Undergrad If AI could replace physicists why should we still research?

[Gjmdp]

I mean, probably in the future an Artificial Intelligence may solve every problem of Physics and even create a theory of everything. I feel like all the progress physicist have done in 100 years could be done in the future by a machine like in 1 second. So why should we keep on doing research at all? And the same applies to Mathematics, Biology... We could just leave this to the machines, right?

 

[.Scott]

Let's get in as much fun (ie, Physics discoveries) as we can before they take over.

 

[phinds]

I mean, probably in the future an Artificial Intelligence may solve every problem of Physics and even create a theory of everything. I feel like all the progress physicist have done in 100 years could be done in the future by a machine like in 1 second. So why should we keep on doing research at all? And the same applies to Mathematics, Biology... We could just leave this to the machines, right?

Well, you should just avoid all scientific research and the rest of us will go happily along without you.

 

[russ_watters]

I mean, probably in the future an Artificial Intelligence may solve every problem of Physics and even create a theory of everything. I feel like all the progress physicist have done in 100 years could be done in the future by a machine like in 1 second. So why should we keep on doing research at all? And the same applies to Mathematics, Biology... We could just leave this to the machines, right?

Because we don't know when or if it will happen and without that research it won't ever happen.

 

[fresh_42]

"Yes, man is gradually becoming superfluous. They say, there is already a computer in Japan that can be used to play chess against." - "Then I'd buy two" Jossele said. "They can play with each other and I go to the movies." - "Alright" I answered. "Let's go." (Ephraim Kishon)

 

[russ_watters]

"Yes, man is gradually becoming superfluous. They say, there is already a computer in Japan that can be used to play chess against." - "Then I'd buy two" Jossele said. "They can play with each other and I go to the movies." - "Alright" I answered. "Let's go." (Ephraim Kishon)

Funny article title* I saw today:
"AI[!] can do an astronomer's job 10 million times faster". Yeah, I've got one of those - it's called "Excel". Never heard it called "AI" before though.
*I didn't bother reading it.

 

[anorlunda]

It is even easier for AI to replace human posters on Internet forums. Why bother to ask questions?

 

[ZapperZ]

I mean, probably in the future an Artificial Intelligence may solve every problem of Physics and even create a theory of everything. I feel like all the progress physicist have done in 100 years could be done in the future by a machine like in 1 second. So why should we keep on doing research at all? And the same applies to Mathematics, Biology... We could just leave this to the machines, right?

This is silly.

Show me a computer that can derive superconductivity by solving EXACTLY the many-body problem, and we'll talk.

Zz.

 

[Gjmdp]

Well, you should just avoid all scientific research and the rest of us will go happily along without you.

Ok that was a little bit rude.

 

[Gjmdp]

It is even easier for AI to replace human posters on Internet forums. Why bother to ask questions?

Still, for now, they can't, or at least they can at a very primitive level hahaha. Tell me when they can.

 

[Gjmdp]

This is silly.

Show me a computer that can derive superconductivity by solving EXACTLY the many-body problem, and we'll talk.

Zz.

I mean in the future. Of course they can't for now. But I'm pretty sure when Super AI becomes real it will manage this task.

 

[Gjmdp]

Funny article title* I saw today:
"AI[!] can do an astronomer's job 10 million times faster". Yeah, I've got one of those - it's called "Excel". Never heard it called "AI" before though.
*I didn't bother reading it.

There are some papers in astronomy that use machine learning (AI). We are letting Artificial intelligence into Physics, and as we know AI will replace physicists in the future (machines will have achieved a superior understanding of the laws of Physics that no human could understand), so physicists are selling his jobs nowadays even though they know it. When I wrote this post I hoped some type of answer that made human research worthly rather than machine research, but as you have said research is now a mix of human and machine intelligence. What I can inference from your posts is that human intelligence will no longer be worth.

Some experts believe this AI will arrive in 2020-2030, so future physics may eventually have to leave their jobs as they will be of no utility anymore. Discouraging right? This is just speculation, but physics students should know that there is a very real chance that in a very close future they will loose job.

Thanks for all your responses, even the rude ones(...).

 

[russ_watters]

We are letting Artificial intelligence into Physics, and as we know AI will replace physicists in the future (machines will have achieved a superior understanding of the laws of Physics that no human could understand), so physicists are selling his jobs nowadays even though they know it.

Until it happens, nobody "knows" this: it is speculation. And I doubt most informed observers consider it likely in the near future.

 

[fresh_42]

Our most advanced instruments in physics (Kepler, VLT/ELT, CERN) all run automatically to some extend, including data analysis. Nevertheless, there are dozens of engineers, astronomers and particle physicists around to manage and evaluate the results. And the waiting lists for using time are considerably long.

I wonder, if an AI would have found the broken connection, when superluminous neutrinos had been measured.

 

[ZapperZ]

I mean in the future. Of course they can't for now. But I'm pretty sure when Super AI becomes real it will manage this task.

Sorry, this isn't a matter of "capability". How does an "AI" hope to do such computation?

Hint: look at the concept of "emergent" phenomenon.

Zz.

 

[Gjmdp]

Sorry, this isn't a matter of "capability". How does an "AI" hope to do such computation?

Hint: look at the concept of "emergent" phenomenon.

Zz.

I knew that already. Super AI would outperform every human ability, such as thinking and theorizing, not just computing, so if machines can't do the task you are talking of, human being will not be able to do that either. I don't see your issue as a problem.

 

[hilbert2]

Would humans end up being mere "pets" for those computers? I think it would be better to somehow "upgrade" the human brain by connecting it to computers that can handle those things that are not easy for a human mind.

 

[Dale]

Artificial Intelligence may solve every problem of Physics

Super AI would outperform every human ability, such as thinking and theorizing,

we know AI will replace physicists in the future

Please PM me with the professional references that support these claims. Until then, please review the rules about personal speculation. Although there is a lot of popular media hype about AI, we require stronger evidence than hype here at PF.