Episode 6 - Facebook Data and the Election, Decentralization to the Rescue
The big news right now is around an analytics company that used a dataset from Facebook - gotten by the company under false pretenses - to help the Trump campaign in 2016. Everyone is SHOCKED - but this kind of data mining was glorified and extolled as a social good when it was used for Obama in 2012. The comparison of the coverage is stark.
How are we going to get out of this? What exactly is the problem? I think that the vast centralization of data services has led to all sorts of problems, and the answer is well articulated in VC Chris Dixon's article from last month about the coming decentralized services. I summarize his article which adds to my optimism about the future.
Here's the corresponding article from the Guardian from the 2012 election.
Machine Learning Symposium in 2014 at the world trade center - the "Machine Learning for Good Talk" it turns out wasn't from a campaign consultant, but actually from Chief Data Scientist on the Obama Campaign itself.
Another article from 2012 in the MIT tech review which takes a more positive tone than we're seeing today.