What happens when AI, Psychology & Big Data drive politics? (Top3ics, February 19)

If there’s a single Top3ic running through the following stories, it’s probably Artificial Intelligence (AI), but I’m deeply into learning about psychology for the moment, so that’s my starting point.

You can have these links delivered on Messenger by subscribing to my CuratorBot (get a preview by asking it for ‘High3lights’).

The common factor behind Trump & Brexit: Psychometrics

We saw last year what happened when psychology and big (social) data analysis are combined and applied to political campaigning:

while other campaigners so far have relied on demographics, Cambridge Analytica was using psychometrics… “We have profiled the personality of every adult in the United States of America—220 million people”
- The Data That Turned the World Upside Down

Cambridge Analytica’s system crunches data to derive psychological models about people “… based on five personality traits… the “Big Five” … also known as OCEAN… openness, conscientiousness, extroversion, agreeableness, neuroticism…”. And then they use that data to run political campaign advertising  targeted to the personality of (potentially) each individual voter. 

Why should you care? Well, their recent clients include Donald Trump and the Brexit Leave campaigns, with sister companies ”involved in elections from Ukraine to Nigeria, helped the Nepalese monarch against the rebels… influence Eastern European and Afghan citizens for NATO…”. 

Forget about fake news. This is individually customised manipulation, operated at population scale for political purposes. 

More reading:

• Google, democracy and the truth about internet search, which includes this immortal quote from Vote Leave’s campaign director: “If you want to make big improvements in communication, my advice is – hire physicists.”
  
• 22 resources tagged privacy and 9 tagged big data

Elementary, dear Watson

Cambridge Analytics isn’t unique. IBM Watson’s Personality Insights also crunches big data to profile people using the Big Five model. Unlike CA, however, it does this upon demand, according to Chatbots Magazine:  

Watson … analyzes the content that you send and returns a personality profile for the author … This kind of finely-grained user segmentation could do wonders for bots. The experience becomes tailored to each user, therefore much more personal — and much more effective. 
-  Tailored Bot Interactions Using IBM Watson

The two posts are about the same technology, applied differently: one to manipulate people, the other to help provide a better service. 

These are, of course, not different at all: I just framed them differently. Like all technologies, outlawing algorithms would outlaw the benefits as well as the perils, something I’ve been discussing with the EU Commission’s Dan Sobovitz on LinkedIn.

(Aside: the Chatbots Magazine post is also exactly what I was looking for when I started seriously playing with bots a few weeks ago: an insight into the role bots could play in public communications. A blog post’s brewing.)

More background:

• 44 resources tagged bot
  
• Blogplug: my first and second posts for Chatbots Magazine
  

When Artificial Intelligence fails, we lose

So what do you get when you combine the above two posts? 

AI software running big data-driven ad targeting campaigns. In other words: 

• opaque, privately-run AI algorithms 
• analysing everything we do to continuously improve personality profiles of each and every one of us
• so they can run manipulative, completely personalised advertising campaigns 24/7, without human supervision. 

If that doesn’t worry you, maybe you need to understand that these AI systems are based on Deep Learning neural networks, so:

Nobody knows quite how they work. And that means no one can predict when they might fail… And when they do they fail … they fail spectacularly disgracefully. 
- Is Artificial Intelligence Permanently Inscrutable?

More reading: 

I cannot recommend that article enough. In fact, AI seems to stimulate many of the best longreads - optimistic and dystopian - I’ve read over the past 12 months. Here are a few more favourites:

• 5 Big Tech Trends That Will Make This Election Look Tame
  
• The Sadness and Beauty of Watching Google’s AI Play Go
  
• Humanity and AI will be inseparable
  
• Machine learning can fix Twitter, Facebook, and maybe even America 
  
• Understanding the four types of AI, from reactive robots to self-aware beings
  
• All 61 articles tagged AI 
  

If you found what I found interesting, get more by email or Facebook chatbot, which gives you interesting content in return for every keyword.