Hacker News Reader

Comment by sleavey | Hacker News Reader

I went to university intending to study computing science, but switched to physics half way through under the assumption that you could teach yourself computing science more easily than physics. I wouldn't change what I learned for the world, and I'm now in physics academia, but it's funny how my job these days involves using some of the algorithms, mathematics and data structures that I was just about to start learning if I had stayed the course in computing.

My few years of study and general interest in computing has already given me a huge advantage in solving certain types of problem. I think computing science (not just software engineering) is slowly but surely becoming an essential part of many areas of modern physics research.

View on HN

Not just research, but applied physics too - at least, that was my experience working in the defence and energy sectors. Looking back, it’s remarkable how little of my physics studies were devoted to learning about the computational techniques I spend the majority of my time working with.

I agree. It seems to me that in those huge physics projects like CERN what most physicists do is programming or data analysis.

> computing science ... is slowly but surely becoming an essential part of many areas of modern physics research.

Also, this means that fundamental assumptions and the calculations are hidden from view. Recently, some physicists questioned the findings of LIGO but if I understand correctly all is decided by data analysis software and outside teams have a disadvantage to decipher it.

The LIGO result is actually excitingly available as an IPython notebook, which provides a tutorial to verify the data yourself, using entirely open source software. https://www.gw-openscience.org/s/events/GW150914/GW150914_tu... The actual signal processing isn't that complex IMO.

I don't know how open the LIGO hardware is, but it's very unlikely you would build one yourself to verify.

This is very true - I was a physics undergrad and learned to code working in a project at cern. While the underlying physics was the foundation of the project, the day to day work was all coding: processing and analyzing the huge amounts of data coming out of the experiment.

> a physics undergrad

Now imagine dozens of people with a PhD in physics doing the same. (Earning a degree in physics is hard.)

yes, that was one of the insights of the work. The people working alongside me were post-docs that were still having to put in their time as workers on other people's research projects. Particularly, in the area of experimental high-energy physics, where accelerators were needed to do the work, everyone had to jockey to get time on the accelerator (even nobel laureates like the leader of the project I worked on). It seemed like every experiment was a compromise of different people's goals in order to combine together to get enough collective clout to get access to the accelerator to even do the project. That meant the guys who had already completed PhDs at prestigious schools were still pretty low on the hierarchy (but ten years older than me). The time spent working with those guys definitely influenced my decision not to continue with a PhD...

> I think computing science (not just software engineering) is slowly but surely becoming an essential part of many areas of modern physics research.

I like to say that physics, mathematics, and computer science are the three pillars of human knowledge. There's some overlap between each pillar, and we don't yet fully understand how each of them fully relate to the others, but each is essential to gathering a full picture of any phenomenon.

I think there must be more pillars than that. Biology and chemistry while related to physics are distinct (in the same way physics is not maths). More broadly, humanities and the arts are still valid areas of knowledge. Hacker news has a bias towards your three pillars, but I think it is important to acknowledge that human knowledge does go beyond CS, maths and physics.

I disagree re: biology and chemistry. Certainly they are epistemically useful simplifications of the other pillars, but they are not foundational or fundamental.

Physics is not math because it's an empirical rather than a priori exploration. We are trying to reproduce the function governing natural laws, where in mathematics we are exploring the properties of functions. Computer science in turn is concerned with the construction of mathematical structures via algorithms, and so physics will be one such construction.

I think trying to place humanities and the arts under "knowledge" is a category error. These fields don't produce knowledge as its commonly understood, as a corpus of mind-independent facts and their relationships. Certainly facts appear in these fields, such as historical facts and authorship, but I don't think building a corpus of facts is not the main purpose.

> simplifications of the other pillars

This fallacy is called "reductionism."

It's not universally accepted that it is fallacy. And for whatever it misses, reductionism is a hell of a tool that has been incredibly useful in understanding the world.

Of course, so are chemistry and biology, so blithely dismissed in the parent comment.

> Of course, so are chemistry and biology, so blithely dismissed in the parent comment.

I didn't dismiss them at all, I explicitly said they are epistemically useful, but that doesn't entail we should accept them into our core ontology.

It's a philosophy not a fallacy: https://plato.stanford.edu/entries/scientific-reduction/

Wrong. Philosophy and Mathematics are the real pillars of knowledge. Everything else can be derived from this including physics.

would you care to elaborate?

Philosophy doesn't really produce knowledge, it explores and frames questions. When we accept that the questions define a coherent domain of discourse, we create a field of study that does produce knowledge. This is how natural philosophy became science, and if ethics ever gets to this point, we'll get a moral science which will produce moral knowledge. Ethics doesn't currently produce moral knowledge.

And that's the problem with living in a philistinic society. When people think the sciences are the only noteworthy achievements of human intellect.

Wait a bit, in a few years physical and computing models will merge. #my2ctsbet

Reply on news.ycombinator.com