Search Engines and Social Media: The Disaster of using Quantitative rather than Qualitative Rubrics for Evaluation

from The True Story of Fake News, Mark Dice

Often exposing bias in the media since forever (ho hum), rather than outright false information, but still, some things I learned …

“Twitter admitted to the U. S. Securities and Exchange Commission that they estimated over 23 million of their active user accounts were ‘bots’ or fake accounts run by automated computer programs which then post spam or are used by people to who buy followers so they can look more popular than they are.” p183

And that was in 2017!

“Bots are also used by services that sell ‘likes’ and ‘retweets’ that some people buy hoping to appear as if they have more ‘fans’ than they actually do.”  p184

In order to get higher on the results lists of search engines.

In order to get more income from the advertisers on their sites.

For the same reason/s, they post exaggerations and outrights lies; the more outrageous, the more clicks.

Which reveals the problem with using a quantitative rather than qualitative rubric for evaluation.  The first is easier and can be done by AI; the second is harder and beyond the capabilities of even most humans.  That is to say ranking by popularity is easier than ranking by relevance.  (And ranking by importance is near impossible without a clear statement of one’s criteria—which is something most people can’t formulate.)

(This is the problem with economists influencing government; the standard of living has little to do with quality of life; read Marily Waring’s If Women Counted.)

“Google is more than just a search engine.  It is the closest thing to an all-powerful information monopoly the planet has ever seen.”  p223

Well-put.

So don’t use it.

Dice presents some interesting comparisons of results on google, yahoo, and bing (233-234).

And here’s an excellent illustration of the importance of context (p288-289):

“NBC’s version of the call had Zimmerman on the phone with the operator saying ‘This guy looks like he’s up to no good.  He looks black,’ but the actual conversation was Zimmerman saying ‘This guy looks like he’s up to no good.  Or he’s on drugs or something.  It’s raining and he’s just walking around looking about.’ / “The dispatcher then replies: ‘OK, and this guy—is he black, white, or Hispanic?’/  “Zimmerman then responds, ‘He looks black.'”