Saturday, October 22, 2005

Too many stupid professors and bureaucrats...

I posted a blog a while ago whining about the annoyingness of the style of writing and thinking most common in academia today.

This is another one, with a slightly different slant.

At the end of the whining, however, I'll include an actual constructive suggestion for how to make some aspects of the academic world better. (Not that I expect my suggestion to have any actual impact!)

As I mentioned before, I've been making a push to submit papers and books for publication recently; something I haven't done much of since leaving academia in the late 90's. It's been quite an experience!

At first I thought I was doing something badly wrong. I have had some publications accepted but my rejection rate has been higher than I thought -- and not because what I'm submitting is bad (really!), mostly just (egads! can you believe it!) because it's unorthodox.

Of course, I'm revising and resubmitting and everything will be published in time. But the process has been educational as well as frustrating. And I've become aware that others whose work is even less radical than mine have been having an even more annoying time with this sort of thing.

I recently got two emails from friends reporting similar experiences to my own.

One is a biologist who recently left a major university for industry and has worked out a truly radical technique for repairing some types of DNA damage. This technique has now been demonstrated in live cells as well as in the test tube. Amazing stuff, with potential to cure some degenerative diseases as well as to slow human aging.

His paper? Rejected without review six times so far. WITHOUT REVIEW each time !!!

Another is an MD who has found particular patterns of DNA mutations that correspond to a couple very well known diseases. But -- oops -- these patterns are more complex than the ones biologists are used to looking at, and they occur in parts of the genome that biologists don't normally like to look at. So, no matter how statistically significant the results, he's got an uphill battle to fight. He's fighting against convention and presupposition. The result: right after he gets some breakthrough results, his government grant funding is cut off.

As compared to in the late 80's and early 90's, it seems much more common now to have things rejected without review. At least, this seems to be happening to me moderately often lately (though not a majority of the time), whereas back then I don't remember it ever happening.

A draft of my book on the Novamente design for general intelligence (not fully polished -- that's still in progress) was rejected by a publisher recently -- the rejection didn't surprise me, but the nature of the rejection did. The book wasn't even sent to a reviewer -- instead the editor just sent back a letter saying that their book series was intended for "serious academic works."

I had a bit of an email conversation with the editor, which revealed that he had shown the book to a "very distinguished AI professor" who had commented that due to the broad scope of the book and its claims to address general intelligence, it couldn't be a very serious academic work. Heh. Well, my ideas might be WRONG, but they're definitely just as serious as a lot of other books published. And the book doesn't contain a lot of mathematical proofs and only a handful of experimental results, but, it has more of both than Minsky's Society of Mind -- which also addresses general intelligence (or tries to) -- but wait, Minsky is old and famous, he's allowed to address big topics.... What we want to avoid is young people addressing big and interesting topics, right? But wait, why?

Please understand the nature of my complaint: I'm not pissed because this publisher rejected my book, I'm pissed because it was rejected without being read or even seriously skimmed over. And note that I've had six academic books published before, so it should be obvious to the publisher (who had my resume') that I'm not a complete raving crackpot.

I had the same experience with a couple bioinformatics papers I recently submitted -- which were nowhere near as eccentric as my book on Novamente, but presented algorithms and approaches radically different from what's typical in the bioinformatics field. Not just rejected --rejected WITHOUT REVIEW.

Of course, I also had some bioinformatics papers rejected after being reviewed, but by reviewers who plainly understood nothing in the paper. Of course, I could have tried to explain my methods more didactically -- but then the papers would have been rejected for being too long! Tricky, tricky....

Yes, I have had some papers accepted this year, and I have couple books (a futurist manifesto of sorts, and an edited volume on AGI) coming out in an academic press later this year. So these are not the whinings of a complete academic failure ;-p

I've been through enough of this crap before to realize that, after enough resubmissions, eventually one's books or papers hit a publisher or journal who sends them to intelligent and open-minded reviewers who actually read the materials they're given and either understand them or admit they don't (so the editor can find someone else who does). Eventually. But it's a long and annoying search process.

The academic community does reward innovators -- sometimes, eventually,.... But more often than not it places huge obstacles in the way of innovation, via a publication process that makes it much easier to publish variations on orthodox ideas than unusual approaches. One might argue that this kind of extremely strong bias is necessary to filter out all the crap in the world. But I don't believe it. Major changes to the reviewing process are in order.

Collaborative filtering technology would seem to provide a fairly easy answer. Suppose one assumes, as a basis, that individuals with PhD's (or MD's or other similar degrees) are, on the whole, reasonably valid raters of academic content. Then one can give each PhD a certain number of rating points to allocate each year, and let them use them to rate each others' work. People can then post their work online in resources like arxiv.org, and ratings can then be used to guide individuals to the most important or interesting works.

Journals aren't needed since the Net and computer printers are so widespread, and book publishers may still exist, but will be able to assume that if a book manuscript has received a reasonable number of rating points in its online version, then it's probably worth publishing.

You can argue that citations play a similar role -- but citations only play a role after a work is published, they don't help with the irritation of getting innovative ideas past conservative referees in the first place.

Anyway I don't have time to work toward implementing an idea like this, so I'll just keep working within the existing, annoying system, unless I manage to gather enough money for my research from business profits or private investments or donations that I don't need to worry about the often-absurd publication game.

Urrrghh!! I can easily see how, facing this kind of crap, young scientists and philosophers give up on trying to think wild and novel thoughts and follow along with everyone else.

Following along certainly would create a lot less hassle.

Or else giving up on the game of seeking reputation and simply wandering around in the woods like Zarathustra (Nietzsche's, not my son; my son Zar only wanders around these days in the simulated woods inside World of Warcraft!) and keeping one's thoughts to oneself (and then foolishly emerging to preach them to the world after a couple decades, only to find that no one understands what the HELL you're talking about...)

Humanity -- gotta love it...

Or -- hmm -- do you ???

Ah well...

7 comments:

LaBlogga said...

John McCarthy said at a recent Stanford lecture that the biggest problem in AI is a lack of radical new ideas. However, it is clear that his definition of 'radical' means within the narrow band of academically acceptable ideas. I like the arxiv.org suggestion and perhaps making more material available on your website, possibly working with a book agent and/or other intermediary/insider champion of your work.

Joel said...

Sounds like a cool idea.

As a bridging point you could include citations and links to the pre-existing journals allowing them to be rated.

One difficulty (among many I suspect) would be getting the print journals to accept references to articles in this rating scheme.

I need to fork() my life so that a copy of me can go do this! :)

ZARZUELAZEN said...

It's incredible Ben.

You're trying to do the most important in the entire history of life on Earth (and possibly in the entire history of the universe) and virtually all of humanity just couldn't give a stuff.

If anything tells us that the current version of humanity is unbelievably stupid and needs to be changed, this is it.

Whilst literally *billions* of dollars are spent by consumers every year on drugs, gambling, sports TV and pornography, Singularitarians can't even raise a couple of bucks, they are laughed at and called crack-pots , ignored and are reduced to scrabbling for every inch of free time, money and hardware.

You know, I used to be really worried that the government or others might go nuts and try to take-over or shut-down AGI projects, but I've actually come to realize that it won't happen, simply become most people just don't give a stuff.

Did you that Bill Gates read Kurzweils book, said 'That's Interesting' and then just moved on. Did you know that Gate's donates large sums to the DISCOVERY INSTITUTE (which is a Creationist institute) and none at all to Singularitarian causes?

And what does someone with Kurzweil's brains do? HE GOES AND WRITES A BOOK. Not trying to create a frigging AGI, not donating any frigging money, WRITING A FRIGGING BOOK.

Seriously Ben, no one will give a stuff about AGI probably until close to the very end (the final one or two years). People will still be calling Singularitarians cranks right up until the day before Singularity and the vast majority of people won't ever know what hit them.

Anonymous said...

elbloggo said "... I like the arxiv.org suggestion ...".

However, that does not work if you are blacklisted by the Cornell arXiv, as I am.

Tony Smith
http://www.valdostamuseum.org/hamsmith/
email f75m17h thenputat mindspring thenput dot com

PS - Ben, have you heard from Onar lately?

Anonymous said...

Ben,
I think the answer lies in sidestepping the academic journals (read: popularity contests) altogether, and using something akin to Robin Hanson's Prediction Markets in science.
Alcibiades

Anonymous said...

yeah, its amazing how many people are so narrow-minded. and what's even more amazing is that there are tons of them in the academe. makes you wonder...

Anonymous said...

I have come to realize that academia is generally an old boys club, and many of those old boys aren't actually all that good at thinking for themselves. University basically teaches us to synthesize other peoples' ideas; as an undergrad you learn not to say anything that you can't cite someone else as having said. Not really conducive to independent thinking.
Recent events in my life have helped me to realize that some categories of people (sorry to be so blunt, but mainly old white male professors, particularly those who are tightly linked with industry) are believed no matter what they say-- no matter how illogical, poorly researched, or lacking in rigor it may be. As a young woman who thinks and researches very rigorously, I am beyond frustrated-- because the shallow thinkers are driving decision making and are given credibility they just don't merit. Believe me, I speak up... but most people seeing this happen are just not smart enough to see what's happening, or can't be bothered to think it through.

Arg!

There's my rant.