= 27d567bc2d48d9f40e9ccc20ea370b15)
Salemicus comments on How minimal is our intelligence? - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
= 783df68a0f980790206b9ea87794c5b6)
Loading…= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
Powered by Reddit
= f037147d6e6c911a85753b9abdedda8d)
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (214)
Not at all. The point is that some academics are useful and some are not; there is no market process that forces them to be so. It may be that some of the academics are able to continue doing exactly what they were doing, just for a private employer. But I would not bet on that outcome for most.
???
Academics need funding. The ability to get funding is well correlated with usefulness in most fields (to be fair, other things are in play, like institutional inertia, fashion, etc. etc.) Useful research also results in spin off companies, and fame. There are all sorts of incentives to be useful in academia.
There is also the issue of hedging, and diversification of intellectual effort -- you want people doing long term payoff and long shot research as well. Modern business culture is generally much worse at this than academic culture is at being useful. Google, a company started by two ex graduate students, is one of the few notable exceptions.
How would you tell?
I could look at output over the last 50 years, comparing useful stuff out of academia vs long term research done by corporate research labs, scaled by funding. Or I could look at incentives high level corporate decision makers have, which heavily favor the short term and empire building. Or I could look at anecdotal evidence based on testimonies of my corporate vs academic acquaintances. Or I could look at universities today that produce useful applied research (basically any major research university) vs companies today that have labs working on fundamental research (Google, Microsoft, possibly some oil companies (?), maybe Honda, maybe some big pharmas and that's about it).
The vast majority of big companies (the only ones who could afford fundamental research) do not engage in fundamental research. The vast majority of research universities do very useful applied work.
This might equally lead to the conclusion that the kind of "fundamental research" you're talking about just isn't very worthwhile.
Ok. Which of the following do you think is a worthwhile research question:
Non-Euclidean geometry (Lobachevsky, 1826).
Galois theory (Galois, 1830).
Complex numbers (Bombelli, 1572).
The periodic table (Mendeleev, 1869).
Interventionist causality (Wright, Neyman, Rubin, Pearl, Robins, etc. 1920-today).
Objectivism (Rand, ~1950s).
None of these were developed by corporations (or similar entities) or corporation sponsored individuals, to my knowledge.
Bombelli predates corporations on any large scale, so including his work seems strange in this context.
Private individuals could not incorporate until fairly late (after Bombelli) but states started establishing commercial entities that play the role modern corporations play in our society fairly early. I edited a little to clarify, though, thanks.
http://en.wikipedia.org/wiki/List_of_oldest_companies
http://en.wikipedia.org/wiki/History_of_corporations
I don't know anything about interventionist causality, and Objectivism seems like a waste of time. The rest all seem to have produced worthwhile results.
But this is costless analysis! Of course if you buy lots of lottery tickets, and look only at the winners, then buying lottery tickets looks worthwhile. You have to consider the opportunity costs, not just of these research projects, but also of every other similarly situated research project. And you also have to do time-discounting. Bombelli discovering complex numbers in 1572 looks like a waste, when they weren't useful for 200 years or more.
Moreover, I don't know why "corporations" is the comparison. The comparison is the private sector generally. Huge amounts of scientific work are done, and continue to be done, by enthusiastic amateurs, and the charitable sector. I am not making the claim that in an ideal world all government academics should be fired (although I do think that would be a big improvement on the current situation). I merely claimed that, on the margin, we are hugely oversupplied with academics, and undersupplied with businessmen.
Ok. So, helpful fundamental research, as it is currently produced: (a) imposes a heavy opportunity cost, (b) has a low success rate, (c) is generally discovered "too early," leading to waste. What is your proposal for doing better? Can you give me some examples of things like complex numbers discovered in the private sector, by charities, or 'enthusiastic amateurs'?
Incidentally, the success rate of fundamental research for a given finite time horizon k is an untestable quantity. Thus, (b) is a weak complaint. (a) is hard to argue also, because you need to construct counterfactual scenarios that people will believe.
Examples that fit Salemicus's narrative in this context aren't non-existent. For example, Fermat was a lawyer by profession and did math as a hobby in his free time. There are many similar examples prior to the 19th century or so. And some major charities have helped fund successful research- one sees a lot of this with a variety of diseases.
How exactly do you measure "success" in this case ? As for me, I find myself hard-pressed to think of any examples of fundamental research that weren't ultimately beneficial -- except perhaps for instances of outright fraud or gross incompetence.
Even if a scientist spent five years and a million dollars trying to discover, say, the link between gene X and phenotype Y, and found no such link, then the work was still not in vain. Firstly, we can now be more certain that gene X does not cause Y; secondly, we can most likely gain a lot of collateral benefits from the work, leading to an increased rate of discovery in the future.
Sure. For a start, complex numbers themselves - Bombelli was in the private sector himself. Some others, taken at random:
For-profit: Lightbulb (Edison, 1881), Propranolol and Cimetedine (Black, 1960s), Fractals (Mandelbrot, 1975)
Amateurs: Evolutionary Theory (Darwin, 1830s-50s), Photoelectric effect, Brownian Motion, Special Relativity, Matter-Energy (Einstein, 1905), Linear B (Ventris, 1951),
Actually, if the success rate of this spending is as unknowable and untestable as you say it is, that's an excellent argument to stop forcing unwilling people to pay for it. But note that although I'm happy to fight you on your strongest ground (pure scientific research), surely you must then concede that the rest of the battlefield is mine, and that all government spending on academia that can't be justified in these terms (e.g. arts, humanities, medicine, space exploration, etc) should be eliminated. I'm not doctrinaire - I'd settle for that compromise.
EDIT: I'd truly be fascinated to know why was this voted down.
I don't know precisely why your comment was voted down, but one obvious guess is factual issues:
Darwin got most of his ideas from his time working the survey/exploration ship the HMS Beagle. As you may gather from the "HMS" in front, this was a ship in the British navy which had specific funding to hire and provide support to a naturalist.
This is why I suspect you are getting downvoted. First, it shows a confused notion of what is "pure scientific research"- a large part of space exploration and medical research counts as pure research for purposes of the arguments being made by IlyaShpitser and others in this context. Moreover, you seem to be wanting a "compromise" as if the truth must be somehow negotiable. The key of discussion is to understand what is likely to be actually true. Settling for a compromise isn't how one reaches truth (and no, Aumann's agreement theorem doesn't apply here).
No. Just... no. What differentiates basic research and applied research is how many years there are until commercial application. For applied research, the number of low- five years is stretching it, and hopefully it'll be less than one. For basic research, the is number is larger- it was around three decades from Einstein's explanation of the photoelectric effect to the commercialization of cameras based on it.
The point that some here considering burning of books a taboo, and that that disagrees with consequentialism, is an interesting and valid point. The point that public goods can be provided without government intervention is an interesting and valid point as well, but you're not arguing it well.
Another problem with fundamental research -- from a commercial corporation's point of view -- is that its results may not be applicable at all to products in your target market. For example, you might start by researching the formation of clouds in the atmosphere, and end up with major breakthroughs in atomic theory. Those are interesting, to be sure, but how are you going to sell that ?
This is but one of the reasons why large corporations tend to stay away from fundamental research, unless they can write it off their taxes or something.
Can you briefly taboo 'useful' for me? What do you mean by that?
In this context, I mean providing a service that someone else is willing to pay for.
By that definition, heroin is useful.
So by that notion, anything that is an externality but can't be captured by market forces is by definition not useful? Does that capture your intuition for the word useful?
I said is willing to pay for - not necessarily that they can pay for it. Any one-sentence definition of a word as complex as useful is going to necessarily be incomplete, but I certainly mean to include externalities in it. If, for example, people value "a sense of belonging to a community" and are willing to give up something meaningful for it, but co-ordination problems or whatever else means it can't be captured by market forces, then I would absolutely view someone who creates "a sense of belonging to a community" as useful - provided that the cost of their doing so is less than the price that the community members would be hypothetically willing to pay.
Fair enough. How do you determine then what people are counterfactually willing to pay?
I don't think anyone has a good way of doing that.
Would you grant that many things are valuable which are nevertheless not useful in that sense?
EDIT: I don't mean anything fancy here. Eating a hot dog, for example, is valuable (I'm willing to pay to do it) but not in any sense useful (no one is willing to pay me to do it).
Yes, sure. But I was talking about useful as a quality of a person, not as a quality of an object.
However, as my comments in this thread are getting voted down, I assume it's not really worthwhile to continue this conversation.
I see. And does the fact that much academic research produces positive externalities/public goods, which thus aren't easily funded by private employers says what in this context?
Deadweight cost of taxation + deadweight costs of political rent-seeking + resource misallocation due to political decision-making + opportunity costs
versus
costs from possibility that private sector is unable to capture all externalities
Markets fail. Use markets.
I'm confused by this remark, given the context is about academic jobs, not government intervention in the market. Can you expand/clarify what you mean?
Government creation of academic jobs is intervention in the market for academic jobs.
Yes, but that's a tiny fraction of the issues you list above. Unless I'm misreading you. By for example deadweight cost of taxation, you mean the deadweight loss of the portion of taxes that go to fund academic research?
Taking that sort of interpretation throughout, I'm still not sure what your point is. Can you be more explicit and maybe use full sentences?
You stated that academics aren't easily funded by the private sector because of an externality argument. I agreed that it is possible to argue that "basic research" or some such is underprovided by the market, because the private sector may not be able to capture all externalities, and that this is in some sense a market failure. However, I am saying that:
So, even though there may be a textbook "market failure," there is no reason for any intervention. Dissolve the modern-day monasteries, and let academics prove their use, if they can. And indeed, I'm sure Alvin Roth would be just fine if we did.
No. We can make such estimates by looking at how helpful basic research was in the past.
Sure. How much?
That's a danger certainly, but what evidence do you have that that's happening?
The areas where politics has heavy aspects are actually the areas with the least government funding. For example, physics has a lot of government funding, whereas most of the humanities and social sciences have comparatively little. Thus, the politics comes into play primarily through the interaction with outside donors with agendas. That's how you get virulently anti-Israel attitudes in Middle-Eastern studies due to funds from rich Saudis and you get Israel studies as a subject which is about as ridiculously biased in the other direction for the same reason. Political problems in the sciences are rare.
We have theorems and a lot of empirical of how externalities interact with markets. If you think there's something wrong with that vast body of literature, feel free to point it out.
Is this a serious argument?
Yes, grant proposal writing is annoying and often a waste of time. Question: What fraction of taxpayer money is going to academic research?
At this point I think I have to cite Use of Knowledge in Society, Hayek, 1945 link.
Where I live, we spend approx $4bn per year (0.64% of GDP) on state-funded research (note that this figure is conservative because it doesn't include the way that higher education funds get siphoned off into research). Conservatively then, let's say $1bn in deadweight loss annually, just in this country - and our state-funded research is low compared to most OECD countries. If we extrapolate this figure to the world economy, we get a deadweight loss of approx $127bn annually, just due to government research spending. That's a lot of bednets.
I am not talking about partisan clashes. I am talking about money being spent on worthless projects because they seem cool or win votes. NASA has a budget of almost $18bn!
Of course it's a serious argument - subsidizing one group of rent-seekers encourages others. I am of course being a little facetious in the sense that both the academics and the farmers already have their snouts deep in the trough.
Sorry, I'm not following. You are citing Hayek to argue what here?
Ok. So we have less than 1% of GDP going to state-funded research. And where is that going to go?
Projects seeming "cool" is a very different claim than political rent-seeking. In this case though, looking at the overall NASA budget isn't very helpful: First, much of that budget is not going to what would be considered academic research. Second, you are talking about the space program of one of the world's largest economies, so the total cost is a misleading metric. Third, technologies developed by the US space program (especially GPS, communication satellites and weather satellites) have had large-scale world-changing impact.
It often doesn't, and the case you've picked is a really good one. In the US, many of the people getting farm subsidies are people who are rural and if anything anti-ivory tower. A large fraction would probably be turned off of the idea of government subsidies if it was compared to what those East Coast intellectuals were doing. (I'm engaging in some broad brush strokes here obviously but some people like this do exist.) This sort of thing is connected to why many groups (including farmers) have tried to get their money through tax breaks rather than direct subsidies. Of course, from an economic perspective, tax expenditures are identical to subsidies. But people don't like to think of themselves as getting handouts so they prefer tax breaks (at least in the US).
By the way, my point earlier about only a small fraction of tax money going to academic research was (to be clear) about the claim that academic research would necessitate tax policy watchdog groups.