top of page

AI versus the Polisters

  • Writer: Mark Skilton
    Mark Skilton
  • Nov 10, 2016
  • 1 min read

The pollsters got it hugely wrong in the US election. Putting the political allegiances aside of the popular tabloids, why was the data so completely off the target?


This could be due to sample bias in the results, even if the data was being interpreted as false positives on the way voters would behave.


Using AI might have been better with hindsight. If past training data used in models was incomplete, the use of "social listening" of social media conversations and behaviour may have been an early warning of possible contradictions from official polls. AI may also help in removing human bias by being factual and specific in its analysis. This may be difficult in practice if the number of dimensions of the problem that could impact the result outcome is hard to measure and define for a. Imputed input


We may not find out fully what happened, millions of dollars had been wasted in the campaigns and this didn't measure the network effect and outcomes successfully.


The AI field is still young in the industry. Massive progress has been made but like elections that only happen every 4 years it may prove harder to predict let along take real time actions to influence and manipulate the behaviour before it's too late.


Many voters may have been last minute choices of behaviour but even so questions must be raised over time the data sources and the algorithms that were used. All this is worthless if the last Mile step of the human editor then added bias into the use of the results

 
 
 

Comentários


Featured Posts
Recent Posts
Archive
Search By Tags
Follow Us
  • Facebook Basic Square
  • Twitter Basic Square
  • Google+ Basic Square

Mark Skilton    Copyright 2019  ©

  • White Twitter Icon
  • White Facebook Icon
  • White LinkedIn Icon
  • White YouTube Icon
  • White LinkedIn Icon
bottom of page