subject without suffering
|
|
Tony Danza
from NAFO Suicide Hotline on 2024-01-08 01:57 [#02631843]
Points: 3623 Status: Lurker
|
|
Yeah I'm inclined to agree that development of real consciousness requires "skin in the game" at least in your ancestors.
Start torturing those neural networks, researchers!
|
|
Messageboard index
|