Clearly since Ms. Livingston was graduated from college, she has learned a lot a...

graycat · on Jan 12, 2015

Future

So, what lessons for the future might a young person draw from this history?

(1) Science. Off and on, some amazing science, especially physics, played a huge role. Maybe that will continue.

(2) Information. The desire for information, create it, transmit it, store it, use it, etc., seems nearly unlimited.

(3) Logic. Want something done? Well, describe the work in clear steps. For a lot of work done manually in offices over the past 100 years, such a description is now usually fairly routine. Then with such a description, fairly routinely can write software to do the work. So, can automate a huge range of old, manual work of office workers. That's a lot of what for some decades made IBM successful.

(4) Social. People are highly social animals. Or to paraphrase E. Fromm, The Art of Loving, "For humans, the fundamental problem in life is doing something effective about feeling alone." In more detail, since humans are also thinking animals, we see that alone we are at risk, that is, vulnerable to, say, the hostile forces of nature (earthquakes, blizzards, tornadoes, floods, wild fires, disease) and society (war, crime, economic depression). Knowing that we are vulnerable, we are worried (have anxiety) and seek security. We feel more vulnerable when alone so want to do something about being alone. From Fromm again, the first recommended solution is a good romantic relationship. Next is a good version of religion -- get all wrapped up. Next is membership in a good group -- get acceptance and approval, a feeling of belonging. Next, not recommended, is what some college students try -- get drunk on alcohol, high on drugs, and go to an orgy. So forget about the worries until recover (but have more worries).

So, to do something about the worries, we want security, financial and emotional, don't want to be lonely, do want to be loved, want a romantic relationship, want to belong, etc.

More generally we will want to form good families and be in good communities.

We will be using computing and the Internet for all they are worth for such things.

(5) Economic Security. Likely second only to love, and maybe more important than love, and maybe essentially a prerequisite to love, people want economic security, and for that there is a famous one word answer "more".

The drive to use logic, software, computing, the Internet, etc. for "more" will remain powerful for decades, maybe centuries.

(6) Information. Now one of the keys to more in economic security, "more", is information, and the drive for that will also continue for decades, etc.

For information, we take in available data, process it, and report the resulting information. This processing is necessarily mathematically something, understood or not, powerful or not. Then clearly one approach to more powerful processing and, thus, more powerful and valuable information, is to use mathematics to determine how to do the processing.

E.g., how to look for oil? Okay, often oil collects in pockets in the subsurface layers. So, let's map the layers and look for pockets. How to do that? On the surface, have something go "boom". Sound waves go into the ground, and they get reflected off the layers so that there is a convolution. So, to find the layers, take the resulting signal and do a deconvolution -- Enders A. Robinson, 'Multichannel Time Series Analysis with Digital Computer Programs'. The fast way to do deconvolution? Sure, the fast Fourier transform.

Once get the oil out, over here have all that oil, from Texas, the Mideast, Venezuela, Canada, etc. -- typically it's all different. Over there know what can sell -- methane, propane, gasoline, Diesel, heating oil, motor oil, etc.

So, how to take the available input and sell the output and make the most money? That's a math problem, in particular in optimization. Long the first-cut approach was via linear optimization (programming in the sense of operational planning). At one time, IBM had fun selling mainframes to Houston for just this work. But linear programming is not quite the right stuff. So, want some non-linear optimization. Well, for more details, see the work of Christodoulos A. Floudas in chemical engineering at Princeton. Houston does know about Professor Floudas.

There's much more to do. Right: Likely not a single VC in the country says that they want to see some especially valuable software based on some especially powerful mathematics. Hardly a one. And they are not comfortable backing something they understand so poorly. So, right, a lot of confused and unhappy VCs (they so richly deserve it!) but: Presto, bingo, opportunity. Besides, the main raw material into original mathematics is paper, pencils, and coffee, and how expensive are those?

Almost inevitably, there will be only a few people going that way with the rest heaping ridicule, etc. Not nearly new: Think of the Mother Goose story The Little Red Hen.

Secret: It turns out, no matter how much advanced and/or original mathematics you use, nearly always a lot of the actual computations will boil down to linear algebra and there, numerical linear algebra. So, take linear algebra, elementary, intermediate, advanced, applied, numerical, and related subjects such as linear programming, non-linear programming, multi-variate statistics, ordinary and partial differential equations and their numerical solutions. For more, study the leading generalizations of linear algebra, functional analysis, e.g., Hilbert and Banach spaces.

(7) Niches. One of the standard ways to make money is to have close to a monopoly, and one of the standard ways to do that is to have a niche of some kind and where the monopoly is protected by, say, a geographical barrier to entry, an especially good product or service, some crucial, core, defensible technology or know-how, a good customer list, some network effect, etc.

So, go for it!