Gee, looks like a LOT of dedicated, devoted English majors here. Who would'a thought?!
31 karma · joined January 21, 2011
"Journalists"? You are being moderated in your high praise of 'journalists', that is, professional members of the highly professional journalistic profession?
Cruel. SO cruel. How can you be so CRUEL?
I mean, while we're discussing particle physics, don't you believe that the English, drama, theater, and communications majors also need to eat?
And it's even worse, you are being even more cruel: You are suggesting that the journalistic profession should pay attention to technical information and, thus, push out the main technique of journalism back over 100 years: Use communications of human experience and emotion to grab people by the heart, the gut, and below the belt.
So CRUEL!
No, again, I just addressed the point of the article, 'productivity'. To move on to address the 'social', etc. results from getting more productivity via information technology would have been 'off the subject'. Or, raw productivity remains important.
For your concerns about China, I'm not much concerned. My biggest concern would be that they will be able to compete more for some of the 'world-market' goods such as oil. But, the US doesn't have to be short of oil: We are just awash in shale oil and coal, both of which can be made into gasoline, at roughly present prices. And longer term, with bioreactions, possibly endothermic and driven by electric power from nuclear reactions, we need not be short on energy in any important forms.
For the US 'standard of living' in, say, GDP per person, one decade at a time, I see that increasing at a nice rate and not falling. Yes, GDP per person is an 'average', and taking the average raises some issues: So on the one hand is someone out of work who made $50,000 a year, and on the other hand is someone who gets a vacation for three months in the south of France in a hotel at $2000 a day. So, how to 'average' that?
Or, after $100 million, who would even walk across the street for another $1 million, yet a poor person would crawl across the street for $1000, $100, maybe $10. So with a high GDP per person, we can have more people not willing to walk across a street for another $1 million while we have others stuck with very low fraction of the GDP.
Broadly, a society awash in cheap robots should not have nearly all the humans destitute and on the streets but with really well off humans being served by the robots!
Really, that desirable future needs more, actually MUCH more, economic productivity because the GDP per person now is not nearly high enough.
So, one issue is increasing the GDP per person; another issue is what to do with the increases.
Here's a first-cut, fairly automatic resolution: Currently an information technology entrepreneur can start a business, get a large increase in productivity, keep the revenue and earnings, and put out of work a lot of other people. Well, as people go out of work, the safety net will cost more and taxes will increase. Then the entrepreneur will not deploy the technology unless the revenue can cover BOTH some earnings AND the taxes for the safety net. So, technology that does get deployed will increase productivity enough to pay for both the technology AND the people put out of work. Or, instead of the Google founders being worth maybe $15 billion each and putting out of work thousands of people in, say, newspaper classified ad departments, the Google founders will pay higher taxes and the safety net for the former newspaper workers will improve. Our politics and society might move to such a situation without ever really agreeing that 'technology should pay for the jobs it kills' or some such.
For the politics, I'm against stupid government that makes things much worse, and we've had WAY to much of that. It would be negligent for me not to mention the real cause of the housing bubble that caused The Great Recession: My view is that the bubble was caused by stupid government, by Obama (when he was a Senator), the Black Congressional Caucus, and Barney Frank telling Fannie and Freddie to back junk paper. It was NOT caused in any significant sense by Wall Street, 'complex' financial instruments, AIG, or the bond rating agencies. I do agree with Greenspan that he was a significant cause.
I do want my slice of the pie to get bigger but from making the whole pie bigger, not from taking from someone else, but this desire does not translate very directly into any feasible, effective 'political action'.
Beyond politics, a crucial issue is, yes, good government and, then, MORE, actually MUCH more, economic productivity. Only with more GDP per person can we have the safety net as strong as we want it. Or "The problem with socialism is that it too soon runs out of other people's money.". Right. So, one solution is enough more economic productivity so that the other people won't run out of money. Simple enough.
Then there will be other problems: The political system can split into (1) the people working and (2) the people in the saftey net. Then the people in (2) can so choke the geese in (1) who are producing the golden eggs that in every sense standard of living stagnates. Indeed, that is a common description of just what is going on in most of the more socialistic European countries.
Yes, a worker makes a big 'investment' in education, training, experience, etc. in their career. In accounting terms, that investment is an asset on their personal, financial balance sheet. Then when they lose their job, they have to 'write off' that asset. That 'write off' shows as a big loss on their income statement for that year. So, who covers that loss? The person and no one else. And, you are correct: We are short of capital to replace that asset on that person's balance sheet. So, the person is now 'under capitalized' and like such a business that has shut down and left a building with a leaky roof, broken windows, and weeds in the parking lot. Not good.
Before the economy gets going again, we may have to open 'camps' where people live in dorms and eat in common cafeterias.
Sadly I blame the UAW, for then and for what's happened to Detroit since, but it still happened. It was grim. Sadly, a UAW like solution solves some problems but creates some bigger ones including just killing off the whole industry it tried to 'organize'.
Or it's "the great wasteland".
Or it's the "message" in "the medium is the message".
But heavily it's the old tried and true of the humanities culture, 'drama' as in 'formula fiction', especially as in the 1000 year old 'morality' plays.
So, I'm reluctant to believe that more of the 'humanities' culture will have a solution.
Or, with high irony, in main results, if not in all intentions, the 'humanities' are a great enemy of humans!
In some ways it's a BAD situation.
But, in 'productivity', that is, economic results per dollar, that remains. It's just that increases in 'productivity' don't always mean an increase in standard of living, quality of life, better society, etc.
All I addressed was productivity, not standard of living, 'the social contract', 'financial security', a 'better society', etc.
You are absolutely correct that much of the technology that has increased productivity has hurt a LOT of people. And it is very true that our society needs to do something about this situation.
And, moreover, if count the cost of the 'safety net' for the people put out of work for the technology, then it may be that, with also the cost of the technology, sometimes there is a net loss in 'productivity'.
For another example, Google, Craig's List, etc. are putting out of work a LOT of people in the classified ads departments of newspapers. Heck, the Internet is putting a lot of newspapers and magazines out of business: Newsweek just sold for, what, $1 to someone who wanted to run it as a personal interest instead of a business?
I haven't gotten a newspaper or magazine on paper in years, and last week I finally hauled out three really heavy, big, green plastic bags of old magazines.
Still, in the long term, larger picture, and in the sense of the article, 'productivity' marches ahead. We don't like all the downside, but, still, we don't really want to go back. So, we need to find better solutions for how to move ahead.
Sure, it's too easy to see that too soon we will have to give everyone a basic 'stipend' so that anyone who wants more can work and make much more. Doing this and keeping a 'productive' society will NOT be easy.
Or, what happens when Robots, Inc. makes a robot for $1000 that can do the work of over 90% of the current jobs?
Still, the article and I were writing about JUST 'productivity' and not at all about your, justified, concerns about 'quality of life'.
"The United States spends more per capita on health care than any other country, yet without producing measurably superior results."
That situation is always easy to obtain: Just do a sufficiently poor job at measuring the results.
The tiny, short-term drop in the quantity he quotes says nothing, not even as much as zip, zilch, zero, about his claim of "trump" and lower productivity for the world. For such a claim, we'd need some data over a longer term and for much of the world and not just for the US. And we'd need more than the data: We'd also need some good candidate explanations. His article has no such and no references to any such.
Instead, with so little data, some of the best data is striking and not in the article:
Look at 'productivity', for the US and the world, in 'information technology'. E.g., it was not so long ago that 1 billion bytes of disk was in a box about the size of a large SUV and would really set you back, and now can get 1 billion bytes in main memory for about $5, retail. For just disk, about $100 will buy about 3 trillion bytes. It may be that of the 1 billion byte boxes, they never sold in total as many as 3000 for 3 trillion bytes. There was a time with reports that FedEx had, in total, 7 trillion bytes of disk. Now that's about $300 in a mid-tower case for about $100, actually, a whole computer with four processor cores for less than $1000. Did someone mention 'productivity'?
Think about telephones: I well remember paying monthly telephone bills of well over $100. Not anymore! Think of mobile phones, useful in many trades, e.g., electricians and plumbers, nearly anyone who in their job spends a lot of time in a car: A mobile phone used to be just some radio in, say, a police car; then a mobile phone was about the size of a brick; then a mobile phone would fit in a shirt pocket; now a mobile phone is a quite capable computer, Web browser, video camera, movie player, electronic funds transfer device, etc.
Think about tire life: It used to be that 10,000 miles was good tire life. Now can routinely get 80,000 miles. Better chemistry for better tires!
Cars are safer with more air bags, body designs better in crashes, anti-lock brakes, stability control, various warning systems, etc. So, auto insurance rates should go down, and that's 'productivity'.
Carbon brake disks are coming to cars, and the disks can be guaranteed for the life of the car. "Look, Ma: No more brake jobs!".
Think about car engine life: With all the computer controls, get much less gas and water in the oil and, thus, get much longer engine life.
Think about the changes in house building materials and tools: Now can put up a house for many fewer worker hours. And house insulation: Have much better insulated roofs, walls, doors, and windows, with much less air leakage, for much lower HVAC bills.
Seen a cargo ship lately? They are HUGE, carrying oil, natural gas, or containers, with much better cost per ton-mile.
Bandwidth should count: Kept track lately? Now can put several trillion bits per second on one wavelength, some dozens of wavelengths on one long-haul fiber, maybe 144 fibers in one cable, and several cables along a road, pipeline, high voltage electric line, river, coastline, etc.
Remember the 'T-1 leased line' from Ma Bell? That was 1.5 million bits per second (Mbps) for over $1000 a month. Now for less than $60 a month I have 15 Mbps download speed and 2 Mbps upload speed. For another $65 a month I can get a static IP address, 101 Mbps download speed, and 15 Mbps upload speed. We're talking at least 10 times the data rate for about one tenth the cost for a 'productivity' increase of 100:1, before counting inflation. We were talking about 'productivity', right?
Heck, now robots are cheap enough that even the Chinese are converting from manual assembly to robots.
A guy had some chest pains. In the hospital they did an angiogram, found a clot in a coronary artery, sucked it out, and sent him home the same day. To get home he drove his own car. Did someone mention "productivity"?
I still have a nice sack full of gorgeous Nikon camera equipment. But it uses just film. It's still okay, but now a digital sensor can have 48 million individual sensors, four for each pixel, that is, two for green and one for each of red and blue in a little square, giving 12 million pixels giving an array of 3000 by 4000 pixels. So, no film developing! Anyone mention 'productivity'? Maybe ask Kodak?
Have a question about something, from how to get stains out of cloth to how to configure SQL Server, etc. Okay. Do a Google search and find blogs, fora, PDF files, etc. Or go to Wikipedia, Quora, or HN. Used to be had to drive to a library, maybe even a university library. Anyone mention 'productivity'?
Ask the USPS about 'productivity': They have discovered that the Internet has replaced a significant fraction of their mail volume, so significant that they are in financial trouble and having to close some post offices.
Work in some technical field with a lot of math, say, math, physics, engineering? Try to get the typing done? Hope you never had to do that; commonly the typing was more work than the technical work. Now with computers, Knuth's TeX, laser printers, and PDF, the typing is reasonable and routine. The productivity gains here are some factors of 10; for technical writing, we're talking another Gutenberg.
Gotten a printed bank statement in the mail lately? I get my bank statements as PDF files over the Internet. 'Productivity' anyone?
It used to be that a bank had to physically sort the checks and have someone visually read the check and manually key in the amount. Not anymore! Now at a retail site, the 'check' is mostly just nonsense, just my receipt, and the transaction is all electronic. The amount is keyed in at the cash register for the first and last time. 'Productivity'? anyone?
My driveway is steep, and last winter I needed tire chains to get out but had no tire chains. But the Internet worked! So, I ordered some tire chains from Amazon, got them, put them on, and drove out! I didn't have to get towed out or drive to an auto parts store for the chains. We were talking about 'productivity'?
Intel is now down to 14 nm line width (Atom processors), and has stated that they see their way clear to putting 1000 cores on a processor and having them work together effectively. We were talking about what, 'productivity'?
Let's see: The key to productivity is automation. Now the key to automation is computer control. Computer control is based on computer hardware, which keeps falling rapidly in price, and computer software. The ability to develop software has increased enormously with progress in, e.g., Windows and .NET. We just now have the computers, disk storage, operating systems, middle-ware, languages, libraries, networks, and Web services for some astounding progress: We should be at the beginning of by a wide margin the greatest increase, in percentage and absolute terms, of productivity in all of history.
Now what was it we were talking about, hmm, 'productivity'?
There is another point on my side: It is commonly accepted that what children do is a mixture of both 'nature' (genes) and 'nurture' (that is, what they learned from their parents). There are common, strong suggestions that a daughter of a women who is really happy being a mommy is more likely to want to be a mommy. So, net, for a woman to WANT to have children can have come down from nature or nurture.
In my experience, some women REALLY want to have children, and some women REALLY do NOT want to have children. When I was in the ninth grade, I met and dated a girl in the seventh grade, natural blond, great figure, very sweet, who REALLY wanted to have children, knew it, admitted it. E.g., each time she saw the face of a baby with its common characteristics, she wanted to have a baby. Now, I'm not saying that she wanted to have a baby there in the seventh grade, but she wanted to be a MOMMY eventually. And she was, eventually.
And I've known women who had children or not and, in both cases, looked at motherhood as "giving up the best years of their lives and their careers doing low grade, menial scut work to take care of some MAN'S [bitter resentment and anger about the man] children". Sure, the position was partly just negotiations as "don't throw me into that briar patch" but also was meaningfully genuine.
Net, now, some women WANT to be mommies, and some don't. Does this situation inherit, via nature and/or nurture? My observations say that it does. My observations are not very good data, but if the observations are roughly correct then we conclude that the effect, even if weak, could be how Mother Nature will "find a way".
More generally I have confidence that Mother Nature will 'win' this one, will meet the challenge, will "find a way"!
While I believe that the 'way' I mentioned, that women WANT to have children, and that this will get passed from mother to daughter, is the one Mother Nature will select, there can be other ways. E.g., a society can get rid of anything like US Social Security and, thus, tell people in stark terms that they need to have children as a source of support in old age. A society can tell girls that it does not support single women or single mothers and, net, girls should look for good husbands or hope that they have good brothers or have good inheritance from their fathers.. There can be a role for some traditional religions that strongly emphasize 'traditional family values' and having children, whether the women want to or not. Some people might say that the result will not be a change in the gene pool but in the popularity of some religious, social, and/or political values.
Or one might just say that the 'Darwinian' effect will apply to whole societies, e.g., Finland: Finland beat Sweden, Russia, and Germany in wars but is losing out to the 'feminists'! Whatever the politics, at 1.5 children per woman, Finland is on the way out of it as surely as some genocide. So, maybe it's countries that will die off instead of people with the wrong genes. Still, I trust the influence of genes more than that of politics.
Your issue of "N generations on" is interesting: In part my analysis is saying that the women were much the same 40,000 years ago, which would be about N = 2000 generations! So, it would seem that we could extrapolate a few generations from the N = 2000 generations! But, in part my analysis is saying that, if you will, the current, big 'environmenal' changes, i.e., from women having jobs, will bring the biggest changes in the gene pool in 40,000 years and bring the changes very quickly, in, say, just a few generations, say, just 100 years. My reading of evolution is that, when the 'enviroment' changes, then genes can change very quickly. E.g., the genes in some of those weak, sick, or dead limbs on the tree are already gone forever.
(1) In the past, e.g., in tribal or agricultural communities, women had children whether they really 'wanted' to or not.
(2) Now in more industrialized countries, a significant fraction of women have some options. Some women still have children but some women do not want to have children and do not.
(3) The change in number of children per woman is a 900 pound gorilla in the room: The article mentioned numbers under 2 children per women with a rate a low as 1 child per woman. There is some recent data that Finland is at 1.5 children per woman.
Of course, for any rate much under 2, each 20 years or so the population will be going down significantly. "Get your old houses, furniture, dishes, baby clothes, etc. cheap, cheap, cheap!".
So, net, heavily women who don't want to have children won't. These women will be 'weak, sick, dead limbs on the tree' and will be pruning their genes from the tree. What will be left are women who, given the choice, actually, effectively WANT to have children.
The big point: After a few such generations, we will be left with a much smaller population with many fewer women but nearly all of whom WANT to have children. Then the population will start growing again.
One more big, surprising point: We are now, in much of the world, in the most rapid change in the gene pool of the last 40,000 or so years.
Where did the 40,000 come from? Humans walked out of Africa back there somewhere. At one point, ballpark 40,000 years ago, they reached, say, India. One branch went west to Western Europe, and another branch went east to China and Japan. Mostly the two branches haven't much mixed since then.
Okay, at the common branch, ballpark 40,000 years ago, what were women like? Well, take women from Japan and women from Western Europe. Take some 'characteristic' in common, say, desire of a major fraction of the women, given an economic opportunity, not to have children. Now, start in either Japan or Western Europe and count genetic 'changes' on this 'characteristic' going backwards in the tree to the common branch, about 40,000 years ago, and, then, continuing to count changes, going forward in the other branch of the tree to the present.
So, if on this characteristic the women in Japan and Western Europe are close, that is, have few changes, then on this characteristic the common ancestor 40,000 years ago has still fewer, that is, is closer to the women in both Japan and Western Europe than they are to each other.
Net, since a significant fraction of women in both Japan and Western Europe will, given an economic opportunity, choose not to have children, that is, these women are close to each other, both are still closer to their common ancestor 40,000 years ago. So, for 40,000 years, many women had children not really because they 'wanted' to but because of economic necessity.
So, now that women who don't really want to have children are being pruned from the tree, we are, on this characteristic, in the most rapid change in the human gene pool of the past 40,000 years. And, the change is VERY rapid, should have a huge effect in just a few generations, say, just 100 years, which on the scale of 40,000 years, is FAST.
The main drive in the economy is more productivity; the main approach there is more automation; the main approach there is computer hardware driven by software.
Next, the main point about software is that it be 'smart' enough to give especially valuable output. For that the main tool is math.
As in this thread, two of the influences are (1) employers want some 'criteria', maybe even 'credentials', that will simplify selecting promising employees and (2) students seeking jobs want to keep down the costs in time, money, and effort to meet such 'criteria' or to get such 'credentials'.
Broadly there should be some 'market forces' that provide answers: (1) If the education employers want is expensive, then employers will have to pay employees enough to pay for the education; (2) if the education costs more than it is worth to the employers, then students and employers will make do with less such education; (3) if students don't come for the education, then some educational institutions will have to make some changes to offer the students more value for the costs.
Such 'market force' influences are easy to see, but there are some other large influences less easy to see:
In the US, the 900 pound gorilla is the interest of Congress and the US DoD in technology for US national security.
Such was not always the case: Indeed, during the rapid rise of engineering in the decades before WWII, schools of engineering actually concentrated on teaching engineering for students seeking careers in engineering! Amazing! Radical! Astounding!
Then we had WWII and radar, sonar, and the bomb, right, the atomic bomb, and Ike and other influentials concluded that "Never again will US academics be permitted to operate independently of the US military" or some such. Congress went along, that is, pulled out the US national checkbook, and started signing.
Then the top three dozen US research universities got an offer they couldn't refuse: Take the US Federal money for research in math, physical science, and engineering, with a tilt toward what might be useful for US national security, or cease to be a leading research university. They took the money!
Now, for such a university, for the money, the most important activity on campus is research, research, and research as in "How do I get a grant?", especially, now, from the NSF, DoE (that is, energy, not education!), or DARPA. Since then Congress has also provided money to the NIH and its grants -- you see, there are a lot of old people in Congress eager to see progress on some of the serious diseases of old age and ..., well you can see the connection!
Now for this 'research', there are some examples and, now, a 'model': The most influential example, for both science and US national security, is physics, especially as for the bomb, right, the atomic bomb and now, too, the hydrogen bomb. So, 'research' has a really severe case of 'physics envy', especially theoretical and mathematical physics envy. So, good 'research' is supposed to 'mathematize' a field.
So, at top three dozen US research universities, math, physical science, and engineering pursue research, research, and research with physics envy.
Yes, there was that report, yes, the David Report, that said that some of math was, well, not so applicable so soon and should, then, maybe get less money. That's why now at some of the research universities the math department has trouble keeping the lights on but is still working on the analytic-algebraic topology of the locally Euclidean metrization of infinitely differentiable Riemannian manifolds (extra credit for the source!). Math is still the most respected field, but for some decades more applicable topics in math -- e.g., probability, stochastic processes, optimization, statistics, control theory, signal processing, numerical analysis, computational fluid dynamics, computational complexity, math for finance, math for theoretical physics -- has been done outside the 'pure' math departments.
Now it turns out that somehow some huge fraction of good students are really eager to get their bachelor's degrees from such research universities. So, the universities can be very 'selective' so that employers can get some of the 'criteria' they want just by a student being admitted to such a university!
But the students are being taught mostly just by professors interested mostly just in narrow, leading edge research to get grants. So, the education is not really about, say, engineering for engineering students who want jobs in engineering! Instead, in, say, computer science, the education might be closer to background for research in 'the fundamentals of computing', not that we really know what they are but some people would like to!
How to work through the security model of SQL Server for SQL Server installation, administration, and management? Professors can't get grants for mud wrestling with the messed up SQL Server security model or its just awful documentation and terrible problems with installation, management, and administration and, thus, mostly DON'T! So if someone actually wants actually to actually work with an actual, real installation of SQL Server for a real, actual, important, practical database application, in the real world, outside of academics, as part of a career, that can pay enough to buy a house and support a family, then the professors in a research university are not a very direct sources of such information!
Still good students like to go to research universities.
Then there is the money, that is, what the universities charge. So, tuition has gone up, way up, over the last few decades, up faster than even health care! Why? Well, there's a dirty little secret! What has gone up is the published, 'list price'! But there are also 'discounts', especially for good students, called 'scholarships'! So, if the student's father is wealthy and the student only so-so as a student, then go ahead and charge list price! While this student pays list price, they can also do well in their studies in beer and bed! But for a really good student, especially with poor parents, there are scholarships.
In the US, there are also many colleges and universities that don't try to be top research universities and concentrate on teaching.
And there are many community colleges where 'job training' is the goal and not a dirty word and where there are courses in auto repair, auto body repair, framing carpentry, finish carpentry, masonry, cosmetology, plumbing, electricity, HVAC, and, yes, computer programming and network management. Tuition is low; the courses are fully intended to be practical; the teachers are not researchers and are often practitioners.
For computer science, there is a secret: The more advanced parts of computer science have been heavily 'mathematized'. So, net, the best background for such material actually is not even in the computer science department but in the math department, especially a course in, say, abstract algebra.
Next, for a career, e.g., in computing, there is now a big, dark, ugly secret: 'Jobs' are no longer such a good idea! E.g., the Stanford AI course got interest from 50,000 students in 175 countries! So, generally, if take some material in college and look for a job, then here in the US, both you and your employer will be competing with thousands of eager people in 175 foreign countries!
Meanwhile here in the US the people buying houses and getting their children through college will, may I have the envelope, please? Yes, here it is: Often they will own their own Main Street business where they have a geographical barrier to entry and, thus, no competition more than, say, 100 miles away. In particular they will have no competition from anyone in any of those 175 foreign countries.
So, what future for computing in the US? Broadly, gotta do something new, out there, on the leading edge in at least some respect, something entrepreneurial, something you can't get hired for because the guy hiring doesn't understand that new thing yet. So, have to be an entrepreneur. For that, something in advanced computing might help. Then, just getting 'skills' with, say, Linux, C++, Java, Python, MySQL, SQL Server, Flash, HTML5, etc. is 'routine', maybe at some point necessary but not sufficient.
Just what should such new stuff be and just how to get help from a research university? No one really knows! Welcome to the challenge of the future!
E.g., it's not the least bit clear that finance needs to attack the worst case problems. Even if so, finance, anywhere close to reality, doesn't need to get solutions that save all of the last tiny fraction of a penny.
If relax worst case problems and saving the last tiny fraction of a penny, then the situation is MUCH different. E.g., once I got a feasible solution to an ILP with 40,000 constraints and 600,000 variables within 0.025% of optimality on a 90 MHz Intel processor in 905 seconds. More generally, here's an easy approach to the ILP problems: Just drop the integer constraints, solve with LP, and then round to integer values. For the LP solution, if insist, then use a polynomial LP solver. Crude? Sometimes. Always crude? No.
He's making a common mistake: He's assuming that because in general ILP is in NP-complete that good work on all practical ILP problems has to be too challenging, and that's nonsense.
You are correct on both points. Early in my learning, I found both points to be important. And, for a beginning student, both points can be crucial: Without the second point, it is far too easy for a student to get into some not very good material or into some good material but, still, lost.
Eventually I got away from your second point.
Still, there is a version of your second point that lasts: For research, seminars and conferences are good, fast ways to keep up, see the forest for the trees, pick new research directions, etc.
When my company is successful and I retire, I will return to mathematical physics and, maybe, attend research seminars in, say, Boston.
If Boyd has some nice engineering applications, terrific!
Yes, I can believe that the Stanford alums have especially good reason and ability to be 'grateful'! Good for Stanford; thank you, thank you, thank you Dean Terman!
Still, nearly all the world's top research universities are in the US, and the main reason is that right after WWII Ike and others concluded that research in math, physical science, and engineering were just crucial for US national security. Congress has gone along. Any significant claim that the US is falling behind in research will open Congress's funding for the usual suspects NSF, NIH, DoE, DARPA, etc.
Net, no way, not a chance, will Congress permit US research universities to fall behind: From Congress all it takes is money, that is easy for Congress to vote, and actually in the big picture it doesn't take very much money.
Also, all things considered, the money is relatively well spent. Evidence: Getting those grants is DARNED competitive. Want to get promoted at a top US research university? Get a grant!
And there is a secret -- hush, hush! About 60% of the grant money, for math, physical science, and engineering, goes to 'overhead' which ends up good for the English, modern dance, French literature, and art history departments! Likely helps pay for the university theater, art gallery, string quartet concert series, etc.!
Don't tell anyone that Federal taxes are supporting string quartets!
Also looked to me that the Stanford lawn was in nice shape!
Convexity is an important concept, and the world is just awash in rock solid treatments.
So, for some positive integer n, consider n-dimensional Euclidean space. Suppose C is a closed, convex subset and point x in the space is not in C. Then there exists a closed half space that covers C and omits x. The boundary of this closed half space can share a point with the boundary of the closed half space so that the plane of the half space is 'supporting' for C. So, this is a 'separation' result. Yes, it generalizes past finite dimensional Euclidean spaces! Yes, it's a darned useful theorem! E.g., can ask for the point in C closest to x -- the point has to exist since C is closed. Then the plane through this point and perpendicular to the line to x is supporting for C. So, can get a nice projection result and use it for optimization and best approximation.
Okay, my point: This result is one of the most important about convexity but is standard in 'analysis'. I first saw the result in the 'advanced calculus' book, Fleming, 'Functions of Several Variables'.
Bigger point: Convexity pops up frequently in analysis. E.g., there is Jensen's inequality. With it can easily knock off a nice list of otherwise difficult to prove inequalities. E.g., in the L^p spaces, we can use Jensen's inequality to get Holder's and Minkowski's inequalities.
When we start with optimization, we should start with convex sets and also convex functions. E.g., there is a nice list of theorems of the alternative that are separation results for cones and polyhedra that can be used to establish some otherwise tricky results about duality in linear programming and are also key to the Kuhn-Tucker conditions in non-linear programming. Of course, the feasible region in linear programming is a convex set.
In non-linear programming maximization, there is a non-linear dual that is to minimize a convex function, and that can be the core of the powerful technique of Lagrangian relaxation.
When we have a norm on a vector space, the locus of all points distance 1 from the origin is convex.
In facility location we can be minimizing a convex function and doing so by constructing a sequence of planes supporting the hypergraph of the convex function.
Alas, I never heard a lecture from Boyd!
There are plenty of rock solid, highly polished books where the role of complexity is made clear!
"#1: Be Narrow
"Focus on the smallest possible problem you could solve that would potentially be useful."
has
"You can much more easily position and market yourself when more focused. And when it comes to partnering, or being acquired, there's less chance for conflict."
Cute. Hadn't thought of that one.
For his
"#2: Be Different
"Ideas are in the air. There are lots of people thinking about—and probably working on—the same thing you are."
So he's saying to give up on trying for something really "different" and, thus, is contradicting his title for his #2.
That's once!
So his #2 is a bit 'confused' or 'mixed' in its message about being "different". The confusion continues with his:
"Second, see #1—the specialist will almost always kick the generalist's ass."
So, now he seems to be saying to be more specialized and, thus, "different". Okay.
That's twice.
Then he has:
"Third, consider doing something that's not so cutting edge."
where he seems to be back to being not so specialized or different.
That's three times he's 'flip-flopped'!
Here's my hot button issue about such lists:
Yes, it's true in maybe 99 44/100% (borrowed from the label of old Ivory soap) of the cases:
"There are lots of people thinking about—and probably working on—the same thing you are."
And two other things are true:
First, maybe 99 44/100% of information technology (IT) startups will never be worth over $1 billion.
Second, there are so many startups that have hardly better aim than blind shots in the dark that maybe over half of successful startups were such blind shots in the dark that just got lucky. Yes, between being smart and being lucky, being lucky is better. Sadly, we can't order up luck on demand. So, for an advantage we are stuck trying to be smart.
So, in trying to plan a successful startup, let's set aside luck.
Next, in trying to be worth, say, over $1 billion, we will be doing something quite rare. So the fact:
"There are lots of people thinking about—and probably working on—the same thing you are."
is not very impressive, compelling, or convincing for our planning.
That is, for our rare, desired success, we should notice the advantage of the old advice to:
(1) Pick a problem without a good solution and where a good solution would be very valuable. E.g., for a Web IT startup, provide something that hundreds of millions of people want and will like much better than anything else. In a biomedical analogy, have a one pill (safe and effective) cure for cancer.
(2) Construct this product or service with an 'unfair advantage', that is, some crucial, core 'secret sauce' that is the key to providing something "much better than anything else". In a biomedical analogy, use some work in biochemistry to construct a one pill cure for cancer.
(3) Have the 'secret sauce' be difficult to duplicate or equal. In a biomedical analogy, use some powerful, difficult, original, advanced work in biochemistry to construct a one pill cure for cancer.
So, when following (1)-(3), is it fair to say:
"There are lots of people thinking about—and probably working on—the same thing you are."
HECK NO!
(A) Are (1)-(3) commonly pursued in IT? No.
(B) Do (1)-(3) represent a major fraction of successful IT companies? No.
Do (A)-(B) mean that (1)-(3) should be ignored? NO!
Can (1)-(3) bring advantages? Should be obvious that they can.
Remember: The successful companies are heavily from luck, only a tiny fraction of startups can be worth over $1 billion, and the world keeps changing and moving on so that what most startups are doing and even what many successful startups did are not very good evidence for what to do to be worth over $1 billion!
There is one more hot button issue: There is advice, say, from Steve Blank, that we have to keep going back to the customer to confirm that the customer will like the product or service. But, for a safe, effective one pill cure for cancer, what do we ask the customer, what shape and color they want for the pill and what color they want on the pill bottle label?
So this importance of going back to the customer over and over means that the product or service being provided is less desirable than a one pill cure for cancer.
So, if possible, quit asking customers about what color they want in the pills and return to the work of constructing the one pill cure for cancer.
Can we get something like a one pill cure for cancer in IT? Hmm ...!
E.g., in 'research', how do you think the necessary background material is obtained? By signing up for 'courses'? Nope! Instead, the researcher gets the background materials, learns what is needed, and then gets on with the research. A researcher can't get paid just for getting the background material; instead, the researcher has to just do that seemingly effortlessly! Well, students, once they reach a suitable level of educational 'maturity', can too!
E,g,, here on HN are many thousands of programmers who use computer hardware and software they never saw in a course. So, HN readers learned on their own. Such self-education has long been a central feature of the US computer industry.
Net, researcher, HN reader, or serious student, all can learn solid material without a course, and do.
Actually, learning outside a course is sometimes in academics considered almost 'cheating'! So, if a course is to be a 'filter', a competition, a 'stress test', then for a student to have worked carefully through solid texts on the material before the course lets them in the course effortlessly blow away all the students who started with the material only in the course. So, can get comments such as "Best student in the class by a wide margin, but apparently knew most of the material before the course.". So, there's commonly no doubt that it's possible to learn the material outside the course and to learn it well enough to lead the class in a corresponding course. Indeed, such self study can move forward at the student's own pace, faster some places, slower others, and, net, learn without gaps or wasted effort.
Net, course or not, the actual learning is mostly or entirely from what the student does with the material alone in a quiet room.
Of course, again, to get this good situation of self learning, need solid material clearly presented in, say, a book. Some subjects, sadly, are so vague that no such book can exist; in that case, maybe a course is 'needed', i.e., so that a student can soak up nonsense that couldn't be written down in solid form! For such, wander over to one of the 'humanities' departments and take one of the courses in 'political correctness'!
We have AI as a field mostly just because of long, aggressive funding by DARPA, early on especially of the MIT AI Lab. Since then DARPA has funded the prize where computer controlled cars chase across a desert. It appears that some people at DARPA are hoping that something good happens from their funding. Tough to say something good will never happen! If something okay leaks out of the research, then some lives of some US soldiers might be saved and, by now, may have already. In that case, good. DARPA is likely some of the very best money spent, even if for some years the work looks weak. At least with the money so far, a lot of quite bright people have TRIED to make progress in AI.
Some years ago, I was one of three researchers in an AI group that did some of the world's best AI work. I had two conclusions: (1) For the practical progress desired now, just use 'traditional' approaches of (A) see problem, (B) analyze problem, (C) find assumptions can make about the problem, (D) do some math starting with the assumptions, (E) take the conclusions of the math as progress on a solution, (F) use the best of what have to wrap up the best solution can see. (2) For the AI part, so far "We dance 'round in a ring and suppose while the secret sits in the middle and knows.".
Maybe 60% of the budget of Stanford is from research grants from the US Federal Government. For the students, there are also scholarships, e.g., from the NSF, NIH, etc., which is more "government" support.
Be not at all in doubt: Look at the Stanford campus with all its buildings, lawns, people, equipment, etc. and know that nearly all of it is paid for by the US Federal government.
So, the core really isn't Gaussian elimination for systems of linear inequalities.
For more, LP and the simplex algorithm are surprisingly powerful: Part of the reason is, for a huge range of optimization problems, taking a local linear approximation and then attacking that with simplex is often the best technique known. There are also surprising connections with combinatorial optimization and some classic NP-complete problems. What LP and simplex do on networks is also surprising and powerful; e.g., there simplex takes on a special form based on spanning trees.
I'm exaggerating, but, still there is a reasonable point here. CS is about some 'science', and call that the mathematical part where we have some solid material worth being called 'science', For the rest, call that 'computer practice' or some such.
The upside of my view is that for some serious progress we're going to have to use some serious methodology. So, I'm proposing if not mathematical physics envy then applied math envy. Applied math didn't take on how to design the display lights in a scientific pocket calculator although without the lights the thing wouldn't work.
My view is not the most extreme: Last year I communicated with a CS prof whose position was that CS is looking for the 'fundamentals of computation'. Hmm .... It sounds like he believes that the P versus NP question should be right at the top of the list, and I don't. I'll settle for anything that is solid and a contribution, even if small, to the 'science' and not just to current practice.
LIkely many people here know the current state of programming language research much better than I do. If that field has gotten nicely mathematical with some solid material, great. The progress from Fortran, Cobol, Algol, Basic, PL/I, Pascal, C, C++, etc. was pragmatic and important and of enormous value to the economy but not much progress in a 'science' of programming languages, and that progress, with poor methodology, has slowed as we might have expected.
Maybe I'm saying that, in 1920, if we wanted a really good airplane, then maybe we should set aside the wood, linen, and glue and go do some aerodynamiocs calculations, discover Reynolds number, and discover that those really thin wings were a mistake. Or, observational astronomy was just a lot of curiosity until Newton came along and made progress in understanding the universe. The practical chemists had discovered a LOT, but by applying quantum mechanics they made HUGE progress.
If we are going to make the huge progress we want in computing, then history suggests that we can't be just pragmatic and that "theory is under-rated". We can't expect that chemistry will do much to help CS, but the obvious tool is math, to turn CS, the science, part into some applied math.
I nailed them on their description that the simplex algorithm was like Gaussian elimination for systems of linear inequalities.
No, The simplex algorithm is selecting elementary row operations to improve the value of the objective function.
Gaussian elimination also uses elementary row operations, but that is about the end of the direct connection. In particular, how Gaussian elimination selects elementary row operations is quite different from how simplex selects them.
For more detail, I explained the role of artificial variables where we just 'glue on' an identity matrix. So, since we just glue on an identity matrix, we never, directly in simplex, go through anything like Gaussian elimination to GET an identity matrix. Yes, at times, for numerical reasons, one should 'reinvert' the basis in simplex, and here Gaussian elimination, etc., might be used, but such 'reinversion' is not really simplex.
The authors mentioned Gaussian elimination when they should have mentioned elementary row operations. Next, what is special about simplex is improving the objective function, not working with linear inequalities. Indeed, simplex doesn't work with linear inequalities; instead, we just use 'slack' and 'surplus' variables to get a system of linear equalities (with non-negative variables) before simplex even gets the problem.
The authors were being sloppy. I called them out on it. It was appropriate that I do so.
If you believe that my remarks about simplex are in error, then by all means make your claims.
But the issue is not me; it's the book.
"I regret to say that that's not as easy as it sounds. There are issues of tractability, and more importantly, there's this not-quite-cottage industry of software development, which needs good-enough languages to use _today_."
I didn't say it was "easy", and clearly my point is progress and not current software development. Sure, for my project, I'm typing in code in Visual Basic .NET. For writing code today, all things considered for my context, that is about the best option.
But what I'm typing in is hardly different from what anyone might have typed in all the way back to Algol.
Here's the point: People type in all that code, and then what? Can some software go over it, report properties, do some transformations with known, useful properties? Not very much. Such things won't be better for Visual Basic than they were for Algol or anything between.
So, they type in the code, and then all they have is just that code. They can desk check it, test it, try it, revise it, etc., but it's all still digging the Panama Canal with a teaspoon one canal at a time. That is, there's no real automation and no significant exploitation of mathematical properties that could lead to automation.
Or, return to most of the rest of engineering where we have specifications of resistors, capacitors, inductors, transistors, fans, copper wire, steel sheets, aluminum bars, etc. with ohms, farads, density, tensile strength, electrical conductivity, etc. Okay, now compare with some DLL or its source code: What engineering properties do we have. Essentially none. It's like building with iron before we knew about tensile strength.
Sure, with some such research progress, eventually programmers would benefit. Obviously the intention is real progress in software productivity. We're not going to get such productivity by more of the same that got us C, C++, Java, Python, Visual Basic .NET, C#, etc.
Also for another of your points, I'm not talking about anything like branch prediction or deep pipelining. That's basically what to do with with the hardware to execute an existing instruction set.
Also the point is not math or not. The point is progress. Math will be necessary but not sufficient. That is, just stuffing in some math won't yield more than the old trick of putting two extra, unused transistors in a radio to claim nine transistors instead of seven.
Your point that there's a lot of good CS to do without 'mathematizing' the field is not promising for research or significant progress.
You are not concerned about the book but are torqued at me for some of what I said about the book. You took what I said personally, as if I was mean to your favorite baby.
You need to grow up.
For the CS topics you listed, e.g., programming languages, that field has for decades just cried out for some good math but gotten nearly none! Clearly compiling and executing a program are necessarily mathematically something, understood or not, powerful or not. For progress, we need to understand the subject mathematically. E.g., given a programming language statement, what are its mathematical properties so that given a sequence of two such statements, what are the mathematical properties? Can't see any? Okay: Change programming languages until can. Then with the properties, what can we conclude about the program? Anything useful, say, about correctness, running time, storage utilization, approaching limitations? Given some properties, can we have some code transformations with some known properties? Are some such useful? If not, can we define some useful programming languages that admit such transformations?
There are some simple, illustrative, pregnant cases that just leap off the screen: Write a collection of routines for sort by surrogate, i.e., finding a permutation, maybe honoring sort 'stability', applying a permutation to an array, chasing through the fact that each permutation is a product of disjoint cycles, etc. Then with such routines, will see some properties that are close to algebraic, e.g., where some combinations of the routines are the identity transformation, where one routine is the algebraic inverse of another, where we might have commutativity and/or associativity, or are equivalent to other combinations, etc. Can we preserve sort 'stability'? So, we see that we should have an 'algebra' of those routines and be able to derive known properties. Then we see that due to 'boundary' limitations, etc., our algebra is only roughly correct. So, maybe we should fix that.
This is old stuff. It has leaped off the screen for decades. We want some known properties we can use to derive new, useful properties. Instead all we get is programming languages that compete in some absurd 'beauty contest' or that 'encourage' programming 'styles' seen as 'helpful'. Sicko.
E.g., it has been common now in server farms to be concerned about reliability. So, we set up systems so that when system A gets too busy or fails, then system B takes over, etc. However, what we've seen, including at some of the most important farms, for decades, up to this year, is that these intuitive approaches far too easily go "splat" in the mud: They are 'unstable', have problems that propagate, and are surprisingly UNreliable.
So, we need some new math of 'reliability' for such server farms, math that can write us some guarantees. That math will start with the real problems, have definitions, theorems, and proofs, and, if successful, some solid tools for building reliable server farms.
All across computing, CS needs to give us progress, and for that they need to proceed essentially only mathematically.
It's the math. The CS is the math. Server farm reliability is just the application of the math and is not the CS. Ignoring math, CS is "A little people. A silly people."! Until CS gets serious about the math, it is writing checks its methodology can't cash.
To get serious about math, CS first has to learn LP well!
The would you believe probability? How about stochastic processes? Any stochastic processes in server farms and networks? Any uses of the axiomatic derivation of the Poisson process? Any roles for one of the strongest inequalities in math, the martingale inequality and the associated convergence theorem? Measure preserving transformations as in ergodic theory -- I'll answer that one, "Yes!". The renewal theorem and loads at server farms? How 'bout a little in probability?
Understanding now?