> targeted surveillance

This one is really interesting. I thought a lot about it after the revelations that Uber were targeting law enforcement based on user behaviour, and refusing to offer them rides in areas where they were breaking regulations.

Of course, one option is that the developers were fully aware of what they were doing, and were perhaps motivated by money or some sort of libertarian ideology.

The other option - more compelling in my mind - is that this functionality was carefully broken down into innocuous pieces and distributed on a need-to-know basis. One feature which groups users in configurable ways based on their history. Another feature that allows on-the-fly ride prioritisation based on user groups. Another feature that localises prioritisation to geographic areas. The upshot is that only some managers knew what the full potential of the functionality was.

Developers can't physically see the people that will be effected and they are usually just following orders from managers. It is shitty but it is way easier to know a bomb you were ordered to make is blowing people up on the other side of the world than it is to shoot someone directly in front of you.

There's a sufficient number of people personally doing targeted surveillance or pulling the trigger on weapons - and I'd assume that the barrier for building the tools for something is even lower than doing that yourself.

It pays well and gives you the certainty of your code having some heavy impact across borders, often against the kind of people you don't like as well.