The DSS team

Cambridge Analytica – has data science gone too far?

Op-ed piece.

3
votes

Disclaimer, I don’t follow the news. You might be wondering then why I’m writing about Cambridge Analytica?

Last week I attended a commission of the faculty of automation at TU-Sofia. One of the items on the agenda was the 4 year outline for my Phd work. When it was being described to the commission, the chairman made the following remark: “so you’re going to be doing similar things to Cambridge Analytica”.

He was joking and even though I wasn’t aware what they had done, I could tell it was something of criminal proportions. The chairman may have been kidding, when he implied that I would be involved in criminal activities, but he was completely serious when he said “there will come a time when we burn people like this at the stake”.

I didn’t have to justify what they did, since everyone knew that I wasn’t involved. Yet the association was there and I had no response beyond “I don’t intend to do something like that”. As far as reassurances go, that’s very weak. Anyone who tries to calm people’s concerns about data science being used in malicious ways better have much stronger points. It most certainly can be used against individual people and even against entire populations, just like any other tool humanity has developed.

This got me thinking about how data science is perceived by most people. Whenever you mention machine learning, AI comes up and with it the ever present SKYNET association. Or the more realistic “losing your job to robots” possibility. Do we as a community have an accomplishment, which is famous yet isn’t viewed in a negative light?

I’m not talking about small benefits people have as a result of analytics. Things like:

  • questionably accurate google translations;
  • youtube recommending similar videos to the one you just watched;
  • e-bay bringing up similar products to the ones you viewed.

I mean a big benefit for regular people, comparable to the harnessing of fire, steam, electricity, the invention of agriculture and so on. Does the data science field have something comparable and more importantly, do people know about it? The only thing which comes to my mind is google’s search engine, if that can be considered a data science accomplishment. Is this everything, or are there things which I’m missing?

Share this

Leave a Reply