Re: Why growth may stay slow

Michael Lorrey (
Sat, 21 Mar 1998 11:08:39 -0500

Dan Clemmensen wrote:

> Robin Hanson wrote:
> > Dan Clemmensen writes:
> > >Thanks for the clarification. I'm interested in growth of the corpus of technical
> > >knowledge and its rate of increase. Your argument appears to apply to growth
> > >of the measured economic value of the technology, and may be very valid.
> >
> > Actually, the corpus of academic knowledge, as measured by the person-years of
> > academic research and by published articles, has been doubling every decade I
> > think for ~150 years. Recently this rate has SLOWED DOWN as academic has
> > reached budget limits. See "Little Science, Big Science" book.
> >
> Hmm, this would mean that we are producing roughly 32,000 times as many papers
> per year as we were in 1848. That's a lotta paper. I haven't read your reference, but
> I do know that the acedemic publishing business percieves itself as being in crisis.
> In today's world there may be disencentives to publication such that the publication
> of a paper is not a linear measure of an increment to the corpus. I have no reasonable
> alternative, though.

The alternative to the academic publishing establishment is self publishing, via xerox,
ditributed print on demand, HTML, etc. Easy to use desktop publishing software is the
threat to the academic publishing houses, not a dearth of publishable material. They just
can't compete as they are structured now. If they wished to get with the program, they'd
evolve themselves into internet content providers of scientific papers online, in HTML,
PDF, PS and other formats, providing a commons for peer review online, using backlinking.
Those that don't will have steadily narrowing business.

> I do know that there is a tremendous amount of technical progress that is not captured
> in papers in peer-reviewed journals. In my field of data communications, a lot of
> stuff is in journals, but much more is in specification documents, code, RFCs, and
> standards documents. A great deal more is in patents, trade secrets, and other
> unpublished material thta is nonetheless part of the technical background information
> that the industry builds on. Programmers and hardware engineers learn this stuff and
> then take it with them from job to job because it sinks into their general body of
> knowledge. It's just not possible to remember where you learned each neat trick and
> what the IPR status of the idea was, so only blatant IPR infringments are likely to be
> prosecuted or even remembered.: this stuff just becomes part of the state of the art.
> Can you recommend other measures of technical progress? Mine are almost all apocriphal
> or qualitative.

a) average speed of various commercial transportationb) average energy consumption per
capita, times the average energy efficiency per capita.
c) in combat: i) ratio of wounded to killed (higher tech will have greater killed, less
ii) survival rate of wounded
iii) number of 'rounds' expended per KIA (this includes everything from
bullets to ICBMs)(smarter weapons will have more kills per 'round')
d) average time to download a 10 file, 300k HTML page, or just average modem speed.
e) average size of the average credit report
f) how much time it takes to clean the mildew out of my tub ;) today is cleanup day....

   Michael Lorrey
------------------------------------------------------------ Inventor of the Lorrey Drive
MikeySoft: Graphic Design/Animation/Publishing/Engineering
How many fnords did you see before breakfast today?