[PREV - PAGES_YELLOWED]    [TOP]

KRUGMAN_WEBBED

                                             August  10, 2015
                                             October 04, 2022


I've got notes from and idea from years ago about
using Paul Krugman's writings as a case study in      This is an idea that
the junkiness of web search results.                  would've needed to be
                                                      done in the moment.
                                                      The wayback machine
                                                      will not get you back
                                                      to the google-eye
                                                      view of a web gone by.



                            (April 07, 2020)

  The thesis is that a websearch on a subject
  yields far too many blatantly wrong ("fake"?)
  hits-- a modest proposal: learn which sites are     ENGINE_EXP
  reliable, and go there directly to do a search.



An idea for a small research program:

  Choose a subject, e.g. Paul Krugman
  Count the wrong links.

You need a standard for choosing
which links are wrong, some way    This is not as hard as you might think,
of evaluating Krugman's remarks    because of the "zombie attack" strategy the
to pick where he's correct.        bad guys keep using.  They personally don't
                                   care if someone in the know has identified one
                                   of their talking points as being completely
 (In my experiments with this      wrong, they keep running with it as long as it
  circa 2015, I kept finding       seems to be effective propaganda.
  that the highly ranked links
  were changing too rapidly to
  come to any conclusion. One
  day the top ten was full of
  lies, the next they might be
  more neutral.)



Consider a few models of what's happening here:

  One is fairly straight-forward: the Right may have hired experts
  in SEO optimization, and may be directly funding the links and
  clicks necessary to bump their criticisms up to the top of the list.

  The other is more indirect, and I think more interesting:
  The Very Serious continue to have a lock on punditry,
  through the big newspaper editorials and cable news channels,
  and the various TV talk shows and so on.  This influences the
  majority of web users strongly, and they follow along linking
  to things that are in sync with that world view.

  Popularity is a weak guide toward truth.




    Which ever theory you back: it appears that the search
    engine, a major part of today's web, is broken as a piece
    of the infrastructure for our collective intellect.

                                                             INFOSTRUT


          If google can get the Krugman search so badly wrong,
          what else might it be getting wrong?


--------
[NEXT - SOUND_THE_ALARMIST]