Is computer science the next physics?

[CyberShot]

Can anyone out there entertain the idea of the likelihood that sometime in the future (say, maybe 50 years or so from today) there will computers that can "conduct" physics experiments. Let's say, for example, we program a machine, with corresponding hardware apparatuses to detect a nearby magnet, record the electric and magnetic fields/potentials/etc at different points using hardware, run it through computational software and do some "best-fit" measurements to conform it to an all-encompassing equation and have the machine spit out one of Maxwell's equations? Surely, this would seem cumbersome at first, but with enough insight and trial and error, we could optimize calculations and generalize the machine to mathematically formulate other complex phenomena. This is one of the reasons that I've switched to computer science from physics, in that cs has so much potential. Has anyone heard of digital physics? Are there graduate programs in that field? Is a degree in CS enough preparation, or do you also need a physics background?

 

[Pengwuino]

In 50 years? How about 2009.

http://www.sciencedaily.com/releases/2009/04/090402143457.htm

 

[CyberShot]

In 50 years? How about 2009.

http://www.sciencedaily.com/releases/2009/04/090402143457.htm

I think I heard about that somewhere. That's pretty neat, but aren't those scientists comparing it with already known formulas? I'm talking about the computer being the one in charge here, and creating truly original expressions by generalizing and molding empirical measurements. Perhaps, that's a bit bold for right now, but it's good to see that progress is being made.

 

[Pengwuino]

I think I heard about that somewhere. That's pretty neat, but aren't those scientists comparing it with already known formulas? I'm talking about the computer being the one in charge here, and creating truly original expressions by generalizing and molding empirical measurements. Perhaps, that's a bit bold for right now, but it's good to see that progress is being made.

Well, of course they compare to already known formulas because we have to check to make sure it knows what it's doing. The key is they didn't tell the computer anything about Newton's laws and it popped out Newton's laws.

 

[DrummingAtom]

In 50 years? How about 2009.

http://www.sciencedaily.com/releases/2009/04/090402143457.htm

Don't show that to the theorists.

 

[flyingpig]

Ah damn it and here I thought I found the new path of life.

 

[leontd]

Computer Science was created by physics. Without physics foundation, this generation wouldn't have much technology as we do now.

 

[zif.]

Computer Science was created by physics. Without physics foundation, this generation wouldn't have much technology as we do now.

Agreed, and what's your point?

 

[chiro]

Can anyone out there entertain the idea of the likelihood that sometime in the future (say, maybe 50 years or so from today) there will computers that can "conduct" physics experiments. Let's say, for example, we program a machine, with corresponding hardware apparatuses to detect a nearby magnet, record the electric and magnetic fields/potentials/etc at different points using hardware, run it through computational software and do some "best-fit" measurements to conform it to an all-encompassing equation and have the machine spit out one of Maxwell's equations? Surely, this would seem cumbersome at first, but with enough insight and trial and error, we could optimize calculations and generalize the machine to mathematically formulate other complex phenomena.


This is one of the reasons that I've switched to computer science from physics, in that cs has so much potential. Has anyone heard of digital physics? Are there graduate programs in that field? Is a degree in CS enough preparation, or do you also need a physics background?

I don't think computer science will be the next physics. The domain and context of the two are so different.

Physicists think, analyze, and solve problems in a very unique way to computer scientists, and computers.

Even if a super-computer somehow magically fitted all the data to some model using advanced algorithms, it doesn't really know how to interpret it. You need physicists for that. You need context and context is not something that can be easily programmed into a computer.

Personally I think every major science has unlimited potential. There are so many unanswered questions, so many opportunities for further abstraction, and just so many opportunities for discovery and advancement.

 

[Monster92]

Yep, of course. Like Pengwuino showed there's already experiments done within the whelm of computer science. Obviously, as computer power increases we'll see more advanced experiments :)