The decline of unfettered research (1995)
dtc.umn.edu
dtc.umn.edu
At IBM. That wasn't a smart move either.
The term "under the radar R&D" has not been unheard of in many corporate research labs.
But what worries me there are reports that the corporate R&D lab as an institution is in decline. I cannot judge whether this is true, since I recently switched back to academia to have a bit more autonomy after a decade in industry R&D.
I've spoken with colleagues over the years, and when I bring up this desire, so many of them feel the same way. Many of these people are incredibly accomplished and come from top institutions. What a waste to not give that magnitude of creativity an outlet! The only path I see to this type of life now, is independent wealth. It just strikes me as such an opportunity to create a place for these people. If you build it they will come!
the problem is finding and filtering the applicants that come to such a place. How do you tell the difference between somebody who just phones it in, from somebody who would actually produce some valid research breakthroughs?
Unless you have unlimited resources to spend, this is a difficult problem to solve.
Managing a process which has no defined outcome, but whose local quality and direction can be measured seems like an interesting research path in itself. I think the institute would have to be bootstrapped and a lot of learning would have to take place to make a sustainable version of it. Also, as you say hiring would be challenging (though maybe no more challenging than hiring professors). However, I'll say that I can recognize at least one phenotype of researcher reliably who would thrive there.
"Success leaves clues" affords some possible hope, though there's also a long history suggesting progress and individual attribution are limited and difficult to assess.
There's a history of major research institutions, public and private, including those of Edison, GE, Dupont, AT&T Bell Labs, Xerox, IBM, Microsoft, Google and other companies, as well as the Manhattan Project, WWII signals and computing research (see especially Norbert Wiener's account), the space race and Apollo project, the Santa Fe Institute, and more. These should afford guidance.
Often one clear trend is that results follow something of a sigmoid curve --- slow initially, a period of strong progress, followed by a decline. Much engineering trends toward an optimum, not some unlimited potential. Even Moore's Law likely ultimately follows this path.
I read a lot of his papers ( http://www.dtc.umn.edu/~odlyzko/ ) in the hopes that I would learn something to improve my exam scores, but he has a knack for asking questions so fundamental that they have almost never even been properly formulated before.
If you have some time, I recommend reading a few of his papers. He completely changed my view of mathematics.
Indeed, for example this page [1] is absolutely fantastic, it hits a lot of right buttons in my case (history of railways, history of finance/economics, a combination between history of railways and the history of finance, which is even more interesting).
https://physicstoday.scitation.org/doi/pdf/10.1063/PT.3.4521
There could be a whole range of topics that could be worked upon and you could allow the researcher to move freely between them,
https://volokh.com/2011/10/02/justice-breyer-on-tenure-stand...
“The highest form which civilization can reach is a seamless web of deserved trust. Not much procedure, just totally reliable people correctly trusting one another. That’s the way an operating room works at the Mayo Clinic.”
I'd guess that the polarization of our society is self reinforcing, and that as it increases, our institutions follow along for one reason or another whether it be some form of "good business" or internal capture. Its probably some Internet Law that any institution not specifically devoted to staying out of it will eventually succumb, but even then you have the ACLU... and there is non-culture war decline in trust - something more like the financialization of everything or the injection of metrics into all parts of life that can be metered....
We don't value competence and accountability first; maybe this was always the case and only our narrative changed, but that's hard to imagine when we used to do things. At this point, why not Research, too?
In research, abolishing the grant system and the business model of universities selling credentials in favour of direct financing of learning and research institutions to teach the best students should be attempted. It probably won't come from the academia, it has to come from the state.
The thing about basic research is that you usually cannot see the consequences down the line.
Few of the original researches that discovered ionizing radiation could anticipate the enormous destructive power of nuclear weapons. And yet their contributions were crucial.
May require extensive trial and error before a good result is uncovered. Don’t let anyone tell you it’s cherry-picking..
https://www.hhmi.org/programs/biomedical-research/investigat...
They also have their problems, but it's good to have some of them in the mix.
If you no longer believe what you proposed is a useful thing to do... send an e-mail. This may be easier to resolve than you think.
Seriously, a lot more phd students just need to embrace their ADHD. Even if you are given a path, there is no path.
Fun times.
(What was the research? Basically taking Heckman's 70s-era selection correction and applying it to nonlinear models. Big fuckin' whoop. My advisor had already written multiple useless papers riffing on the idea anyway. He's probably still doing it to this day.)
On the other side I have multiple projects, the biggest one already presented in a conference. I meet multiple people my university and other big ones that are very enthusiastic about either that project or the official one (which got none in my own lab) so it's not like what I'm doing was totally dumb.
I'm not sure what to do now. I have 3 more years to write the thesis, but would need to find a job to stay in the country. Hopefully I used web technologies I can market on my CV but I'm bit sour about the whole thing (and life in general).
This hits so close to home that I think I just heard a knock on my front door.
I was diagnosed with ADHD about a month before I started a PhD (in my mid 30s). In my 3rd year, I finally took the required course on theory of computation and almost failed it because I spent all my time reading SEP entries about the foundations of mathematics and the origins of computing. I even started reading the Homotopy Type Theory book, despite a woefully inadequate formal mathematics background.
Meanwhile, I’ve got two papers in submission on a topic that I think is a dead end and a waste of time, and I’m hoping that my advisor won’t be too upset if I just refuse to work on any of the follow-up papers that he is considering. I prefer to hang out with math PhDs rather than my lab coworkers, so I can pester them with questions about category theory. And I’m supposed to scrape together a dissertation in, like, 18 months.
eh... I know of one similar instance (student had found that what they were looking for has been proven false by another team) - the advisor basically said "okay, we're stopping the phd there". Two years to the drain.
I wouldn't be surprised if some of them choose to create university positions for themselves, or otherwise "self-fund" their own prestige projects.
One would need a very complete picture to be able to accurately generalize in such an absurdly broad area.
I don't buy that because it's very expensive to experimentally prove the difference between various theories.
Best regards, A purebred experimentalist from the theoretical institute of physics at Blegdamsvej 17 (Niels Bohr’s Institute).
Right, and the mathematical field where you use axioms matching physics experiments is called theoretical physics. Currently there is a lot of work to do in that domain, as currently we don't have any theories that matches well with all experimental data we have. We have small scale theories and big scale theories, but there still needs people to work out theories and formulas for combining the two. You might call that mathematics, other calls it theoretical physics, either way they will help further the field of physics and science in general.
A professional is someone you trust to do their job. Like a doctor or a lawyer. If you have a medical or a legal problem, you go to a professional and the professional uses their professional judgment to make a decision and works on your behalf to try and solve that problem. And they take personal responsibility for their professional decisions. For instance, when a professional engineer signs off on building plans, he is saying, "this building is not going to collapse and kill people, and if it does, I will take personal responsibility".
A bureaucratic environment is an environment where processes and controls have supreme authority and there are no professionals. You have to jump through hoop A, fill out form B, and have everything reviewed by committee C to do anything because you are not a professional and your judgment isn't trusted. To some degree, this means doctors aren't fully professional anymore.
As you observed, by this definition surgeons are not professionals but the teenager who's the sole employee at a lemonade stand is a professional. The former is extremely constrained by bureaucracy while the latter can do pretty much whatever they want as long as nothing burns down or gets too many people too sick.
Bureaucracy and professionalism are largely orthogonal. There are horrible bureaucracies where some individuals have immense power. In fact, that's probably way more common than not. There are also relative anarchies where no one has any real power because the whole org is completely beholden to the market in every aspect of its operation; e.g., most corner pubs on a crowded business street. Even the owners have at best marginal control over their employees and rented space.
A professional is just someone who does the same sort of skilled work year over year for pay. Most blue collar workers think of themselves as professionals, and you'll see plenty of discussion of "professionalism" in any trades training program.
Historically, the connotative notion of a "professional" that you're using here -- basically, upper-middle class professions with a certain amount of social esteem -- were always the most bureaucratic occupations. Have they gotten even more bureaucratic with time? Sure. But they were always more bureaucratic than other occupations of their time (mostly farming). Medicine or law being more bureaucratic than than farming is not a new thing.
I wonder what such people currently do with their time? Actually, HN is a great place to ask this, as lots of them are here! wait
Right after WWII with the planet in shambles, living in the "winning" country, you're working on computers and found unfettered access to funds and investment? No surprise.
GE in 1956? To leave out the massive macroeconomic power of GE in that day and age is shortsighted. Same with Bell Labs, et all. This was an age where military spending rose from 1% of GDP to 10%. It was military spending and military might that bought you that "unfettered research". Yes, society should be better at allocating for the long term regarding research and tech -- but 1956 GE was not some sort of utopia.
We're just in a lower part of the cycle right now. Unfortunately, the only reliable "reset" button society has found seems to be war. Hopefully modern financial markets will be able to create those cycles without as much bloodshed.
Expensive drone industry: US government buying unmanned air vehicles, totally unrelated to quadcopters.
> Unfortunately, the only reliable "reset" button society has found seems to be war.
In my opinion this claim needs significantly more justification even though it is frequently tossed around.
>In this style of work, the researcher is allowed, and even required, to select problems for investigation, without having to justify their relevance for the institution, and without negotiating a set of objectives with management. The value of the research is determined by other scientists, again without looking for its immediate effect on the bottom line of the employer. The assumption that justifies such a policy is that "scientific progress on a broad front results from the free play of free intellects, working on subjects of their own choice, in the manner dictated by their curiosity."
It's sad the current state of science doesn't appreciate the work done purely through curiosity, and instead want to milk professionals for other means and agendas. The paragraph sheds a light on what science really is, and what's kept fueling it for millenia, curiosity. Some of the greatest scientific discoveries have come from curiosity in answering burning questions. Yes we still have some great discoveries, but not as much now I would think. Most of what science today seems to be is just proving or disproving agendas with clear incentives. There are some that seem to be born out of organically produced work, but it's hard to know because who knows the incentives and agendas behind the scenes.
Practically speaking this means that my work is restricted in that it deliver something resembling a product or service within a couple of years. Even if only the thinnest of MVPs of a concept. String enough of those together and you get pretty close to something resembling “unfettered”.
I suppose my point is that industry R&D is not all that bad right now, you just need to be a reasonably responsible corporate citizen.
So fettered, as per the articles definition of the word fettered.
It really just sounds like you don't like the word fettered for some reason. Maybe it has a negative connotation to you?
The Decline of Unfettered Research (1995) - https://news.ycombinator.com/item?id=2952423 - Sept 2011 (2 comments)
Now that a long time has passed, I would be interested in hearing an analysis of this exactly as it is phrased: financially, with respect to profits vs valuation, and not in terms of historical impact.
[0]: https://web.archive.org/web/20171107021707/http://news.morni...
[1]: https://www.fool.com/investing/general/2013/08/09/the-ipo-th...
I suppose that at this point it is lost in the digital accounting noise, but I bet there are some people left who saw it happen and have a perspective on it.
At last, the internet, opium of the masses, finally neutralizes the intelligent classes.
Wealth will not care for genuine research until two aliens races war bitterly over some unique natural resource we take for granted. We will not recover.
(I’m always amused that the Lambda-the-Ultimate papers were paid for by the Office of Naval Research.)
This could be fixed with the commonsense solution of requiring public access to publicly financed research and patents without burdensome fees. This would have allowed numerous private entities to begin their own development and manufacturing processes without worrying about patents and IP lawsuits.
Unfortunately, the end result wa the rise in power of the academic Intellectual Property Office which oversees the licensing (and gathers in the percentages). The UC and MIT systems are most notorious for pushing this approach in the 1990s, but it seems to be everywhere these days.
Effectively, the administrators put researchers with patent-generation potential at the front of the food line, and also pushed hard for public-private partnerships with mega-corps (UC Berkeley and BP, Stanford and Exxon, etc.) As part of that mentality, short-term profitability became the guiding light, not basic blue-skies research into whatever the researchers wanted to look at. Since much basic research generates nothing of immediate commercial interest, it was viewed as less important and even a drag on the bottom line.
Running with this mentality equates to killing the goose that laid the golden eggs, as groundbreaking discoveries leading to truly new technologies then become much less likely.
On the business side, the large private research centers of the post WWII era seem much diminished. Giving corporations a tax exemption for increasing their R&D spending while also raising their taxes to 1960 levels might be an efficient way to reverse that trend. Elon Musk could then avoid the tax bill by putting all his money into a SpaceX R&D facility to rival the old Bell Labs, which is kind of a good idea anyway isn't it?
When the drone lands on a platform atop a building, the prong connects to a computer connected to the platform and sitting inside the building triggering a mount action.
The data gets transferred to the computer. Now, whoever needs to transfer data from this building to another will use the same method to upload it.
I do not think this will be the best way to move data but if a protocol is in place it could be used as a basis for future intra-campus data movement.
The proposal was shot down. A schematic of the idea is drawn here: https://github.com/ketancmaheshwari/datadrone/blob/main/sche...
And for other practical applications related to this idea: https://aws.amazon.com/snowmobile/
-- Cost: Drones are getting faster, accurate, reliable and cheap while storage devices are getting lighter and denser. If a protocol is in place, vast amount of data could be moved relatively cheaper at a faster rate.
-- Auxiliary Medium: If a campus network is down due to security threats or assessment or some other reason, this protocol may be used to pass critical data around.
-- Remote, inaccessible (edge) locations: Places where conventional network is difficult to setup due to temporal nature or hazardous conditions etc.
Never underestimate the bandwidth of a station wagon full of hard drives.
But really, for practical applications I would probably try to use UUCP.