Progress Is Not Enough

Scott Aaronson on The Myth of the Ivory Tower:

I agree with Robin that academic science is often tradition-bound to the point of absurdity, and that its institutions ought to be as open to scrutiny and replacement as its theories. But I don’t go as far as he apparently does in the direction of the Myth of the Ivory Tower. For me, the interesting thing about science is not that it’s broken, but rather that it’s about the least broken enterprise in the whole sorry history of our species.

Arnold Kling on Heterodox Economics:

In my view, the reason that mainstream economics is so difficult to dislodge is the sheer inertia built into the system. In this post, I pointed out how the highly unequal distribution of quality of graduate students results in tight in-breeding. The result is a profession that is very slow to change.  Tyler Cowen’s view is that the best ideas win out, regardless. Maybe that’s true in the very long run. But I think that the process is too sluggish.

Academia may well be the best institution we have so far to produce innovation, and truth probably wins out in the end, at least for most academic topics.  But the mere fact of progress says little; neither of these facts tells us much about how inefficient academia is in an absolute sense, relative to some optimal way to organize innovation.  My guess is that future institutions for innovation will produce many orders of magnitude more rapid progress, given a comparable input of resources.  There will be much new to see on the road ahead; and my bet is that betting markets will be an important part of that new mix. 

GD Star Rating
loading...
Tagged as:
Trackback URL:
  • Greg Marsh

    “Academia may well be the best institution we have so far to produce innovation”

    If you replaced the ‘innovation’ with ‘new ideas’, this strikes me less controvertial. However where innovation is generally understood also to involve the creation of new methods and devices, I think this is less obviously true. If anything it is modern capital markets that are responsible for most product and service innovation. Granted the academy helps train the minds of employees of R&D labs and venture-financed start-ups. But most of the work to convert insight into stuff happens within commercial product development teams. And it is the market that enforces selection pressure to constrain, refine and evolve those products to make them useful.

    So while Stanford can certainly claim to have helped nurture Google’s PageRank algorithm (success has many fathers), it was not the institution of peer review that sparked or honed that invention, but the white heat of industrial design and product engineering, fuelled by speculative investment from Sequoia Capital and Kleiner Perkins, that turned Page and Brin’s insight into an innovation that changed the world.

  • Pete

    Any thoughts on how betting markets might align/misalign with Caplan’s irrational voter?

    Pete

  • http://manufacturedgods.blogspot.com/ LP

    Greg wrote: ” If anything it is modern capital markets that are responsible for most product and service innovation… most of hte work to convert insight into stuff happens within commercial product development teams.”

    It seems like alot depends on the nature of the industry here. In IT, for example, large companies tend to do their own R&D and are rarely interested in technologies developed at universities. In biotech, though, collaboration and co-development with universities is a crucial part of any R&D strategy. Why? Well, IT companies can employ a virtually unlimited supply of hackers, and can develop new stuff quickly without huge expenditure of resources, while pharma companies need access to university resources (tissue banks, patients) and specialized expertise (for a research doc to become expert in a particular disease takes decades). In essence, pharma companies specialize in development of early stage compounds and targets discovered by the disease experts (medical researchers).

    Interestingly, universities and state governments have figured out that they need more of the ‘white heat’ of product development, and are starting to provide proof of concept funds to improve the commercial relevance of university inventions.

  • Doug S.

    The best thing about living in an ivory tower is that it gives you a great view.

  • eric

    Keynes’s said good policies are usually adopted for the wrong reason (or something to that effect). And I think that’s true, in that if you ever tried to pass legislation only with those who agree with you because of the same underlying theory, you would lose (and thus the dirty hands of any politician, because he must collaborate with people who have quite different intentions).

    So if we think of policies being proposed in a quasi-random process, and then growth flowing from good ones, perhaps foresight is over-rated, and ‘successful’ ideas are merely those with the greatest ‘present value’ (as William James would argue for pragmatism) as measured ex post. A good process would allow for more sampling, openness to new ideas, a robust system to replace or abandon inferior ideas.

    Think of ideas like job hires. You can say, we want job hires to be chosen based on merit. In practice they are for all the wrong reasons: he knows my brother in law, he loved my 1998 paper on increasing returns to scale (must be smart!), she’s hot, etc. Hopefully, if you have an exit plan (dismiss if performance is poor), and a broad pool of applicants (you don’t exclude, say, all black people), the broad pool and exit policy will create a ‘pretty good’ set of employees, in spite of a flawed initial selection criterion.

  • TGGP

    Pete, Caplan asserts that irrationality is a private good that people consume, but that through voting its costs are shifted to others (including non-consumers). People are not bearing the full cost of their irrationality. With prediction markets, they are. Also, prediction markets have empirically been shown to be better predictors of elections than polls and a variety of other phenomenon compared to people whose jobs are to predict said phenomena. You might want to ask Caplan himself. He blogs at econlog.econlib.org

  • http://lawlegislationandlunacy.blogspot.com/ David Youngberg

    I’ll have to agree with Greg. Academia (and the military) might do a lot of basic research but it seems that the private sector add much more development much faster. The institution of tenure (which the private sector’s closest relative is the patent) is one big reason to think that competing firms are more inventive than the ivory tower.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    All, by “may well” I mean that academia is near the best, but is not obviously the very best. Evaluating that relative ranking was not the issue I was focusing on in the post.

  • IWantToPersist

    I think one thing that would help is more robust reputational competition between universities.
    For example, companies can lose (or win) reputations very quickly. I don’t see a rational reason that reputation competition can’t be as robust between universities. Yet institutions like Harvard, Oxford, MIT, and Cambridge don’t seem to be experiencing the same types of reputational challenges as are Microsoft or Yahoo. Where are the ambitious start-up research universities, looking to displace Harvard within 20 years? I guess fast moving, ambitious people start companies instead of attempting to reform basic research and biomedical research scientific institutions.

  • http://www.amara.com Amara

    Hi Robin,

    Now that I’ve finished writing grant proposals for a while, I can help someone else with hers. Appropriate to this topic, is someone wishing to establish an institute in Europe for interdisplinary studies between the physical sciences and social sciences, for studies of how science works, and how to improve it.

    http://lightconeinstitute.de/index.php?option=com_content&task=view&id=26&Itemid=41

    You and I know that Sabine is not as market oriented as you, but I still see overlap in yours and her long-term interests for finding ways for scientific resources to be applied in better ways.

    The following is a note I’m passing around.

    Ciao,
    Amara

    —-
    A colleague of mine (me = Amara Graps) is working on a grant proposal to the European Research Council to establish an institute for interdisciplinary research between social and natural sciences, with special emphasis on the sociology of sciences (the way science works). She is seeking active physics researchers, as well as people working on mathematical sociology or the sociology of sciences, who would would be willing to write a support letter. At this point, no commitment is required. The focus is on collecting support throughout the European Union and from all fields participating in these research areas.

    If this sounds interesting to you, and you would like to obtain further information, please contact Sabine Hossenfelder at sabine@perimeterinstitute.ca . Feel free to forward this note to anybody who might fit the description.

    —-

  • http://bbnflstats.blogspot.com Brian

    Why does orthodox economics need dislodging?

  • http://gondwanaland.com/mlog/ Mike Linksvayer

    “My guess is that future institutions for innovation will produce many orders of magnitude more rapid progress, given a comparable input of resources.”

    This is incredibly optimistic about future institutions, incredibly cynical about current academia, missing “over centuries” after “rapid progress”, or all three.

  • http://www.scottaaronson.com Scott Aaronson

    Where are the ambitious start-up research universities, looking to displace Harvard within 20 years?

    This seems to me to be very simply explained by a difference in dynamics: in the corporate world, there’s a strong pressure for people with great new ideas to start companies (the goal being to bring the ideas to market), whereas in academia, there’s an equally strong pressure for those people to join the best existing universities (the goal being to be around other people with great ideas). Department and university rankings can and do change, but it takes a long time: some top faculty have to be the first ones to “defect,” thereby forming a nucleus that will attract other top faculty. So it’s not that other universities don’t want to displace Harvard, Stanford, Princeton, etc. — they’d love to! — it’s just that on a 20-year timescale, they can’t.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Brain, that was not my claim; follow the link to see their reasons.

    Amara, is the claim that more funding should go to 1) sociology of science in general, 2) physicists trying to do sociology of science, or 3) her particular (unclear to me) approach to physicists trying sociology of science? I might support more for social science of academia, including economics of academia, but I’m not sure why funding should focus more particularly than that on her approach. If there is a canonical place to discuss this, do point us to it.

  • Douglas Knight

    Scott Aaronson,
    It seems to me that there are two different sources of “a strong pressure for people with great new ideas to start companies,” both of which are probably true, but one indicates an institutional failure, while the other says that people with different goals have different results.

    There is pressure to start new companies because it’s a high-risk high-reward gamble. New ideas in the academy may also be gambles, but successful academics are remain cheap and can be bought out by old universities.

    But there is also pressure to start new companies because old companies don’t pursue them. Companies should be less risk-averse than individuals, but the existence of start-ups shows that they aren’t. I think it’s best explained as a principal-agent conflict, but it could also be explained by founders being at the tail end of irrationally optimistic people.

  • Bruce G Charlton

    The US academic system the most dynamic generator of scientific innovation, and this dynamism shows in terms of up- and down- movement in the rankings, and the pulling-away of the US from all other nations

    http://modernizationimperative.blogspot.com/2006/12/nobel-prizes-as-scientometric-measure.html

    But we have to recognize that the pace of institutional change is measured in decades not years; due to the constraints of planning, funding, executing, writing, disseminating and responding-to research projects (so that a cycle of communications takes many years to complete) – and also the relatively slow turnover in expert manpower.

    There is no such thing as optimal efficiency – all efficiencies are relative; but it would be amazing if the US academic system didn’t improve further – so long as competition, selection and evolution continue.

    It is the rest of the world’s universities that should be the focus of our concern…

  • http://www.daublin.org Daublin

    Please separate science from academia. A great thing about science is that not only is it good, but its processes are improving. The positivist view was the norm for a long time, but is fading in prominence. Falsifiability was only proposed in the 30’s. I am sure there are plenty more examples.

    In academia, on the other hand…. Well, when was the last time a popular university went bust? Universities do not improve due to selection. So how else do they improve? Academic acclaim is like watching the Oscars: the people making the judgments are the people involved in the system. Where is the external validation?

    I am not saying that we should burn the schools or anything, but we should be wary of assigning too much wisdom to these institutions. They are strong on tradition, and it’s a good tradition, but they are weak on improving how they work.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    Daublin, I’m talking about research, not teaching. And it is very hard to create a meaningful distinction between science and non-science academia.

  • http://backreaction.blogspot.com/ Bee

    Hi Robin:

    Amara sent me a link to your comment above. There is not yet a place to discuss. I will probably have a post introducing my proposal over at my blog and will send you a note then, feedback is welcome. To briefly address your questions:

    is the claim that more funding should go to 1) sociology of science in general, 2) physicists trying to do sociology of science, or 3) her particular (unclear to me) approach to physicists trying sociology of science?

    1) Yes, we are claiming that more funding go to the sociology of science in general because the field is becoming increasingly important, which imo has not fully been realized yet. However, we can’t very well write a proposal just wishing for ‘more funding’. Therefore we have worked out a very precise context that we hope will find support and which embeds this research.

    2) No, this must be a misunderstanding. The institute does interdisciplinary research between physics and sociology/politics, but a) this does of course include people working in all of these areas, not only physicists (this would be a disaster) and b) though the interdisciplinary part is of course in various regards related to the interdisciplinary one we did not plan to have physicists in this area.

    3) The institute IS my particular approach. I have a personal opinion about what I think will work, so do you I presume, and I also have an opinion about your opinion (that I believe I made quite clear). However, the institute is not about what I think will work, but about providing an environment to explore/work out proposals that can improve the situation. (Your proposal might have fallen into this area. I say ‘might’ because I don’t wish to be the one and only person who makes a decision about which projects would be the most promising to work out).

    Best regards,

    Sabine

  • William Newman

    Many orders of magnitude? Even one order of magnitude faster would’ve gotten us from Galileo to quantum electrodynamics in one ordinary person’s professional lifetime. That already seems awfully fast for institutions whose inputs are at all comparable (anything based on plausible numbers of human beings, as opposed to super-fast AI or something). Two orders of magnitude would do it in about the time it takes most people to get a university degree. With three orders of magnitude you couldn’t take a sabbatical without having a couple of major scientific revolutions while you were gone. And “one, two, many” is only a joke: usually “many” means more than three…

  • http://profile.typekey.com/sentience/ Eliezer Yudkowsky

    I have to agree with Newman. I can visualize a system that uses sensory inputs and physical resources (and raw computing cycles) roughly comparable to modern academia, which produces “orders of magnitude faster” progress in the sense of literally requiring less than a month to go from Galileo to electrodynamics. But it wouldn’t be remotely human.

    Presumably “orders of magnitude” was meant in a rhetorical sense. I recently made a conjugate mistake during an interview when I referred to the billions of dollars spent yearly on marketing lipstick, and called the Singularity / intelligence explosion “a billion times more important”. If you try to calculate the expected value of an intergalactic civilization, it seems likely that I dropped at least fifty zeroes for the sake of compact rhetoric.

  • http://profile.typekey.com/robinhanson/ Robin Hanson

    William and Eliezer, I did mean to say at least a factor of one hundred, but I didn’t mean to give the impression it would happen soon.

    Bee, I don’t know enough yet to endorse your particular approach, but as I said I can endorse more social science of academia.

  • Buzzcut

    Academia may well be the best institution we have so far to produce innovation

    That’s laughable. In terms of innovation, we can’t hold a candle to the period from the end of the civil war to WW1. Academia was no where to be found in terms of innovation at that time.

    Even today, it is where academia and the market intersect that innovation occurs. A place like Silicon Valley is innovative because of its university connections, but also because of its corporate and venture capital connections. I wouldn’t give academia all, or even most, of the credit.

  • http://amnap.blogspot.com/ Matthew C

    That’s laughable. In terms of innovation, we can’t hold a candle to the period from the end of the civil war to WW1. Academia was no where to be found in terms of innovation at that time.

    However academia has lots of cushy high-paid and high-status jobs to offer, fancy stone buildings to work out of, and a non-stop conveyor belt of self-congratulatory puffery to offer about how smart academics are and how foolish everyone else is. What’s not to like for the beneficiaries of such a system?