物理学家

9 (没有机会)说

We teach AI. That's how it understands. Although it can learn, it can't accumulate pure factual knowledge by itself.

Jamie (极有可能)说

Hard science fields will be the easiest fields for AI to take-over.

Zuzia (适度)说

AI can already teach itself new things and it escalates very quickly, it probably will be able to analyze all the knowledge we have on Earth and come to some important conclusions.

Samuel (没有机会)说

It requires thinking outside the box, solving new problems, writing new programs. It has already implemented computers for calculations.

Toast (没有机会)说

Being a physicist requires, at least to a certain degree, being able to come up with purely original ideas, rather than interpolating the existing body of knowledge.

Quarked_Out (低)说

Hmm, a lot of areas in physics do incorporate ML techniques and AI to some degree. However, people with physics training play an undisputedly dominant role in research.

If anything, I can see some simulation aspects or redundant experimental procedures being automated in the near term. But parts that incorporate creative problem solving or the physical intuition needed in determining directions to take research are things that are pretty safeguarded to humans for a bit.

I think those "intuitions" are very difficult to map to general problem-solving algorithms.

Brian (No Chance)说

Many other experimental physicists and I already automate every measurement we can, but there is still plenty of work to do.

David (没有机会)说

I wonder what all those people were thinking when they said there was a realistic possibility that physicists will be obsolete in 20 years. That's absurd on its face and every physicist I have ever met would agree.

We can have a discussion on whether it is possible in the next 200 years, sure, but 20 years? That's laughable. Given that only physicists are qualified to write, train, and optimize the algorithms that would be used to replace them, it will take a long, long, long time.

I suspect that we would need true machine sentience before we could actually start to talk about replacing theoretical and mathematical physicists.

Bimsara Bodaragama (没有机会)说

It's more about intuition and innovativeness. Of course, we will use more tools, but with Physics, since we interpret as we observe (there is a little problem with that conclusion, though), AI may not be able to take it over.

Wesley I (没有机会)说

I believe that it will be possible but not for a very long time, the process of positing new questions and then solving them is rather complex and I'm guessing that it will be at least 100 years before the jobs of theoretical physicists start to become threatened.

just bored (没有机会)说

Even though computers may get more intelligent than us humans, there is still a very small chance because computers don't have the basic questioning ability which we humans have

Rowan (不确定)说

The development of AI is rapidly improving, AI maybe 10 years in the future being able to predict or understand the universe better through random generation or pure knowledge is very probable. However I doubt they will replace Physicists in the near decade it should be very increasingly possible.

rafel说

We are users of AI to improve some calculations and we need robots and AI for experimental precision and performance. Until the singularity, we will be in charge.

Pinaki Patra说

A huge section of theoretical physicists do algorithm based research, which can easily be replaced by automation.
However, Philosophy based theoretical research is difficult to be replaced.

Physics Boi (低)说

Doubt it, anyway we need physicists to understand what the AI is discovering anyway

Anonymous (没有机会)说

It is scientifically proven that the human brain is MUCH more complicated than any AI, and it is very hard for AI to ask a question and solve it on their own, so I think AI stands no chance on taking over the complex job of science.

Tom (没有机会)说

Whilst experimental physics can be automated, I doubt theoretical physics will be automated

Mark (没有机会)说

Most physicists I know are already good with experimental automation and AI and there's still plenty of work for everyone.

Nine说

Scientific discovery can be easily automated. It has a rather rigorous process. And if not, then the computer can use randomness like humans do.

Me (低)说

Too chaotic