Is Theory Essential for Success in Industry?

[Crass_Oscillator]

In general, engineering and scientific disciplines (including mathematics) have a theory wing. Commonly I encounter people in the trenches who seem to think theory is useless and that only heuristics get things done. I'm fairly sure this is false, but I want to know what other people think. I'm posting this in the career section because I personally avoided pursuing theory in physics because I wasn't sure it was relevant, in some way, to applications. Some people of course will argue that its relevance to applications isn't necessary, and I'm curious about that too.

 

[Dr Transport]

Theoreticians get jobs in industry to program, they don't do pure theory.

 

[Crass_Oscillator]

Right, I wasn't clear. It's not what an individual with a theory background does in industry, it's how that theory work in academia is actually useful in industry. The (American) government, for instance, does not like funding theory in mathematics, physics, even engineering for instance, as far as I know because it doesn't associate that with results.

 

[phinds]

Without the theory, there wouldn't BE any applications. For example, the transistor, which is at the heart of all modern electronics, was not developed by accident nor was it done without knowledge of underlying theory.

 

[symbolipoint]

In general, engineering and scientific disciplines (including mathematics) have a theory wing. Commonly I encounter people in the trenches who seem to think theory is useless and that only heuristics get things done. I'm fairly sure this is false, but I want to know what other people think. I'm posting this in the career section because I personally avoided pursuing theory in physics because I wasn't sure it was relevant, in some way, to applications. Some people of course will argue that its relevance to applications isn't necessary, and I'm curious about that too.

Theories are applied. They really are! You learned to think analytically using equations and other relationships which were derived from theories. You did this and will do this in order TO APPLY your theoretical knowledge.

 

[Dr Transport]

true, but i am not aware of any company paying a theorist to do pure theory that doesn't have a possibility of producing a profit margin at the end via a device or something similar at the end.

The US department of defense which funds a huge amount of R&D has strings tied to whether or not the theory is tied to a device. Pure theory for knowledge sake just isn't wanted.

 

[symbolipoint]

One would apply the theories which are relevant to the goal of ones current project, assignment, or study. I really do mean, in the real, working world.

 

[Crass_Oscillator]

true, but i am not aware of any company paying a theorist to do pure theory that doesn't have a possibility of producing a profit margin at the end via a device or something similar at the end.

The US department of defense which funds a huge amount of R&D has strings tied to whether or not the theory is tied to a device. Pure theory for knowledge sake just isn't wanted.

Well, they do pay people to do theory, which is good to hear.

 

[rootone]

Well, they do pay people to do theory, which is good to hear.

Only if a theory is necessary in order to achieve a predetermined goal.
In commerce most new development is biased towards improvement of models that are known to be reliable.
Shareholders and all that.

 

[Crass_Oscillator]

Only if a theory is necessary in order to achieve a predetermined goal.
In commerce most new development is biased towards improvement of models that are known to be reliable.
Shareholders and all that.

I guess when I look at Hamilton's equations or homotopy groups I get depressed when I ask myself the question "Does anybody use this to do something 'useful'" and assume the answer is 'No'. It's more a philosophical question. What's the point of being a professor working on homotopy theory (disclaimer: I don't know if that's a finished subject or not) or lattice QCD if it just tells you about experiments or proves a little detail of some math but does not really provide one power to do something like, I dunno, transport people across a large country in fewer hours? Does it really mean anything at all if none of it leads to greater greater control over Nature?

 

[phinds]

I guess when I look at Hamilton's equations or homotopy groups I get depressed when I ask myself the question "Does anybody use this to do something 'useful'" and assume the answer is 'No'. It's more a philosophical question. What's the point of being a professor working on homotopy theory (disclaimer: I don't know if that's a finished subject or not) or lattice QCD if it just tells you about experiments or proves a little detail of some math but does not really provide one power to do something like, I dunno, transport people across a large country in fewer hours? Does it really mean anything at all if none of it leads to greater greater control over Nature?

Science has a long history of "useless" theoretical research turning out to lead to unexpected practical applications. One never knows.

 

[Vanadium 50]

The (American) government, for instance, does not like funding theory in mathematics, physics, even engineering for instance, as far as I know because it doesn't associate that with results.

Not true. The Department of Energy supports HEP theory at the level of just under $50M per year. About 30% of this is formal theory and string theory. There is also theory support from NSF, DOE-BES for condensed matter, etc.

 

[Crass_Oscillator]

That's very exciting to hear Vanadium. I guess I just assumed that HEP was so poorly funded that it was impossible to become a HEP theorist or that there might not even be funding. I don't know where I heard that. I personally avoided HEP because I thought it had little to do with experiment, although it was very interesting. Am I also being naive about that?

I had completely closed off the idea of pursuing HEP for those reasons and wanted to work on condensed phase/biological systems (although I really hated doing simulation work on biological systems and theoretical work is very rare).

EDIT: Two more quick remarks. I never thought HEP was more or less interesting than condensed matter although for reasons I don't quite understand people seem more into it. Also at this point I think it would be very hard to get into a top university. For some reason I assumed that going to a top program like MIT or Princeton was a necessity to do HEP theory (and honestly most other forms of physics).

 

[radium]

I actually found out recently about some interesting pen and paper theory type work one can do in industry in consulting like positions. It apparently pays very well.

 

[Fervent Freyja]

I don't understand why professionals would pick on other professionals to such an extent. When you do that, you are also putting down the majority of people in the world that hold more 'common' jobs that certainly aren't considered careers. Be happy for each other instead! There are plenty of other job titles that could be considered excessive, or 'useless' in terms of meeting basic survival needs for the world.

From that perspective, even all science and technology fields in application are useless; especially, when at least 500,000,000 people don't have access to clean water-- what does ya' fancy iPhone 6 or talking refrigerator do for them? What will solve that issue? A theoretical solution that can eliminate a certain bastard fly from the Earth would be a real good start, only then can something real be applied to the issues those people face.

Backwards, this world is.

 

[Student100]

The US department of defense which funds a huge amount of R&D has strings tied to whether or not the theory is tied to a device. Pure theory for knowledge sake just isn't wanted.

Not quite true. I use to work for the DoD, they do a surprising amount of purely theoretical work that has no readily apparent application. They even fund what I would consider "questionable" basic research.

They can fund such research because the DoD has a lot of money- a lot - and is publicly operated.

 

[Crass_Oscillator]

I actually found out recently about some interesting pen and paper theory type work one can do in industry in consulting like positions. It apparently pays very well.

Elaborate?

Not quite true. I use to work for the DoD, they do a surprising amount of purely theoretical work that has no readily apparent application. They even fund what I would consider "questionable" basic research.

They can fund such research because the DoD has a lot of money- a lot - and is publicly operated.

Such as?

 

[ZapperZ]

I think that at this point, a few things need to be clarified (just like butter) so that people have a clear idea of the topic that is being discussed.

There is a difference between the topic of "the value of theory" versus "the value of theorists" in industries. The first is the value of theoretical development to a for-profit entity. The latter is the value of the people who specializes in theories and their skills to those said entities. The argument that theorists can do mathematical modeling in many different areas, etc. are examples of the latter, i.e. those are about the people and their skills, not about the value of theoretical scientific ideas and their development.

Whenever we deal with a topic like this, the only convincing argument to support anything is via bringing up clear examples. There is never going to be a "universal principle" in which something like this can be applied to. In other words, if you say one thing, someone else can equally counter that with something else. A theory may be important in that company, but in another company, it couldn't be any more worthless. So I will claim that it is impossible to make a blanket statement about this topic in any shape or form. The best one can do is to bring up specific examples to show that it CAN happen. And that is what I will try to do. It isn't meant as a universal rule.

The clearest example of where both theory and theorists thrived was at Bell Labs before the breakup of AT&T (i.e. when it was still a monopoly on the US telecommunication system). Bell Labs was arguably THE most prestigious science laboratory in the world at that time. Some of the biggest names in physics came out of there. And this includes theorists. Phil Anderson worked on his Nobel Prize winning theory while at Bell Labs. John Bardeen carried out the theoretical calculations and worked on the first solid-state transistor at Bell Labs. etc.. etc.

So is it merely a coincidence that the glorious years of Bell Labs and the enormous body of patents and inventions came out during this period in which both theory and applications went hand-in-hand and were both living under the same roof?

High-tech industries, for example, depend on theory and advancement in theory. The question is, do they want to do this in-house, or can they just rely on government labs and universities to do such things? The Bell Labs of old did many of their own R&D, which includes theory. But these are long-term investments, and investments in which a profitable outcome isn't guaranteed. Will the greed of today allow for something like that? We have seen what happened when Bell Labs became part of Lucent, and now, all of the sudden, the issue of profit margins becomes the focus.

But another issue here has been overlooked. It can be that a particular company wants to include theoretical development that are in-line with what they do. Maybe a company is developing a new material, and requires theoretical band structure calculations to guide them in the fabrication of this material. But do they really need a theorist to do that? Many of us who graduated in experimental condensed matter physics can also do such theoretical calculations, and more. In other words, the company can hire an experimentalist who can do a lot of theory development, but is also handy with the experimental aspect of fabricating these material. What could be better than that?

So I can easily see a scenario where, yes, industries may value theoretical work, but they need not want only theorists. Therefore, in this case, a company may value theory, but they may not value theorists, simply because they want someone who can do a lot of other things as well.

Zz.

 

[StatGuy2000]

The clearest example of where both theory and theorists thrived was at Bell Labs before the breakup of AT&T (i.e. when it was still a monopoly on the US telecommunication system). Bell Labs was arguably THE most prestigious science laboratory in the world at that time. Some of the biggest names in physics came out of there. And this includes theorists. Phil Anderson worked on his Nobel Prize winning theory while at Bell Labs. John Bardeen carried out the theoretical calculations and worked on the first solid-state transistor at Bell Labs. etc.. etc.

So is it merely a coincidence that the glorious years of Bell Labs and the enormous body of patents and inventions came out during this period in which both theory and applications went hand-in-hand and were both living under the same roof?

High-tech industries, for example, depend on theory and advancement in theory. The question is, do they want to do this in-house, or can they just rely on government labs and universities to do such things? The Bell Labs of old did many of their own R&D, which includes theory. But these are long-term investments, and investments in which a profitable outcome isn't guaranteed. Will the greed of today allow for something like that? We have seen what happened when Bell Labs became part of Lucent, and now, all of the sudden, the issue of profit margins becomes the focus.
Zz.

In the case of Lucent, the answer will probably be a "no" (at least for physics research), but in areas like theoretical computer science and various branches of mathematics (including probability theory) which also are not immediately "applicable", Microsoft Research has filled the void left by the downfall of Bell Labs of old. There may be other examples of in-house corporate research labs working on theoretical areas (perhaps Google Labs?), but I do agree that these are much less common than in the past.

 

[radium]

The type of theory I was referring to has condensed matter applications but I don't know the exact details.

In terms of computational materials science, there are theory groups that are solely/almost solely computational (DFT etc.) so those people are definitely employable. They may work in areas like materials design for example. I think theorists are very valuable in that area.

 

[Crass_Oscillator]

Very intriguing and insightful remark ZapperZ. I actually have a very specific question for you and anybody else.

What role did theoretical calculations play in the invention of the transistor? Of course nobody was using supercomputers to do DFT in those days since there weren't any supercomputers. Does anybody know if the theoretical contribution involved only phenomenological/qualitative ideas from quantum mechanics, or more?

This is very specific and perhaps deserves its own thread.

 

[ZapperZ]

Very intriguing and insightful remark ZapperZ. I actually have a very specific question for you and anybody else.

What role did theoretical calculations play in the invention of the transistor? Of course nobody was using supercomputers to do DFT in those days since there weren't any supercomputers. Does anybody know if the theoretical contribution involved only phenomenological/qualitative ideas from quantum mechanics, or more?

One of the most significant contribution by Bardeen at that time was his proposal of a surface state to explain the discrepancy in the thin-layer field effect measurement that lead to the initial failure. It allowed them to find ways to work around that. This was a theoretical proposal based on what was understood at that time in the burgeoning field of solid-state physics.

If you have time, I highly recommend reading Hoddeson and Daitch's biography of Bardeen "True Genius: The Life and Science of John Bardeen: The Only Winner of Two Nobel Prizes in Physics". His role, and the role of theoretical calculations in the invention of the semiconductor, are clearly described in there.

Zz.

 

[Chestermiller]

When I worked in industry, I saw theory applied routinely in many different areas. Here is a small selection:

Thermodynamics, chemical reaction kinetics, and mass transfer operations modeling in the design and operation of chemical processing equipment
Chemical reaction kinetics, radiation transport and scattering, photochemistry, dispersion dynamics in atmospheric modeling of CFC and global warming impact
Groundwater flow and transport modeling in subsurface contaminant transport and remedeiation
Seismic modeling in oil exploration
Fluid dynamics and heat transfer modeling in polymer processing and man-made fiber processing
Solid mechanics modeling in development of composites, mechanical parts, and processing equipment

 

[Crass_Oscillator]

Thanks for all the information and advice everyone.