Don't be evil

Data analysis and personalized content can be used for evil purposes, but also to achieve the opposite.

Don't be evil

Video transcript

Hi, my name is Sander, CEO at Unless.com. We provide personalization for websites.

The past month, the internet industry took a pretty bad blow. You probably heard of the Cambridge Analytica scandal. Aside from their CEO declaring that his company used honey traps, bribery, and even prostitutes for "opposition research", the allegations against Cambridge Analytica are mostly about privacy and ethical stuff.

So, what did they do? Cambridge Analytica collected voter data from demographics, consumer behaviour, internet activity, and many other sources. Their goal was to do "psychographic analysis". The company then used those psychological profiles to influence individual people using microtargeting - by sending them the messages that they were most likely to respond to.

Now, where did it go wrong? I mean, aside from the prostitutes and bribery, of course. Cambridge Analytica used data from millions and millions of Facebook users, without their permission or even knowledge. That's probably against the law in the US, but most certainly illegal in Europe with its far stricter privacy laws.

And it gets worse. Cambridge Analytica then used this data to manipulate voters in the UK Brexit campaigns and US elections. With fake news, echo chambers and filter bubbles on social media, the company manipulated voters with the most vulnerable psychological profiles. And now, there is public outrage, and rightfully so.

Where does that leave us? I mean, we are in the personalization game, too. Unless aims to achieve exactly the opposite of what Cambridge Analytica did. I will explain why. By agnostically adapting internet content to match your context, we make complex subjects easier to relate to - including opinions that are perpendicular to your beliefs. This counters the echo chamber effect and leads to improved decision-making.

So in short, we will never change WHAT you read, but instead we aim to improve HOW you see a message. For all sides involved.