MLQ3 on Tumblr!

Should information flows be controlled in the network age? Who should get to decide who gets access to what information? It’s not as if these questions have only been asked for the first time because of the Internet. The many generations of people that learned how to build democracies wrestled with them over the centuries.

Privacy is not about anachronistic prohibitions on information flow, but about personhood.We know what the answers are. If the secret is about something that isn’t a vital interest for other people, then everyone has a right to keep a private sphere private. If the secret is about something of vital interest to other people, then secrets can be kept by those who are sanctioned and accountable to keep them within the bounds of a reasonably functional democratic process.
Both of these answers are under assault by the ideology of nerd supremacy which I understand well, since I was part of it in its early days.

You need to have a private sphere to be a person, or for that matter for anything creative to happen in any domain. This is the principle I described as “encapsulation” in You Are Not a Gadget. I have written about this idea in various ways, but I’d like to try another way here, addressed to the truest believers. Let’s consider encapsulation in computer code.

There was a time when computer code was messier, in that any piece of code could read or write to any other part. That didn’t work out well. Programs were too tangled and impossible to maintain.

So a movement to add structure to programming took root. For instance, the idea of “object oriented” code breaks a program up into encapsulated modules centered on chunks of data and code related specifically to that data. If you program in an object oriented way, you are not allowed to make the code in one object directly manipulate the interior of another. Instead, everything has to go through the proper channels.

A great many programmers hated the object oriented idea in the early days. It seemed like nothing but prissy restrictions. To others, it was simply incomprehensible how restrictions would do you any good. Wasn’t the point to be able to program anything? How could a negative be a positive? How could restrictions improve results?

And yet, ideas like object oriented programming were essential to making big programs reliable. The world we know today couldn’t exist if code had stayed as messy as it used to be. Structure is what makes information usable. Making everything totally connected and open to everything destroys structure. This principle works for code, but it is also cosmic.

Even we people need structure in our affairs. Imagine openness extrapolated to an extreme. What if we come to be able to read each other’s thoughts? Then there would be no thoughts. Your head has to be different from mine if you are to be a person with something to say to me. You need an interior space that is different from mine in order to have a different, exotic model of the world, so that our two models can meet, and have a conversation.

Privacy is not about anachronistic prohibitions on information flow, but about personhood. I was one of those young hackers who didn’t get this point for a long time. I used to think that an open world would favor the honest and the true, and disfavor the schemers and the scammers. In moderation this idea has some value, but if privacy were to be vanquished, people would initially become dull, then incompetent, and then cease to exist. Hidden in the idea of radical openness is an allegiance to machines instead of people.

Improving access to information can be a very good thing in the right circumstances. For instance, another huge factor in making code better (in addition to structure) was a flow of information feedback from the real world.

We sanction secretive spheres in order to have our civilian sphere.
Coding used to be based on hope. You’d code something and someone else would experience whether it crashed or not, and while they would let you know, it was hard to learn much from their tales of woe. With the arrival of the Internet, crash logs could be reported back to the programmers automatically, so software engineering became a closed loop feedback system. I well remember Steven Sinofsky showing me the early results of this flow of data about crashes in the early Windows operating system. It was as if a new sense organ had suddenly sprouted on one’s face.

I bring this up to say that asking whether secrets in the abstract are good or bad is ridiculous. A huge flow of data that one doesn’t know how to interpret in context is either useless or worse than useless, if you let it impress you too much. A contextualized flow of data that answers a question you know how to ask can be invaluable.