OT: One of the Most Interesting Topics I Have Read Recently

[Paul Alciatore]

I love it when science of math advances in unexpected ways.

https://www.wired.com/2015/12/outsid...-math-problem/

[lynnl]

"Ooops! Something went wrong when processing my request."

(later)
OK, now it worked.

Not just math. What I find even more fascinating is how once you delve down into many, or even most, advancements of a technical nature you will find that it's the cross-discipline idea exchange that was so instrumental. The most obvious of course is the role computers and computer science has had in enhancing every area of science and research. Or the effects of metallurgical advances propagating through transportation and most every industry.

[RB211]

Well, the article never said what the exact problem was, just called it out by name, and never mentioned the solution...

[garyhlucas]

If you have to ask you wouldn’t understand it!

[aostling]

I love it when science of math advances in unexpected ways.

https://www.wired.com/2015/12/outsid...-math-problem/

The Wired article was borrowed from the Quanta website, which published it seven years ago.

https://www.quantamagazine.org/compu...oblem-20151124

[aostling]

Here is an article Quanta published today. It is about "machine scientists." This is something new to me.

https://www.quantamagazine.org/machi...data-20220510/

[Rich Carlstedt]

Allan.
It's sort of hilarious to me that these scientists are looking for the ultimate "Objective" math solution to problems
then use "Subjective" reasoning as the answer to "which approach" is best. I find that a contradiction.

Rich

[aostling]

Allan.
It's sort of hilarious to me that these scientists are looking for the ultimate "Objective" math solution to problems
then use "Subjective" reasoning as the answer to "which approach" is best. I find that a contradiction.

Rich,

I don't see a contradiction. The objective algorithm separates the wheat from the chaff, but is unable to find the choicest kernel. That's where we come in.

I wonder if the machine scientist can be fed volumes of data from, say, a magnetohydrodynamics facility, and come up with the Maxwell Equations. James Clerk Maxwell had a strange model for deriving these equations -- it involved rotating hexagonal cells filled with an elastic media in a honeycomb pattern, with ball bearings (I'm not kidding!) in the space between the cells. From this he was also able to calculate the speed of light within about 1-2%. His inspired approach is long-overdue for a a magazine article explaining how it worked.

[RB211]

My brain isn't wired for theoretical sandboxes, why I loved anatomy but hated physiology.