Eureka Machine Discovers Newton's Laws in 2 Hours

It probably still conforms to set routines however. In my theory for example gravity is a push, and not an attraction. I doubt that the programmers allowed it this freedom, and most likely only allowed it to conform to Newton, and Einstein. Now my theory will alter what happens to the formula of protein X when enzyme A increases, as I have an outer, and inner pressure to work with, but the standard model only has the inner pressures. The same with Dark Matter through Galaxies, I have both pressures to work with, the standard model only allows sticky attraction at best. And another example.. is it allowed to use negative mass?

 

Log in or Sign up to hide all adverts.

You're showing you don't even know what they did, even on a conceptual level. They gave the algorithm a basic mathematical vocabulary like "add" or "divide" or trig functions or polynomials and allowed it to form ANY combination of those. The genetic algorithm then searches the infinite space of all possible combinations of such algebraic objects looking for the ones which best describe the data. Newton and Einstein have nothing to do with it.

You're attempting to understand something clearly beyond your comprehension. You're so wrapped up in your delusional little world where you're all knowing with all the answers you seem to have lost the ability to read. You claim to understand neural networks etc so the work Cornell should be well within your grasp yet not only have you not understand the algorithm you haven't even understand what the concepts involved are.

Matter, Einstein, Newton, pressure, mass, none of those have anything to do with the algorithm and the methodology. Unfortunately you're too detached from reality to understand that.

 

Log in or Sign up to hide all adverts.

You have stated the obvious, and missed the point. In fact what you have said is like a 12 year old schoolboy's recount of what he saw. You are my proof that this will not work, because you have the same basically, faulty functionality. Everything is pressure, and by ignoring that you get nowhere. But to be fair, I'm sure that the programmers know when to add -m.

You don't know the difference between these do you?...

F=G*m1m2/r^2

-F=G*-m1-m2/r^2

 

Log in or Sign up to hide all adverts.

You are the one missing the point. The programmers don't out -m or +m in at all. They give the program a set of basic operations, +, - , / , * , a set of basic functions, \(\sin\) , \(\cos\), \(x^{n}\), \(\sqrt[n]{x}\) and the program itself decides what to use. If the mass of the pendulum is 5 then the program will consider both +5 and -5, searching through the space of equations which allows BOTH possibilities. The program, theoretically, can consider ANY possibility.

You're not even understanding the concept here. You're complaining "Oh the programmers might have just put in Newton and Einstein" and "What about -m". The whole point is that the program isn't restricted AT ALL, all it has is the data and it is left to find the most appropriate description.

I hardly think you're in a position to be trying "I understand this more than you". Finished wasting your time designing a robot in CAD yet?

 

Pincho Paxton's point more or less is:

1.5x / (2/3) * 8 might be less physically representative than 18x and is not as deconstruct-able as composed equations.

 

Well that is part of it, but here are 4 examples....

F1=G*m1m2/r^2

-F1=G*-m1-m2/r^2

F2=G*-m1-m2/r^2

-F2=G*m1m2/r^2

When the program can produce all 4 of those is anyone going to edit them? If they edit them they are reproducing Einstein. But I can draw a diagram of each one, and each diagram is different. The current one you are using..

F1=G*m1m2/r^2

... is for Dark Matter lensing. It's not even the right one for G force. So will the program pick the right one, and will anyone edit it back to the original form?...

F=G*m1m2/r^2

 

That's not the point of it; no one is going to accept a monstrous equation which describes motion as a physical solution, however it can present certain characteristics of what the solution should look like and find patterns indicative of known phenomena.

 

Someone will try it out on Galaxy clusters, it's just a matter of time, and then you will get these odd results. (Which aren't odd to me).

 

Firstly you're showing how terrible your maths skills are in that you don't realise they aren't all the same. You do realise that -1*-1 = +1, right? Perhaps you don't.

As such two of them cannot be output by the program because it'll reject them when it considers them because they predict objects repel one another. When you drop a pencil does it float up to the ceiling? Maybe it does in your weird little world, you don't seem particularly interested in reality or truth.

I really do wonder how you get by in the real world. Do you tell people you supposedly work with that you're a physics genius who understands the secrets of the universe? I can't imagine many people wanting to work along side such a person. If you tell people you think you've got all this insight do they ever ask you why you've accomplished nothing? What do you answer?

You've shown in just the last handful of posts you can't even grasp basic concepts and reasoning. You're also more than willing to utterly misrepresent actual science.

 

Actually, your version is the repelling version....
F=G*m1m2/r^2

I could draw you a diagram of each one and prove that your version repels. But for some reason you allow mass to attract to get around this, but by using standard physics your version repels, and attraction is never visible, so I don't count that as normal physics. But my version...

F=G*-m1-m2/r^2

In standard physics two holes never repel a solid. and your biggest attractive force of all you call a Black HOLE.

 

I'll be banned anyway because you think that maths is the only proof, but anyway here is my proof by using full scale physics as opposed to imaginary invisible physics like attraction. So you see footballs, and a custard powder type substance (Gravity is like a bumping liquid), not invisible waves...

Please Register or Log in to view the hidden image!

As you can see, even though some of the results are identical, you are moving the observer around.

 

Anyway this thread is about a camera changing physics into maths. So try a camera on a Galaxy and you will get visible physics like mine, the computer will not invent invisible attraction. You will get my formula with -m, because there is no visible attractive force, only footballs, and water etc.

Please Register or Log in to view the hidden image!

 

give it space travel or HIV or cancer if it could find the cures for thoes how awesome would that be

 

Yes, it moved all threads to Cesspool, now Sciforums moderators are out of a job.

 

In Essence it's a Physics "Bruteforce" method using a Rainbow array filled with "fundamental laws that are empirically proven" or simply a "Crackers" methods to Physics.

 

It produces equations that best fit results of experiments, testing and refining them, matching them against results given to it, rejecting or adapting those that don't work.
It could not find an equation without data being presented to it.

It works along Darwinistic lines.
Watch the video, it is interesting.

I don't think it has found anything new, but it might do, given enough time and a powerful enough computer..


Regarding the shape of galaxies.
Good question.
If you gave it that data, I wonder whether the equations it came up with would infer dark matter, or something else.

 

I just noticed that the video is from 2009. That's quite old.. this is an old thread. My main point is that the maths will be back to front, and electrons will be given a mass. If the program is allowed to store visual data like in the video it will work. If it is given electron mass it will fail. Somebody needs to put a minus sign in front of the electron mass (and any other negative particles) to get a realistic result.