Saturday, 12 May 2018

Weekend Footnote: Friction among the Physicists

Along with the laughable Co-op document, something else that dropped through the letterbox this week was a newsletter containing a fulsome tribute to Stephen Hawking by his former tutor and long-time collaborator, cosmologist and first-rank mathematician Roger Penrose.

It makes for interesting reading (sorry, I haven’t found an online link - though he also wrote a straight obit in the Graun) and certainly contains glowing praise for Hawking’s determination, skills in maths and physics, insight, profundity and enjoyment of life.   But … away from the formalities of a national newspaper write-up, Physicist Penrose can’t help himself.  In the newsletter - you have to smile - he goes on to set out how he reckons Hawking took a wrong turn in his work on black holes; how he was barking up the wrong tree when he ought to have stuck to the earlier conclusions he reached when they were working together; and how Penrose’s own alternative theory would have it.

Never let a good intellectual feud drop ...  these academics, eh?



david morris said...

Or as Sam Johnson once said

"It is not easy to discover from what cause the acrimony of a scholiast can naturally proceed. The subjects to be discussed by him are of very small importance; they involve neither property nor liberty; nor favour the interest of sect or party. The various readings of copies, and different interpretations of a passage, seem to be questions that might exercise the wit, without engaging the passions.

But whether it be, that small things make mean men proud, and vanity catches small occasions; or that all contrariety of opinion, even in those that can defend it no longer, makes proud men angry; there is often found in commentaries a spontaneous strain of invective and contempt, more eager and venomous than is vented by the most furious controvertist in politicks against those whom he is hired to defame."

Electro-Kevin said...

How disappointing that the multiverse theory is to be replaced by the 2D hologram theory.

I want my money back on all of Hawking's books.

dearieme said...

My money is on an 11-D hologram. Strings, innit?

James Higham said...

Should academics be strangled at birth?

Anonymous said...

Great thinkers don't have to be academics.

The great thinkers know simplicity and that only comes through humility thus are, humble men.

If you cannot accept humility, if you cannot accept your ultimate destination, you should be rightfully buried by your fellow men.

All men can look to the firmament and stars, its how you see them which really counts.

Hawking, a very decent mathematician but still, he was a third rater because he was a very bitter man, understandably some would posit.

Nick Drew said...

@ don't have to be academics

It's always interesting to note "great thinkers" who did their work/ writing outside of academia. More easily done before the 19thC, of course.

Almost no maths these days**, and very little science (though some 'private' research labs exist - often, though, very much tied in to the academic mainstream).

The humanities are interesting. Would Dostoevsky count? As someone who takes an ongoing interest in philosophy I can tell you the uni-based philosophers are 100% of the view only academic work counts.

Footnote: Enron had pure mathematicians doing original work. I'm guessing a handful of the banks do, too

K said...

I was recently doing warehouse optimisation (pathfinding, routing, etc) and understandably there's very little public data as anyone with warehouses big enough to research the issue doesn't want to release their results to competitors.

But I did find quite a few academic papers on the subject. On the one hand I'm grateful that someone did some field work and released the results publicly but on the other hand I was surprised that a few hours programming is considered degree level in academia (also their warehouse was smaller and yet their solutions were apparently 100x slower than ours).

I wonder if this is also true in other fields. How much normal day-to-day business work would academics consider serious research?

Electro-Kevin said...

The language of the Universe is maths so, by it's very nature, it is esoteric and excludes the vast majority.

I still go on the multiverse theory and - like Douglas Adams - think the answer is in a number. Mine is not 42 (as it happens, the age at which one's own mortality becomes obvious) but infinity. Infinity is not actually a number but it is the point at which our existence transitions from probable to inevitable.

The answer as to why the Universe exists is simply binary. It is either "on" or it is "off". If it is "on" then we have the Feynmann "everything that can happen does happen" situation.

His slit experiment has an unanswered question. How can a particle's position become inexact when we observe it. I posit that it's our historical perspective which changes when we choose to oberve it - not the state of the particle itself because of some mystical intervention on our part.

I call bullshit on Hawking's 2D position. He threw this into the mix on his departure because he was a Remain voting cunt and thus proven he wasn't right all the time.

My Time Sausage Theory makes perfect sense and I am no academic (but my offspring are.)

I think my sample must have got mixed up with someone else's in the insemination clinic.

Nick Drew said...

For a post of no consequence that was a most enjoyable batch of replies!

thanks, all

Nick Drew said...

For a post of no consequence that was a most enjoyable batch of replies!

thanks, all

K - any further and better particulars? that sounds really interesting (for those who like that kind of thing ...)

Kev - I might almost say the same to you! mmm, sausage ...

Electro-Kevin said...

That a guy walking down the street would look like a sausage to an observer freed from time.

That every moment exists in perpetuity in its own slice of spacetime - and that there are infinite versions of those moments. What we see is the most likely result in a probability cloud.

I like mine spicey.

I'm also coming to the belief that - rather like gravity around planets - our very presence shapes the universe around us in a relativistic way, a bit like Einstein's theory. Observers see a very different version of reality to our own.

At least that's my explanation to my boss when I've fucked things up royally.

dearieme said...

"not the state of the particle itself because of some mystical intervention on our part": good God, it's not mystical. It's because you've whacked it with a photon. You should search out that old classic The Strange Story of the Quantum and enlighten yourself. Any intelligent fifteen year old should cope with it easily.

P.S. The author is wonderfully memorable - Banesh Hoffman.

E-K said...

At point of impact we should know position and velocity. By *deciding* to look at it we somehow change its history.

E-K said...

The particle is, paradoxically, wave and particle and takes all routes on its journey in that instant.

So hit what with a photon, exactly ?

I say sausages !

K said...

@Nick Drew

Basically we saw the warehouse as a small static map like a video game and copied what video games have been doing for 20+ years.

The academics could only see the warehouse as a real location and got stuck in algorithm tunnel vision. They kept trying to use different travelling salesmen solutions which are appropriate for a whole city or country but is overkill for indoor routing.

Pretty much a symptom of how computer science is taught: too much focus on algorithms and maths and not enough actual programming. They'd probably consider our solution a dirty hack but who cares if it can find real answers 100x faster.

Charlie said...

K - exactly this! Back in the late 90s I did a module on VB as part of my Comp Sci degree (God knows why, it was apparent even back then that the language was awful). An early assignment had us create a function that would take an integer as input, return true if prime, false if not. The first couple of lines of my code checked if the input was 2 (return true) or even (return false). My lecturer marked me down for polluting my algorithm with such dirty logic!

I went off and wrote another version without those steps, which was obviously, on average, about half as fast to return a result as my "ugly" version.

Nick Drew said...

Thanks, K & C both

At Enron we were dissatisfied with the standard 'academic' modelling of gas & electricity prices - particularly elec, because we were convinced the algorithms used by the grids would mean negative prices were possible (at the time that was viewed as a theoretical poss, but nowadays it's almost an every occurrence, esp. in Germany). The then-standard representation of volatility (geometric brownian motion using log functions) wouldn't allow for negative prices. Our in-house quants came up with sinh-1, basically by looking through maths text-books to see what functions had correctish-looking shape but carried on through 0 and out the other side. It worked brilliantly, and we were even able to post-rationalise why

Another one I recall was the puzzling challenge of how to value options in a market where the holders frequently exercised them irrationally. All the Nobel prizes for option valuation (Fisher/Black/Scholes etc) assume - natch - rational exercise. But many players in the energy markets are both (a) reacting to drivers that are not strictly financial and (b) often actually quite stupid. So there was a perennial tendency to over-value options

Our people discovered there was an excellent body of work - highly empirical, of course - done by some of the banks on mortgage repayment data, which turns out to be a very similar problem. In the US you can re-mortgage at the drop of a hat, and logically you'd expect people to do this whenever interest rates change by some threshhold amount. In fact, people repay mortgages (or don't) for all sorts of reasons that aren't strictly 'financial', e.g. when they move, or die, or because a salesman has persuaded them to, or because they are stupid etc etc. This analysis turned out to be very useful

Final story: when I was an undergraduate (1970s), the (world-renowned) Maths department was starting to get mercenary. The applied maths people got a contract from British Rail to model what was going on with the electricity current when an electric train gets one of those massive sparks at the live contact.

They worked away for 2 years and came up with ... absolutely nothing of any use. (I don't know how that story ends - if indeed it does. They probably use an ultra-fast camera these days)

Anomalous Cowshed said...

"I wonder if this is also true in other fields. How much normal day-to-day business work would academics consider serious research?"

I suspect rather a lot.


"Another one I recall was the puzzling challenge of how to value options in a market where the holders frequently exercised them irrationally."

Something similar - back in the 90s we used to see warrants being exercised by investors in this fashion. It didn't make any sense and certainly a lot of it was small retail, so just naive, but every so often we'd come across situations where institutional investors would do it. And they certainly should have known better.

Around the same time, we were looking at a couple of other things, like options embedded in convertibles, and there was a bit of history there, what with the Jap warrant market an' all. What academic research we could find was at least two years behind what the guys on the other desks had been doing for a while, and you could find functions on the Bloomberg terminal that were well ahead of the academic models.

Synchronicity - I sort of get involved in warehousing systems, very peripherally, about once every three years for some bizarre reason, so it's a bit odd that they've got a mention here and here in the last month. Note the comment about customers having paper based systems - it's 2018 for god's sake.

I'd hazard a guess that the problem is with the "pure" research role that academia has taken on - research groups have become discrete silos, isolated from real world practice. If the definition of an expert is someone who has explored all the possible ways of fucking something up, and although it's good to learn from your mistakes, it's much better to learn from someone else's, then the value of a university education is in being taught about the failures, not solely about the solutions. The implication here is that academia is a store of practice, and a source for the dissemination of that practice; not a high priest.

And O/T; nothing wrong with VB. One of the best ways of using it was for quickly producing front-end stuff for multi-tiered systems, with all the "hard" stuff being embedded in the back-ends, using languages or tools that had much smaller footprints (no dirty great run-time exec, see?). But, around the time VB6 was released, a bit of benchmarking showed that it was as fast as the C/C++ compiler of the same generation; since then, we have developed a metric fuck-ton of new languages and libraries whose performance is entirely dependent on their RTE. Under Windows, everything is .NET with the sodding CLR. I weep for the processor cycles we have sacrificed.

Nick Drew said...

Anomalous - you should come back here more often!

Anonymous said...

I was under the impression that Enron's mathematicians were all specialists in irrational and imaginary numbers... Which then became the basis for the company's accounts.