Log in

No account? Create an account

Long Legs, Short Patience


  • I'm not applying for even close to every job I'm qualified in PHL
  • Confirmed that I'll be keeping my mini-job in CMU-Psychology for the time being, as I can do it remotely
  • Had another gahhhh experience last night trying to reconstruct several years of attendance from several years of emails. As I expected, I have severely underused my vacation days.
  • Seeing if I can find a new fulltime appointment at CMU that can be done remotely. This is very hard, but CMU seems to have a lot more need for people like me than many of the universities/colleges in the PHL area. No surprise; I think Boston, Stanford, and a few other Universities are similar enough (or a lot of places in Europe).
  • I'm starting to put more effort into getting other remote work. If I can make the numbers add up, I can avoid eating savings and maybe even buy insurance until the next stage of my life.
  • Hoping to have a graceful exit to my job here without working ridiculously long days.

I like Larry Sanger's blog. So far he's mainly been writing about the growth of anti-intellectualism in the US, a trend I also recognise and am worried by. I think some of this is cultural. We're being too sensitive to people who don't stretch their minds, by saying that it's just as good not to. Americans are also excessively individualist, and we love innovation to the extent that we don't consider what we giving up when we replace traditional good ways of doing things. There are a lot of institutions that become obsolete and need to be discarded, but other institutions (like schools and universities) need smaller changes (if anything at all). Raising kids to thirst for knowledge and have attention spans longer than gnats is a better plan than designing ADD-compliant educational curricula. In practical terms, using KhanAcademy to help kids learn in a relatively standard classroom environment with well-designed lesson plans might make things better. Having kids spend the majority of their school time in either homeschools or 'unschools' with self-directed learning will lead to a lot of wasted potential, a lack of discipline, and gaps in knowledge. Unstructured learning can happen around structured learning.

I don't think one must go to college to be a good/educated person, but that it is better if one does. Universities/colleges are about self-betterment, and they (ideally) expose one to perspectives and ideas that one might not willingly be exposed to by one's parents/community or one's own exploration. The structure of higher education also exposes one to the best methods our society has to determine truth. I don't buy that "college just isn't for some people". It should be funded by taxes, and it should enrich everyone, at least at the 2-year level.

Also, IBM's Watson has been adapted successfully towards medical evaluation. I've occasionally talked about this kind of thing with big-picture-people working in the field of medicine. I never anticipated that exposing multiple answers with confidence-levels would be a big breakthrough, although in retrospect it's very important.


  • having doctors keep in their head a staggering (and growing) list of illnesses and partly overlapping lists of indicators. As time goes on, the time needed to reach knowledge enough to do a decent job at illness-determination may require excessively long training for doctors (10 years? 20?)
  • effectively judging which illness is most likely in specific cases, possibly ordering tests to get more information
  • This is very difficult!
Watson isn't exactly an expert system, but it should be very good at this specific type of human judgement. This problem domain reminds me of a trickier one in academia: the ever-increasing depth of each field bottoms out as it becomes impossible to draw on sufficiently broad domain knowledge to make more progress. Coming up to speed in a domain requires ever increasing numbers of years (or levels of specificity that stifle the science). Problem!


Anti-intellectualism is a problem which prevents the sharing of knowledge effectively with the general public; but intellectual snobbery and elitism is equally a problem which has exactly the same results. I say this because I get the impression that you and Sanger think Wikipedia has the first problem in spades, whereas I (and presumably Wales) think it avoids the second one effectively.
People need to get used to the idea that they are not born with all the knowledge in the world, and that they can better themselves and become more qualified to comment on certain matters. Justified expertise in the context of a society that respects (and defers to) it when appropriate is what we should aim for. I don't think these are equal problems.

I recognise that Sanger's approach would not have had as wide an engagement with the broader community as Wikipedia had. There's a balance to be struck, I just think Wikipedia got it pretty wrong.
Justified expertise is good, but I think that 1) in the past, a lot more science was done by amateurs than is the case right now, and 2) in the future, a lot more science was done by amateurs than is the case right now. Right now I see a lot of institutions -- journal publishers, and even universities -- locking up scientific knowledge. I agree that it's important to accept the knowledge of people who know more than oneself, but I also think that Wikipedia's success vindicates that idea that it's more important to welcome people with a good attitude towards helping others and sharing knowledge, than people who know a lot but are elitists about how they share it.
Er, s/was done/will be done/ in point 2.
Wikipedia might be a success, but I'm not sure it's a success at what it was trying to succeed at. Knowledge should be shared broadly but produced carefully.
I do not think siloed production of knowledge is scalable in the modern era. Perhaps this is our fundamental disagreement. I would much rather see knowledge produced broadly by anyone who thinks they have something to share, and then validated by trust, or by science, or by critical thinking and common sense.
I think you're right in how you boil the difference down.

I'm okay with the blogosphere and other types of public brainstorming, so long as we have a very very strong tradition of academia as the heart of how knowledge is produced, and that tradition is given primacy.

I'm not sure what validating by trust would mean, and I really worry about validation by "critical thinking/common sense" --- these are good things, but they discount expert judgement and heavy domain knowledge (in a time when problem domains are deeper than ever). These methods are also not guaranteed to be empirical and easily are sidetracked by people playing word games. The traditional training and collaborative methods of academia so far seem to be a good foil to the thorny bits of human nature that get in the way of science. Decentralising science feels risky at best; I don't think most people are up to coming to reasonable scientific (or medical) conclusions on their own.
Well, validating by trust means "I accept the things this person says as probably true because I have some reason to believe that this person typically says true things." Academic status, which you favor, is one reason to trust someone, but other reasons include "has said true things in the past", or "is spoken highly of by people who typically say true things", or "is someone I know personally, and thus have a good sense for what they know", and so on.

Anyway, I see the fundamental core of our disagreement very clearly in your last sentence, as I think that decentralizing science is a great idea and hope to see more of it in the future -- and indeed I think we are already seeing it.

As regards your final clause, this is slightly glib, but: ultimately people are going to make their own decisions on whatever basis they want, whether you like it or not, so you might as well give them the tools to be good at it. If people are bad at making medical decisions, that doesn't mean they're going to defer all their daily medical decisions to doctors; it just means they're going to make bad decisions. Better to help them make good decisions.
Or "(s)he is my preacher"? There are organised groups of ignorant people out there, and at least academia offers us (and the law) a known group of people with time-tested methods that already produce the best results. We exclude intelligent design, homeopathy, and other things from legal recognition, and don't need to waste time arguing with them anymore. I'm worried about losing the ivory towers as a thing to point to, and I'm worried that decentralisation will destroy science as an independent enterprise that guides society.
I think you are making what is from my perspective a classic mistake, which is to say 'wouldn't the world be good if', and proposing a scenario that isn't going to happen as a counterproposal to one that is. The world where everyone listens to the ivory towers is not a possible world, and it's not the world we're in. The world where we give everyone tools to try to find truth _is_ a possible world. The only way it can lose is if people who _already_ would look to the ivory towers for truth would now fail to do so, _and_ fail to find something equally good.

People who would say "(s)he is my preacher" are _already_ saying that. That they might continue to do so is not an argument for anything, as far as I can see.
Knocking down the ivory towers would constitute a loss. Right now, we're a discrete entity that's doing ok on its own, and we're hopefully holding our own against the barbarians outside. If we're to stand for future generations, we need to preserve that, and hopefully wait patiently by while other methods of seeking truth fall on their face. Crowdsourcing, in my estimation, will simply destroy academia if it makes it easier for politicians to paint intellectual achievement, scientific rigour, and the institutions needed for academia as unnecessary.

I don't buy that my hopes can't be achieved. A world where academia is respected is a possible world and one worth striving for.