I’ve noticed that search engines, especially Google but also DuckDuckGo, seem to be optimizing for common types of queries, at the expense of less common ones. So if I want to find an Indian restaurant in a specific city, that works great. But if I want to find a restaurant with bench seating, that’s a much harder query to get answered. Google will fight you with your terms, replace with similar words, or ignore search terms, even if you use the plus operator, even if you put quotes around the term you don’t want dropped. Based on so much search data, Google’s algorithm now has a very strong opinion about what it thinks I want, and if the exact combination of words I’m looking for falls outside that scope, my query is internally munged to fit a more common pathway.
I’m going to call this the Expert Locksmith’s Dilemma, or just Locksmith’s Dilemma for short.
Suppose you decide to become a locksmith because you enjoy picking locks. It’s, challenging, fun, a nice mix of problem solving and manual dexterity, that ends with a nice endorphin rush when you “win”.
As a novice to journeyman locksmith, your job is a lot like how you imagined it to be. Maybe there’s more sitting in traffic than you expected, as you drive from one client to another, but there’s also lots of time spent puzzling out how to get people back into their cars and houses, fiddling with locks, learning new techniques and hardware, and getting to use cool tools like the Slim Jim (that long metal that slides into a car door to pop open the lock).
As you get better, you’re able to open most locks within a couple minutes. You’re faster, more efficient. This allows you to take more calls per day, and make more money, but it also means that the ratio of time you spend picking locks has gone down, relative to the time you spend driving around, talking on the phone, and invoicing clients. Over time, you find that the vast majority of calls fall into a few categories: trivial lock-outs you can resolve quickly, problems you flat-out can’t fix, and nasty slogs that require a lot of time but no interesting techniques, like gum stuffed into a key hole (bunch of savages in your town!). You still get the odd call that stretches your lock picking abilities, but these grow ever rarer, and problems fall ever more into either common locks you could pick in your sleep, and exotic setups that either have an electronic component and require the manufacturers assistance to fix, or hardware that has to be brute forced (drill baby, drill!).
Sadly, the better you get at your job, the less time you spend on the sweet spot of challenges that drew you into the profession in the first place. In all likelihood, you’ve turned your job into a lovely golden cage. As an expert locksmith who now understands the business side as well, you net too much per hour to walk away without severe angst, but the longer you do your job the lower your satisfaction with it. No matter how much creme and sugar you add to your morning coffee, it still has a strong aftertaste of burnout.
I suspect a large number of jobs suffer from the Locksmith’s Dilemma. I also suspect these jobs are much more common in the technium, because the technium is where craftsmanship goes to die, and craftsmanship is what prevents jobs from suffering from he Locksmith’s Dilemma.
+ + +
Compare the locksmith with the master carpenter. No matter how skilled, the carpenter is still going to take a lot of time to do his job. High quality work, craftsmanship, takes time. A beautifully curved, constructed, and stained staircase is impossible for a novice, and a multi-day task for the master. Expertise opens up more possibilities to work on projects that are likely to please the carpenter who chose her job because she loves building gorgeous wooden structures. There’s no dilemma here, only a virtuous circle as better skill opens up the chance to work for richer clients who are happy to let her spend three weeks crafting their grand master staircase.
Every individual carpenter’s experience in the field will vary, of course. Some may burn out even though they get to work on interesting project, and everyone who works on houses is subject to the feast and famine cycles the industry goes through. But even if the job is to build generic cabinets for a generic kitchen, the majority of our carpenter’s time is still spent actually building cabinets, and the overall picture remains: jobs with a high degree of craftsmanship are resistant to the Locksmith’s Dilemma.
The tools our carpenter uses are for sure technology, but they’re tech that hasn’t changed much in decades, and they require a high level of manual dexterity and effort to operate. The distinguishing feature of technium jobs is that the majority of the effort is applied indirectly, at a meta level.
What might it look like if our carpenter’s job might job drifts into the technium? Imagine — and this shouldn’t seem like that much of a stretch from where we are now — that all of the cutting, sanding, and staining work for a build can be automated using machines controlled by a digital interface. This brings with it a massive increase in efficiency. That grand master staircase now takes a day to design using CAD (computer aided design) software, or just a couple hours to browse pre-existing files from a library of designs and tweak one of them for the needed dimensions. The program generates a complete set of instructions for lumber preparation, and these are sent out for manufacture. When they arrive at the job site, our carpenter spends a couple days assembling all those perfectly sized, uniformly stained pieces. No measuring, no cutting, no sanding or finishing.
Our carpenter has gone from software operator to furniture assembler. And while there may be pleasure in putting together a gorgeous, thousand piece item that arrives like a giant Ikea flatpack, it’s clear that technology has shifted how she does her job. It has much less shaping wood, and much more shaping virtual bits. Simple tools that take a lifetime to master have been replaced with complex software that is awesomely powerful, wonderfully efficient when working, and yet likely exhibits a large number of quirks (or bugs, if you will), that will frustrate our carpenter on many occasions, and are likely to shift with version update. Our carpenter now spends a lot of time on forums learning how to program a dovetail joint, and less time with a chisel perfecting them.
The technium automates the mechanical and drives effort up to the level of tool and process management, leaving untouched only those tasks we haven’t figured out how to handle with robots, at least not yet. Tasks like driving to job sites (until we get self-driving cars), communicating with clients (until our company is big and powerful enough to force them into voice mail hell), and carefully picking bubble gum out of a lock without damaging the tumblers.
+ + +
This is not intended as a Luddite lament. The technium has an upside, and it’s huge. That gorgeous staircase that cost $30,000 can now be build for under $5,000, meaning it can be added to many more homes. I may prefer to never buy another piece of prefab, pressed-wood Ikea crap, but over the years I’ve relied on dozens of Malms, Dagstorps, Fartygs and Billy Bookshelves as reasonably priced, reasonably attractive, highly convenient furniture. My twenty four year old self was more than happy to buy a dresser for two weeks worth of savings instead of two months.
To a large degree, and I am by no means the first person to suggest this, our lives have become much better as consumers, but worse as producers, in particular if our job could have once been described as a craft, before the skills it required became more meta than physical.
As I work on my escape from the technium, I’m trying to go the other way. Less time spent fighting with my Ubuntu web server, more time spent laying down ink on paper with a lovely and reliable Cross pen (I wrote the draft to this post on paper). At best, programming is still a craft (and there’s lots of good writing out there about the craft of coding), but the job suffers from a particularly nasty version of the Locksmith Dilemma, which I’ll explore in future posts. Writing, at least for now, is still a craft.
Maybe it’s time to clear out the canon. The old and the new standards. Adios Homer, auf Wiedersehen Vonnegut, fuck off Orwell, Ellison and Morrison. Don’t even mention Shakespeare. All gone. Let’s start from scratch, wipe the literary slate clean, start anew. Make room for new voices and new ideas. The conversation has gotten stale.
Everything takes longer, even when you take that into account. One of the iron-clad rules of the world. Taught to me by my ex.
What happens when all those lessons add up? Episode of Freakanomics Podcast this week takes on the “planning fallacy”, which is a weaker form of the rule I was gifted two decades ago. It says we suffer from an optimism bias, which leads to overestimation of how much we’ll get done and how long it will take. Apparently, this bias has some big upsides. Optimists live longer, are happier, and are generally, just, more optimistic.
So what happens if that optimism bias finally burns off? What if you’ve run through the cycle of over-predicting how well some new project will go so many times, that you no longer believe your own hype? What happens if your own reality distortion field suffers a breach?
We pretend that definitive, discrete, non-overlapping categories can be defined without ambiguity. We pretend we don’t have to talk about the meaning of words, that it’s self evident. We pretend that what we’ve built has a solid foundation, and isn’t just turtles all the way down.