A while back I blogged about riding the bus to work. we passed a car on fire in the opposite direction, two bus radio reports got the info correct, the third one, a woman, mistakenly said "eastbound lanes." Chatty McBusguy was all "ohhh, the *eastbound* lanes"
I was gobsmacked. it wasn't five minutes ago we saw the thing with our own eyes, in the lane and direction that wasn't ours, what was going west, in fact, and two reports matched the facts that you observed first hand, and still, one person on the radio misspoke and he restructured his whole reality. It knocked me on my ass because I realized that if a few people do that with things they *know* to be true, how many people surrender their reason to things they only think to be true because the tvbox says so?
worst of all, there's actually a logical argument that can be brought to bear in situations of institutional malfeasance like those he describes: the actions of one person don't have any bearing on the institution as a whole, because it is a fallacy to assign guilt by association. as a result, those who identify with the institution are able to displace the credibility problems and continue to identify with the larger entity. it doesn't matter if the protections and privileges afforded by the institution in question might be proximate causes for bad behaviors, since some people don't take advantage of those privileges
( ... )
that would be my assessment, as well. what i was trying to point out, though, is that there are alternative ways of approaching the question, and those alternatives may have something to do with why people continue to trust institutions that have proven themselves unreliable according to some methods of assessing them.
in addition, you rely on a "consistent pattern" in your assessment tools. while i agree with such a tool of assessment, it is one of the most easily arguable ones available when dealing with the public, as it usually requires serious analysis (to a level which is not available to the casual observer) to demonstrate a consistent pattern.
The fundamental unit of thought...bitobearMay 19 2010, 07:20:47 UTC
is the "do not want". It is the proton of cognition in that it firmly anchors all the molecules of opinion. The "do want" is the electron, in that it skitters here and there holding things together. The very first thought is "hunger? do not want" or "pain? do not want". The idea of the unfeeling artificial intelligence is bull, the unfeeling machine trope was about the giant wheels of the steam age picking up people and grinding them to death. The same can be said of the information age in a semantic sense. But our inability to transition into the AI age is that all the current programmers have fallen into the trap that intelligence can be modeled by pure rational logic
( ... )
Comments 5
I was gobsmacked. it wasn't five minutes ago we saw the thing with our own eyes, in the lane and direction that wasn't ours, what was going west, in fact, and two reports matched the facts that you observed first hand, and still, one person on the radio misspoke and he restructured his whole reality. It knocked me on my ass because I realized that if a few people do that with things they *know* to be true, how many people surrender their reason to things they only think to be true because the tvbox says so?
Reply
Reply
Reply
in addition, you rely on a "consistent pattern" in your assessment tools. while i agree with such a tool of assessment, it is one of the most easily arguable ones available when dealing with the public, as it usually requires serious analysis (to a level which is not available to the casual observer) to demonstrate a consistent pattern.
Reply
Reply
Leave a comment