@TrevorGoodchild
I see, that you guys are still too retarded to understand the concept of "I don't fucking know, I am not better then the best of our physicists combined"
@JoshuaSlocum @TrevorGoodchild @Jewpacabra @LukeAlmighty These niggas can't even figure out matter. When the math breaks down they blame "dark matter", a substance they've never seen and shouldn't exist, because their gravity silly string equation says it's there.
the maths working out to galactic scale means the math is pretty good; newtonian gravity failed at the solar system scale
doesn't really have anything to do with string theory per se
but it's hard to explain and it sounds retarded, agreed
We haven't yet figured out what we are missing in the current models but there's obviously something not right. Plugging in an imaginary substance makes the math work, but anyone who doesn't depend on his grant money requiring it is pretty sure that it's just a shorthand for "we don't know yet."
the mofos who named them so similarly prob regret it now
also, you haven't heard? "the aether" is still kicking it on web sites that look like they came from 1997
God told us to take dominion over the earth, and that's what we're going to do. It just takes a while.
It will figure physics so advanced we barely understand it, if we ever do. Like explaining quantum gravity at a five year old's level, only we will always be five years old.
We might still be able to use it and travel to the stars. But we would never truly understand the technology.
AI-generated science will turn out the equivalent of the AI drawing of a woman with big tits and 3 hands.
With the deluge of Indians using AI too generate training content for AI, we will shortly reach the point where AI is widely seen as the joke that it is.
@Snidely_Whiplash @WilhelmIII @JoshuaSlocum @petra
Well, that is kinda wrong.
The AI does have a model literally based on scientific method. (Observe, create a theory, measure, assess difference and alter theory...)
Therefore, it does come up with its own descriptions of what it does observe. We didn't teach it what a car is, we only showed it 100 000 000 pictures of cars and not cars.
So, I believe, that physics might be the best problem for AI to solve, since we are working on it in the exact same way. All it needs is its own interface to observe. We don't need it to create a new physics, but to analyze what physics is.
@Snidely_Whiplash @WilhelmIII @JoshuaSlocum @petra
What do you mean that AIs cannot observe?
literally all additional data are observation.
And the "set of lines in relations to each other" is one kind of an abstraction for a car.
I get your criticism, but the issue with AI is, that it has no concept of itself, and therefore it cannot has its own goals. Literally all learning is from a human set utility function. And that is a fundamental problem, that will not be solved by just more nodes. But to say, that it cannot abstract ideas is just wrong.
Observation requires conciousness. AIs do nit have conciousness and cannot observe.
The issue is not the concept of self. The issue is the lack of concept AT ALL. AI cannot conceptualize.
And no, AI cannot abstract concepts from the training set. This statement is entirely 1000% correct.
AIs do not even have the idea of truth vs falsity. It is a category error to believe they do.
People are overwhelmed by the AI's ability to compose English sentences, which is used by the programner to MIMIC intelligence.
The whole thing will come crashing down soon enough.
This is why it's so terrible with hands. I has no abstract idea of what a hand it, what it does, what it looks like. It can only concatenate several millions of images tagged with the word "hand" and calculate an average of what lines, shapes and coloration correspond to that tag. But because hands are so mobile, fluid and expressive, the examples do not do much to constrain the image generator.
Likewise, it has no idea of what a "car" is. It has a list of compositional elements, shapes, lines, curves, etc that correspond to the tag "car". The image generator has no concept of what a car is, or even what a car looks like, because concept is itself outside the programming.
This is a category error.
And AIs can't observe.