Interesting interview on the state of science

[Jedi_Master]

If gturner is saying we need more funding for blue sky research, well, duh. I don't think you'll find anyone here who'd be against that.

I don't think it's a coincidence that what public science funding the US does have is "results-oriented," as he put it. No kidding. Corporate lobbyists have themselves pushed the dialogue in that direction and gotten sympathetic politicians to cut funding for pure research. Once again, gturner makes the claim that inept government is the problem, when in reality it's corporate interference with government spending that's at the root of all this, aided and abetted by conservative politicians who think pure research is a waste of money.

I was unaware that corporate lobbyists controlled all the spending at universities and government research institutions.

It's not that the government is inept (thought it generally is), it's that it's going to be using much the same system it uses for Department of Defense grants. When they're not looking for bang for the buck and quick results, they're trying a mega-project like a space telescope or a fusion reactor.

A few months ago I was chatting with the head of facilities at our local university, and he talked about how it was his job to support the researchers who are bringing in the biggest grants, since the university has to make a profit by taking a cut of what they bring in. In the case he was complaining about, the school was trying to save money by being green and controling the temperature of the research buildings to save fuel, but some of the researchers were doing experiments whose temperature couldn't be allowed to vary that much. This could cause some of the more profitable researchers to leave (and we're talking pure government grant money, not corporate), which would leave offices vacant and thus non-productive.

It's like the university is farming or share-cropping or something, providing space, power, and communication services in return for a percentage of the grant harvest.

What the hell are you talking about? Government grants are one of the best resources for a "pure" researcher, and often the government is the only source for grant money for research projects that don't have a clear research goal or are not immediately profitable.

Your anecdote doesn't mean anything either, other than show that the university Regent's Board likes to skim off the top.

You still have not offered any possible solution.

 

[gturner]

Maybe for the "pure" researcher in Asian lesbian film studies to show that society is oppressed by the white male patriarchy, or the disease crisis of the day, or a spanking cool new way to blow people up, but other than that, good luck if you don't write up a sparkling and rather dishonest submission.

The key flaw in relying on the government to provide all the research dollars at the right places is the same flaw it has in allocating economic activity to the most efficient places. As established generations ago, this presupposes that a few people can be as informed on all subjects as millions of people, which is the fundamental information problem that's allowed capitalism to far outstrip socialism. Having a small number of old failed scientists (otherwise why would they be wasting their brilliant minds reviewing reams of grant applications instead of going after Nobel prizes) trying to understand what a hundredfold more smart young scientists are thinking doesn't really work any better than Soviet efforts to improve the potato harvest.

That system doesn't remotely work in economics (central planning is an abysmal failure that just produces mass graves, endemic poverty, and starvation), so why should it work with an even sketchier field like innovation, much less ground breaking innovation?

Geniuses are almost impossible to predict, much less to manage, much less to turn into cogs in a big funding wheel. We've somewhat managed to do that, just so they don't create as many disruptive innovations, but this slows the pace of innovation, especially radical ones. That was much the point of shunting really bright thinkers into studying bees and fruit flies, where they couldn't cause many problems. The downside is that they've managed to institutionalize not causing many problems by making disruptive innovations.

Long ago Europe faced this problem regarding sons who weren't the first-born, and those were shunted to the military or clergy. We've gone one better, making them prostrate themselves before Kafkaesque committees to beg permission to study foot fungus, the societal implications of jogging shoes.

The elites make tent-fingers and say that all has unfolded as they have foreseen, and laugh demonically.

There are alternatives, and I offered one to Judith Curry yesterday. She said "sign me up." Unfortunately I now have to round up undergrads who are willing to work as lady's maids, footmen, valets, and cooks.

 

[Robert Maxwell]

If all you're going to do is string together a bunch of wacky non sequiturs and call it an argument, maybe you shouldn't bother.

 

[Gov Kodos]

So the argument is science doesn't work like [yt]http://www.youtube.com/watch?v=wTP2RUD_cL0[/yt]

 

[Jedi_Master]

Maybe for the "pure" researcher in Asian lesbian film studies to show that society is oppressed by the white male patriarchy, or the disease crisis of the day, or a spanking cool new way to blow people up, but other than that, good luck if you don't write up a sparkling and rather dishonest submission.

The key flaw in relying on the government to provide all the research dollars at the right places is the same flaw it has in allocating economic activity to the most efficient places. As established generations ago, this presupposes that a few people can be as informed on all subjects as millions of people, which is the fundamental information problem that's allowed capitalism to far outstrip socialism. Having a small number of old failed scientists (otherwise why would they be wasting their brilliant minds reviewing reams of grant applications instead of going after Nobel prizes) trying to understand what a hundredfold more smart young scientists are thinking doesn't really work any better than Soviet efforts to improve the potato harvest.

That system doesn't remotely work in economics (central planning is an abysmal failure that just produces mass graves, endemic poverty, and starvation), so why should it work with an even sketchier field like innovation, much less ground breaking innovation?

Geniuses are almost impossible to predict, much less to manage, much less to turn into cogs in a big funding wheel. We've somewhat managed to do that, just so they don't create as many disruptive innovations, but this slows the pace of innovation, especially radical ones. That was much the point of shunting really bright thinkers into studying bees and fruit flies, where they couldn't cause many problems. The downside is that they've managed to institutionalize not causing many problems by making disruptive innovations.

Long ago Europe faced this problem regarding sons who weren't the first-born, and those were shunted to the military or clergy. We've gone one better, making them prostrate themselves before Kafkaesque committees to beg permission to study foot fungus, the societal implications of jogging shoes.

The elites make tent-fingers and say that all has unfolded as they have foreseen, and laugh demonically.

There are alternatives, and I offered one to Judith Curry yesterday. She said "sign me up." Unfortunately I now have to round up undergrads who are willing to work as lady's maids, footmen, valets, and cooks.

Let me sum up your response

Paragraph 1:Idiotic bullcrap completely fabricated in your mind

Paragraph 2:More idiotic bullcrap, this time with random historical "references"

Paragraph 3: Horrifically idiotic bullcrap that sounds like you are quoting a dude trying to pick up a chick at bar by being "intellectual"

Paragraph 4: You really have no clue about how the American educational system works do you?

Paragraph 5,6: were you high went you wrote this.

Paragraph 7: You are proof that some people simply should not be allowed to have a computer with Internet access.

 

[Edit_XYZ]

Maybe for the "pure" researcher in Asian lesbian film studies to show that society is oppressed by the white male patriarchy, or the disease crisis of the day, or a spanking cool new way to blow people up, but other than that, good luck if you don't write up a sparkling and rather dishonest submission.

The key flaw in relying on the government to provide all the research dollars at the right places is the same flaw it has in allocating economic activity to the most efficient places. As established generations ago, this presupposes that a few people can be as informed on all subjects as millions of people, which is the fundamental information problem that's allowed capitalism to far outstrip socialism. Having a small number of old failed scientists (otherwise why would they be wasting their brilliant minds reviewing reams of grant applications instead of going after Nobel prizes) trying to understand what a hundredfold more smart young scientists are thinking doesn't really work any better than Soviet efforts to improve the potato harvest.

That system doesn't remotely work in economics (central planning is an abysmal failure that just produces mass graves, endemic poverty, and starvation), so why should it work with an even sketchier field like innovation, much less ground breaking innovation?

Geniuses are almost impossible to predict, much less to manage, much less to turn into cogs in a big funding wheel. We've somewhat managed to do that, just so they don't create as many disruptive innovations, but this slows the pace of innovation, especially radical ones. That was much the point of shunting really bright thinkers into studying bees and fruit flies, where they couldn't cause many problems. The downside is that they've managed to institutionalize not causing many problems by making disruptive innovations.

Long ago Europe faced this problem regarding sons who weren't the first-born, and those were shunted to the military or clergy. We've gone one better, making them prostrate themselves before Kafkaesque committees to beg permission to study foot fungus, the societal implications of jogging shoes.

The elites make tent-fingers and say that all has unfolded as they have foreseen, and laugh demonically.

There are alternatives, and I offered one to Judith Curry yesterday. She said "sign me up." Unfortunately I now have to round up undergrads who are willing to work as lady's maids, footmen, valets, and cooks.

Let me sum up your response

Paragraph 1:Idiotic bullcrap completely fabricated in your mind

Paragraph 2:More idiotic bullcrap, this time with random historical "references"

Paragraph 3: Horrifically idiotic bullcrap that sounds like you are quoting a dude trying to pick up a chick at bar by being "intellectual"

Paragraph 4: You really have no clue about how the American educational system works do you?

Paragraph 5,6: were you high went you wrote this.

Paragraph 7: You are proof that some people simply should not be allowed to have a computer with Internet access.

'Violence is the last refuge of the incompetent.' Old news.

 

[Jedi_Master]

'Violence is the last refuge of the incompetent.' Old news.

 

[gturner]

Paragraph 2:More idiotic bullcrap, this time with random historical "references"

Hrm... There weren't any historical references in that paragraph, just a note that central planning was an abysmal failure due to the information problem, which was well explained by the 1940's and backed up by history. You should try reading something written on the subject since then, and try to do a better job of summarizing.

Instead of blaming "corporations" for the problems in research funding, which is juvenile, consider how diverse our corporations are. We have tens of thousands of small ones, each trying to come up with new things, and some of which got huge by doing just that and doing it very well, like Apple, Microsoft, IBM, Intel, Google, and hundreds of other innovative giants. Each of those will have their own tricks to allocating personnel or granting funding, and even in the same corporation there will be many multiple methods, some very formal, some informal.

In the field of government defense scientific research, even by the late 1940's some of the labs had developed lots of notes on what methods work and what don't. Researchers at the Navy Weapons lab incorporated many of these lessons in developing the Sidewinder as a completely off-the-books project in its beginnings. It was a secret not just from the Soviets, but also from the management at the Navy Weapons lab, who would never approved it because it obviously couldn't work. As the seeker and missile progressed to the point of letting management in on it (who called it a 'proximity fuse with a tendency to decrease dispersion' because they dared not call it a 'guided missile - unauthorized'), and then progressed further to where Washington was informed, they were ordered to suspend all work on the project because Washington's top expert showed that an IR homing missile was physically impossible. So they mailed Washington the video of a B-17 being blown out of the sky that very morning by an IR homing missile, and Washington's order was rescinded.

I point this out because centralizing the funding bureaucracy in a pyramid of experts reduces the thousands of research project funding and management methods currently in use down to one method, codified in a vast shelf of books that contain pages and pages of rules, any one of which can stop a project in its tracks or not let it even get started. This happens a lot at places like IBM, where they've built up management rules by noted every project failure they've had and penning a new policy to make sure it doesn't happen again. That's not a good way to accumulate rules because every possible way a project might be run will be ruled out because some long-forgotten project that was run that way had failed, and eventually you exhaust the possible ways to run a project. I encountered this many many times at IBM RTP.

Paragraph 4: You really have no clue about how the American educational system works do you?

I'm repeating what the founders of many top-tier and Ivy League schools said as their justification, providing a safer outlet for bright and ambitious men, just like the old choice of joining the clergy. My old roommate and caving partner (who started out in EE and cancer research before going on to John's Hopkins and the Salk Institute before becoming a professor at UCSF) illustrated this with the story of John Hunt Morgan, who killed scads of people, terrorized an entire region (we have a big statue of him in town), and whose extremely powerful family looked set to cause all sorts of political trouble. In contrast, his nephew went into research and won the Nobel Prize in medicine for figuring out how chromosomes control heredity by painstakingly studying fruit flies.

The reason the US has such intense college athletics in comparison to European universities is that many of these same university founders wanted their creations to compete with each other in sports, where they could bet lots and lots of money it.

Apparently universities today don't bother teaching anything about the history of universities (which oddly enough is a word that used to be correctly pronounced as 'univarsity', just like hicks still say it, and why we still have 'varsity' sports.

 

[Jedi_Master]

Wall of text

Here is my challenge to you.Can you provide a solution to the problem you started a thread to discuss?

If so can you make it one paragraph in length, maximum 200 words?

If you cannot do either of those things, then I don't see the value of this discussion.

 

[gturner]

It would just fly right over your head.

Now, could you conceive of a world where perhaps a big Pentagon-style bureaucracy isn't the most effective, cost-efficient, absolutely optimal way of conducting research and development? Despite their impressive track record, are all the conservatives 100 percent correct in thinking that the old tried-and-true ways of doing things are best, where insight naturally arises according to rank, position, and seniority? That by staffing a large enough bureaucracy of idea reviewers, all the good ideas will sail through and all the bad ideas will be rejected?

The alternative, of course, is chaos, where disruptive hippies with strange ideas about innovation just run around trying all sorts crazy things. We all know that can't possibly work because hippies don't wear ties and appropriate footwear, and certainly don't realize how far down they are in the scientific pecking order. They'd throw all the well-established rules and procedures about research out the window, invent some revolutionary and wildly popular technologies, get rich, and then throw lots of money at the brightest scientists they'd ever met, causing further disruptions to the slow, steady, planned pace of scientific developments, perhaps causing things to be discovered in the wrong order, and perhaps even discovering things without a massive input of taxpayer dollars and the corresponding employment at research labs named after esteemed US senators from good families. That would suck.

The old ways are always best, because if they weren't best why would we use them?

 

[Jedi_Master]

It would just fly right over your head.

Try me I await your wisdom.

 

[gturner]

Well, first off you've asked

Here is my challenge to you.Can you provide a solution to the problem you started a thread to discuss?

If so can you make it one paragraph in length, maximum 200 words?

If a CEO asked me that (and I get cussed out a lot by them), I'd have to reply "Well sir, that's retarded."

There isn't one method that is optimal for funding, managing, and conducting all research, despite what you might be inclined to think. That's a common failing with bad management, who tend to seize on a method, codify it, and turn it into "the system." The idea that there's the same 200-word solution to a nearly infinite universe of circumstances (well-understood problems, poorly understood problems, unknown problems, wicked problems), organizations (research institutes, wings, sites, and departments, along with hospitals, production facilities, nuclear weapon labs, rain forest field stations), priorities (need it now or we all die?), staffing levels, and goals is, quite honestly, retarded, and comes from an alternate universe where the Soviet Ministry of All-Innovations isn't just a joke.

Government minded folks tend to think there's one best way to do something, if not everything. In the real world there are countless of different circumstances and a vast range of possible management structures, and the best one for a particular problem is often shifting in real time as the problem evolves, which is often how real-world research problems are addressed. During the Apollo program both government and industry learned a vast new array of management techniques, because the prior ways of organizing large research and engineering projects were completely inadequate to the task. I gave a good book on Apollo System Management secrets to a highly-stressed executive at Dell, which is still in the PC business, while the management at IBM stayed with their system of holding endless meetings (where anyone in a white shirt could veto a solution until they retired by referencing the appropriate codicils from the shelf of rule books) until IBM sold the PC line to the Chinese, who quickly gave up in frustration.

And often when you do rely on centralized decision making and bureaucracy, much of what's generated is just really brilliant ways to outwit it. When I was eighteen my first job was programming in a biomedical research lab (the same lab were Story Musgrave worked), which I got because they needed a Forth programmer for a data analysis package and I had just written an elaborate Forth interpreter in assembly language. But the university had a hiring freeze, so they cleverly ruled me to be a computer peripheral that was attached to a keyboard. Anyway, we kept getting high-end PC's for data collection while the rest of the university suffered under an extended hold on any new computer purchases, because every department was in a stampede to get their secretaries and other staffs equipped with laser printers and Microsoft Word. One day some other department heads were visiting our lab and asked how we were managing to get so many PC's when they were stymied. We said that we just tack on a couple extra thousand for high-end A/D converters and data collection equipment. Over the next six months, until the bureaucracy figured out what was going on, secretaries all over the university ended up with really nice computers with laser printers and high-speed A/D cards that took up two slots.

In research, everyone has an almost endless supply of anecdotes and stories about the strange ways around roadblocks, and the bizarre things that somehow get funded and published, and all the little tricks to getting a grant approved.

So no, there's not going to be a 200-word solution to all the world's research problems. Only Dilbert's pointy-haired boss would even ask such a question.

 

[JarodRussell]

That kinda reminded me of this one.




What are you gonna do this weekend?

That's a great question and an important one. And I WILL do something this weekend. But let me take a step back and answer a broader question. What are we ALL doing this weekend? As a nation? As a world? This weekend I will do something comprehensive and robust, yet fun. We all should.

But, what are you going to do?

What I am going to do involves three things. First, it's going to be relaxing. Second, it's going to be enjoyable. Lastly, I'm going to make sure that it is cost-effective so I don't get into a deficit. Four weeks ago, I said I was going to do something - and I did. This weekend will be no different!

 

[gturner]

Well, this weekend I'm slated to do some experiments on artificial muscles, round up materials to make a new type of spacesuit joint that doesn't require constant applied torque to remain flexed, and then some random engineering things like stripping electronics out of a wire-EDM machine.

I was going to have to continue the underground search for 8,200 gallons of gasoline that spilled into a cave system, but that got put on hold till next weekend at the earliest.

Meanwhile, I'm sure folks at Apple, Google, Siemens, GE, Microsoft, HP, BASF, Boeing, and 3M will have to be shooed out of their labs this weekend so the cleaning crews can empty the trash and sweep the floors. Passion has a high correlation to successful innovations, and making people go through a laborious and often humiliating grant process can kind of kill that feeling, which is one reason why the most innovative companies don't rely on it for research.

There was a recent NSF sponsored Berkley study covering twenty years of research and patents in the California university system that found that federally funded research was much less likely to result in patents across all technical fields, and those patents were half as likely to be cited by other patents (a measure of patent quality and broadness). Additionally, the corporate-sponsored research patents were less likely than federally sponsored patents to be tied up in exclusive licensing agreements that limit their widespread use. Admitting that must've been like root canal.

link

 

[Jedi_Master]

That kinda reminded me of this one.




What are you gonna do this weekend?

That's a great question and an important one. And I WILL do something this weekend. But let me take a step back and answer a broader question. What are we ALL doing this weekend? As a nation? As a world? This weekend I will do something comprehensive and robust, yet fun. We all should.

But, what are you going to do?

What I am going to do involves three things. First, it's going to be relaxing. Second, it's going to be enjoyable. Lastly, I'm going to make sure that it is cost-effective so I don't get into a deficit. Four weeks ago, I said I was going to do something - and I did. This weekend will be no different!

I love this. Definitely gturner's style. Start a thread, make vague comments about the issue, refuse to answer direct questions, insult those who ask, and then ramble on about his own life.

 

[JarodRussell]

I'm still trying to decipher his last post about artificial muscles, gasoline spills and cleaning crews at Apple.

 

[gturner]

I love this. Definitely gturner's style. Start a thread, make vague comments about the issue, refuse to answer direct questions, insult those who ask, and then ramble on about his own life.

Confess. You are Dilbert's pointy-headed boss.

The big advantage the US currently has in research is that we're much more lax about seniority and position than Europe (we'll give a newly minted PhD from Europe a big lab and a staff, whereas he might have to spend 20-years working his way up the ladder to get the same facilities back home), and we have a lot more money to throw around.

But we do make really bright grad students slave away on projects where their professor will take the credit, and the professor might be "doing" the research because he managed to get a grant for it through the big grant pyramid.

Here's how the NIH lays out the suggested grant process:

******

Planning: Applicant should start early, collect preliminary data, and determine internal deadlines.

Writing: Applicant often begins writing application several months prior to application due date.

Submitting: Applicant organization submits most applications to NIH through Federal portal, Grants.gov.

Months 1 to 3:

Applications compliant with NIH policies are assigned for review by the Division of Receipt and Referral in the Center of Scientific Review (CSR).

CSR assigns application to an NIH Institute/Center (IC) and a Scientific Review Group (SRG).

Scientific Review Officer (SRO) assigns applications to reviewers and readers.

Months 4 to 8:

Initial level of review: SRG members review and evaluate applications for scientific merit.

Impact scores: Available to Principal Investigator on eRA Commons.

Second level of review: Advisory council/board reviews applications.

Summary statement: Available to Principal Investigator on eRA Commons.

Months 9 to 10:

Pre-award process: IC grants management staff conducts final administrative review and negotiates award.*

Notification of award: NIH Institute/Center (IC) director makes funding decision. IC staff issues and sends Notice of Award (NoA) to applicant institution/organization.

Congratulations: Project period official begins!

Post-award management: Administrative and fiscal monitoring, reporting, and compliance.

******

Imagine if we used the same procedure for art, or writing new iPhone aps. But with that much money being tossed around, the public demands strict accountability, because it's public money. As they say, you can go broke saving money, and staffing a bureaucracy with people whose inclination lies in running a bureaucracy (instead of doing science themselves), and tasking it with maximizing the number of fundamental insights generated per dollar by eliminating waste through multiple layers of review (by people whose inclination lies in running a bureaucracy) is not the best way to foster staggering feats of genius.

Little known fact: The time clock, and working a fixed number of hours, was invented by the US military to satisfy Congress that weapon production was being run in a highly-organized, cost-managed fashion, because Congress kept auditing the arsenals looking for waste - in terms of man-hours of labor per weapon. The arsenal response was to make all the local gunsmiths clock in and out instead of paying them by the piece.

So too we try to make insight a 9 to 5 job that can be managed, except for grad students who should work 9 to 9, and then to 11 so they don't have the free-time to calculate that they're making well below the minimum wage, an epiphany that sometimes leads them to take up a career in the food-service industry.

 

[publiusr]

As they say, you can go broke saving money.

And that's the problem. We don't spend enough money on science publically, so all these folks have to compete.

Frankly, if you want to reduce corporate-minded motives in scientific research, you need more public funding of said research.

bingo

Since it's public money provided through a bureaucracy...

Everything is a bureaucracy--and suffers from the Peter principle. Take a look at Dragon's Den/Shark Tank.

As Fallows said in his book "Free Flight" about the air taxi Very Light Jet debacle, entreprenuers and venture capitalists have a class grudge against each other.
http://en.wikipedia.org/wiki/Very_light_jet

One wants more money and a freer hand, the other will nickel and dime you to death, a 'la Bob Lutz's book "Car Guys vs Bean Counters"
http://www.amazon.com/Car-Guys-vs-B...=sr_1_1?ie=UTF8&s=books&qid=1309808368&sr=1-1

You get dullards and road-blocks in both the public and private worlds.

There will never be a private NIF, Large Hadron, etc. And folks won't invest in green tech unless it is required.

You have to have strong public funding.

Imagine if we used the same procedure for art, or writing new iPhone aps.

But that's the problem. You see, the reason the VLJ/Air Taxi model never really worked is that venture vultures would rather invest in something like an app--because there is no factory floor.

Musk really didn't invent anything. The ARPANET, AT&T's wiring, peoples computers--all this infrastructure was already bought and paid for before paypal came along.

Obama and Elizabeth Warren were right. He (Musk) really didn't build that.

What we have are profits without products--and that is why you have to go the public route, because the venture capitalists want to invest in something with the least up front costs and the most profit. Musk had no up front costs.

You don't make great strides doing that, except to come up with more crap like "Candy Crush" to waste peoples time. And people thought missile defense was a brain drain--quite the opposite, with Clementine.

So help me, I don't think Newton could have finished his principia had he had a mobile cell phone and e-mail to distract him. He'd be so busy in flame wars and end times stuff that he wouldn't have time to do anything. Wiles at least had the good sense to shut the world out so he could solve for Fermat.

 

[gturner]

But that strong public funding has to compete with Social Security, Medicaid, and Medicare, which together are 45% of spending, DoD at 18%, and interest payments at 7%, and then fund HHS, State, Veterans Affairs, Homeland Security, Intelligence, HUD, Transportation, Justice, Treasury, Labor, Interior, and Commerce, which together make up 12.6% of spending, at which point the spending is already tens of billions of dollars higher than revenues.

We never sit around worrying about whether Google will have their budget slashed, because we don't have the power to slash it and because Google handles their own problems. But once you invest your everything in one giant budgetary process, hundreds of thousands of people start looking to raid your kitty for their own projects, and they'll keep at it seven days a week, 365 days a year, because that's what they get paid to do. The people in Congress, meanwhile, have to try and balance priorities between wildly different interests and activities, such as either upping SS payments by 1.1% or zeroing out NASA in its entirety (which represent the same amount of revenue).

 

[publiusr]

Well, Musk had that divorce, and I hear Drexler was all but wiped out by one. A private outfit is more vulnerable to that sort of thing. The USAF wouldn't have been hurt if Lemay had fooled around.

I want that same power structure for NASA. It will take some doing though.

Remember, where VLJ failed, F-35 looks unstoppable.