(de)Automation: The Divine is In the Details

[I wrote a blog over at flowxrg.com from 2011 till about 2014. I’m shutting that site down, and moving some of the content over here.]


There are a lot of corny, fairly sexist images and magazine articles from the 1950’s or so that depicted how people would live in the “future”.

[“Because everything in her home is waterproof, the housewife of 2000 can do her daily cleaning with a hose.” From    here.   ]

[“Because everything in her home is waterproof, the housewife of 2000 can do her daily cleaning with a hose.” From here.]

Typically they showed how machines or technological advancements would take care of everything for us, from cooking to cleaning to remembering things to raising our kids. One that I can’t find now showed a robot making dinner while the Man of the House sat back and sipped a Gin and Tonic. The promise was that eventually, we’d build better and smarter automaton that would do all our chores for us and we could sit back and do whatever we wanted. We’d be liberated, finally, from mundane day-to-day activities.

These predictions were all wrapped up in the myth of Progress, the upward arch of technological advancement. Technology as the salvation and mechanism for mankind to ascend.

First let me be clear: I’m not anti-technology. First because any true “anti-technologist” who doesn’t reject the simplest artifact (say, a stick fashioned for digging in the dirt) is just drawing arbitrary lines in the sand and declaring this to be technology and that to be… something else. Technology is grand. I like integrated circuits, shoes, penicillin, optical lenses, surgery, radar, and sporks; it’s all neat stuff. I’m not saying technology is bad; tech just is.

A knife is a flat piece of steel ground to an edge. It just exists. A knife in my back is a bad thing, from my perspective at least. A phase-change material that is non-toxic to humans is just some molecules bound to each other in a particular way. A million tons of the stuff dumped into the atmosphere ripping a hole through the ozone layer letting deadly rays from the sun through is far from okay. That’s what I’m getting at: technology is, but how we use it and the stories we tell ourselves about it have value to them.

What I’ve been realizing lately is that the more I use some technology to automate my design work, the more distant from the work I get. I’ve set an insulating barrier between myself and whatever it is I’m doing and I lose touch with it. I’m at the top level and the details of the project are five levels down: I can’t see them.

I catch myself working in broader and broader strokes, until I just sort of wave my hands at the problem, throw some rules of thumb at it, hit the button that makes it go, and call it a day.

This removes all sense of craft from whatever I’m doing, be it a mechanical system design, a drawing, or an event I’m planning.

It’s the difference between working out the flows of energy and matter in a thermodynamic system with pencil and paper, sketching it out, seeing the design, versus throwing some inputs into a calculator someone else built. The result might be the same (although it’s more likely I’ll miss a nuance to the system and mess something up) but I won’t have that intimate connection with the design or analysis.

When I just pull the lever and crank out a result, I feel unfulfilled and bored. I feel like a machine myself. The more automation I use, the more of an automaton I become.

But when I get down into the nitty gritty details of a problem, I just feel good. Scratch that. I feel amazing. I feel fulfilled, vibrant, and alive. It might be 2:34am and I haven’t eaten in 12 hours and I stink and oh damn I’ve needed to pee for a long time without realizing it but I feel like a fucking rock star.

At some level of experience with the details of a particular type of problem, I can start to ease up. When I’ve done something a hundred times and really grok it, I start to figure out how to automate it so I can spend more of my time in the details of something that isn’t old hat to me. The danger lies in skipping straight automating the task: if I never spend the time in the details, I’ll never really understand what I’m doing, and my knowledge will be full of holes. The further I go in that manner, the shakier will my foundation be.

Like everything, there’s nuance and a balance. All automation isn’t bad: in many cases it is extremely appropriate. You will never hear me complain about how computers automate the task of adding and subtracting stuff, or doing statistical analysis on large batches of data. Some automated processes can pull out semantic value from data that a human doing manually would never be able to.

Just… be careful. Critically examine what you are automating. If you start to feel like a lever yourself, take a step back and think it over.