Bandwagon effect: Bias in Evaluating AGI X-Risks

post by Remmelt (remmelt-ellen), flandry19 · 2022-12-28T07:54:50.669Z · LW · GW · 0 comments

Contents

No comments

Note on unusual formatting:  Sentences are split into lines so you can parse parts precisely.  

  The tendency to do (or believe) things
  because many other people do (or believe) the same.

  While this is similar in effect 
  to the 'Availability Cascade [LW · GW]',
  this bias effect relates 
  more specifically to the relation 
  between 'expert' and 'non-expert' opinions,
  whereas Availability Cascade is more in relation
  to the arguments and discussions among experts
  who are knowledgeable and qualified
  enough to consider and assess 
  the information directly.

  The concern here is that
  once the general non-expert public
  has been drawn into accepting
  a given proposal, proposition, or belief,
  the apparent boundary between

  1. the opinions resulting from
    real evaluations; and;
  2. the opinions resulting from
    people simply quoting other people,

  ... becomes very blurred.
  

  This means that it is  
  no longer possible for any party,
  regardless of all other factors,
  to easily tell if the analysis and evaluation
  has been independently replicated/validated,
  or is merely being quoted, copied 
  from one person to another.
  The net effect is that it becomes
  increasingly difficult to determine if
  the actual strength of the evaluation itself
  is due to multiple concordant validations,
  or that validation strength is actually absent.

  If most people are simply copying
  their results from someone else,
  that means we are all equally likely
  to have the same incorrect 
  or incomplete answer.
  This can easily lead 
  to a false sense of security:
  everyone believes we are safe
  because everyone else believes we are safe 
  – a result that could easily become
  completely detached from any
  sort of objective real basis
  or evaluative grounding.
 

  - link Wikipedia:  Bandwagon effect 
  - an item on Forrest Landry [LW · GW]'s compiled list of biases in evaluating extinction risks.

0 comments

Comments sorted by top scores.