s08e03: Flint and Kindling
0.0 Context setting
Monday January 20, 2020. It’s Martin Luther King Day in America, so here is a thread of book recommendations for children from Sarah Tuttle about the civil rights movement that’s a bit more realistic than MLK being about “Peace & Love” and more along the lines of how “I don’t see colour” is harmful.
I’m slightly more anxious than usual because my talk for Interaction 20 is due in (checks watch) about 48 hours and I am totally and absolutely ready for it, no doubt about it. If you are in Milan on Wednesday 5 February and happen to be at IXDA, then you can stop by and listen to my talk Computer, Stop: Why Star Trek only goes so far and we need to try harder than science fiction. Can I just say, unrelatedly, how completely anxiety-inducing it is to prepare a talk for an [interaction] design conference and be insecure about how well-or-not designed your talk is? I am also paying absolutely no attention whatsoever to the people following my talk, and the fact that my talk was scheduled to start the Futures section of the program.
1.0 Things that have caught, etc.
Look, I should just set a timer for these things to stop me from writing forever. I take back everything I’ve ever said about Neal Stephenson.
1.1 Flint and Kindling
This one is a bit more personal, so if you’re a recent subscriber, it might feel a little weird. Don’t worry though, it’ll all tie together in the end.
One particular either/or way of looking at yourself is to see yourself as a starter or a finisher. At first glance, this might feel pretty harmless, but I suspect that a number of people might grimace a little at the implication. A starter is someone who doesn’t finish things, which can feel very different from… whatever the opposite is? Can you have a finisher who doesn’t start things? (The answer, of course, is yes, but I’m getting ahead of myself).
A few years back, my therapist referred me to someone after wondering whether I had ADHD. The usual set of questions followed, starting through what your childhood experience of school has been, what it was like going through college, do you feel like you find it hard to stop talking and not interrupt people and so on, do you get distracted easily, do you leave tasks to the last minute, do you find that you perform well under pressure. Long story short: hey, you check off many of these, looks like you have ADHD, let’s try some medication.
Having some sort of theory behind your behavior — whether it’s right or wrong — means you start to have language and a framework to understand what’s going on. Turns out there are a number of underlying causes that can produce symptoms similar to ADHD without sharing the same brain chemistry.
The framework and scaffolding combined with therapy works a bit like this. Being distractible, from a therapy point of view isn’t inherently a bad thing. It just is. It might make some things easier and some things harder, and undoubtedly if you’re in therapy, it might be because you’re just noticing the things that are harder. It’s not like modern life doesn’t require significant amounts of focus and concentration to accomplish what might on the surface look like simple tasks: “What do you mean you haven’t done it? You just have to call up the health insurance company and fix the mistake in the date of birth?”
For me, the distractible part also works like a giant pattern recognition engine with its sensitivity dialed way up high. It’s easy for me to jump from one concept to another and find similarities, where those similarity signals might not reach a threshold of relevance of usefulness to someone else, or where they might just seem too disparate. The problem with being stuck in our own heads and our experiences being singular is that we have no idea if this is normal or not, and this ends up being compounded with whether or how comfortable you feel sharing your inner experience with, well, anyone else at all.
Turns out, jumping from concept to concept and holding them together in some sort of associative web isn’t something that lots of people do really easily. I can sort of say this now without feeling horrible because I’ve had a good seven years or so of noticing evidence that whenever I do it, other people seem to be surprised, or it’s a connection that hadn’t occurred to them and they find it relevant and useful. But… it’s just there for me.
For me, this ability to mix things together and relatively effortlessly find interesting new associations isn’t that great because the intuition pump is just too strong. I can jump and jump and jump and before you know it have created a chain of persuasive reasoning and associations that goes from what-we-have into what-might-be that, if I’m lucky, feels really new and turns out to be new-enough after a bit more validation and research.
The problem is the finishing part. I’m good at the beginning part, but less so at the filling in the detail — sometimes. The novelty of finding new connections and associations, of assembling a plan and drawing people together is a massive hit in the reward center. My brain likes [potentially bullshit evolutionary psychology citation needed] finding patterns and creating new ones. The existence of the pattern and external validation of the pattern can frequently be enough. I can stay inside my head and get reward after reward after reward for generating novelty.
Where it can all come crashing down is when you take a step back and look at this behavior and think to yourself: well then. I start things but I’m not great at finishing them. Worst of all, the thought that I can start things but I can’t finish them, which is pretty self-incriminating.
I‘ve done enough work to know that internalizing the phrase I can’t finish things is what we call negative self-talk these days. It’s also not true, right? I totally finish things all the time. The phrase is an example of a belief—a thought—that a brain can generate that is not the truth. It’s just a thought.
That’s the regular brain version. But the insight I had recently was that I can’t finish things isn’t just inaccurate, but harmful in a different, more insidious way than I had previously realized, because it wipes out the value of any of what I do. It implies that there’s no value in generative work, that all of the prep work and pattern recognition, generation of new patterns and so on is worthless without the input of someone else. Which, you know. Kind of toxic?
Anyway, here’s where I tie it all back together. You get personal essays and work-related insight here!
In “digital” and “internet stuff”, and the area that I’ve been working in for the last few years, one of the overriding philosophies is that nothing really counts unless and until you’re shipping. Software that exists in a human world for humans to use is based on a bunch of increasingly abstract assumptions that may well completely fall over on first contact with actual humans, instead of imagined humans.
Digression: are you the kind of person (sometimes stereotypically male) who imagines conversations with someone and those imaginary conversations never get anywhere, so you freeze up? And all the meanwhile, the person you’re supposed to having the conversation with is getting increasingly frustrated that you’re not communicating with them? If I may offer an analogy: you’re doing the same equivalent as writing a thousand-row Excel specification, attempting to de-risk and game-out what would be involved in a “successful” conversation. You’re literally speccing out a waterfall conversation that will be rolled out in a big bang, with just one shot at user acceptance testing. This is also a horrible metaphor for a number of different reasons (please don’t treat this as actual advice, etc.) but… maybe this is a useful way of looking at communication?
Some of the work I did that was at a genuine enterprise level was getting to the root of what people thought they would get by adopting an agile/iterative development methodology. First there was the whole business of getting past Agile and then into little-a agile. And then, there was this big part: if you’re not understanding user needs, doing the research to generate hypotheses and validate them, then all iterative delivery will get you is failing faster.
Failing faster is minimally better than just failing slowly, but it doesn’t get you any closer to your (hopeful) outcome, which I would assume is succeeding at delivering some sort of outcome. Without generative insight, you’re just flailing around in the dark like some kind of evolutionary algorithm and you’re not even sure what your fitness function is because your business customer isn’t articulating what outcome they’re aiming for, either.
All of this is to say: if your focus is solely upon shipping and finishing, I think you’re missing at least half the picture. Your research will generate insights. You and your team will need to come up with hypotheses, whether you realize you’re doing that explicitly or doing it implicitly “because this is how you build software”. But there’s a creative, generative process involved, and that creative process isn’t easy and it isn’t worthless without shipping.
For me, this has meant a very slow and gradual calibration and quite vigirously latching on to an analogy made by my partner after one after-dinner unloading of my frustrations. You can think of the generative part of work, or a generative person as flint. You can think of building on that spark as kindling and that together these are complementary skills and people. Not enough flint? Your kindling won’t get anywhere. Not enough kindling? Your flint is going to just keep generating sparks that don’t get anywhere (and, in my experience, will just get bored and go elsewhere).
I’m not going to pretend that other people haven’t come up with this metaphor. I’m aware of Simon Wardley’s categorization for, uh, digital transformation work, of pioneers, settlers and town planners. Simon is careful and deliberate in his piece to emphasize that pioneers are brilliant people; settlers are brilliant people; and town planners are brilliant people who all do different things at different times.
So, personal appeal: if you think of yourself as a starter and not a finisher, and you think this is a terrible thing, please consider the opposite.
And the professional, work version? Perhaps think about how these different skills are recognized in ways that they might not currently be, and how they fit together at different times.
… and if you’re looking for flint, get in touch. I’ll be available in May.
1.2 A (Short) Way Of Thinking About Service Design
A while back, I had the realization during a meeting that one of the reasons why digital transformation and legacy system modernization is difficult in government, especially in the area of using data to drive outcome-driven policymaking and delivery (sigh, so many… words) is that the legacy systems are systems of administration.
The challenge with encouraging organizations to adopt “service design”, say, is that roughly speaking, you’re probably moving from administering and operating a program to delivering and improving outcomes.
A very high-level (so, useful only to a certain extent) is to see how computer systems have been used until relatively recently as methods of administration and automation:
Service design for government services is the difference between paying attention to whether qualitative, non-administrative outcomes are achieved, and using a computer system as a very big and very fast filing cabinet
I realize this is yet another way of framing the policymaking versus delivery argument. But it’s fundamental, at least, in the current work that I’m doing. You will get very, very different results if your goal is to:
a) deliver an improved administrative system; versus
b) deliver a policy outcome
and that’s before you even get into the concept of improving those outcomes over time.
There’s a related thought — actually, the precipitating one — which was recognizing the idea that many [citation needed] people think many needs are just CRM, when really, the CRM positioning is barely above the hierarchy of understanding the problem than saying everything is just a database.
I mean, yes, you probably need data to operate on in order to do something but, you know, it’s probably a bit more complicated than that?
There’s yet another related thought here of the tyranny of SQL datatypes constraining the development of digital products, an argument of which I’ve made before where the ease of storing gender as a binary versus, say, a float, results in a whole stack of implications once real people start using software. In other words: what if people don’t fit inside the datatypes you first think of?
1.3 Smaller things
The New York Times had a piece on Ring with a droll subhead: “Ring offers a front-door view of a country where millions of Amazon customers use Amazon cameras to watch Amazon contractors deliver Amazon packages” of which my only comment is that the webcomic Spiders (2002) foretold crowdsourced citizens watching drone feeds for enemy combatants in the Forever War, but instead in 2020 we’re just using the internet and cameras to spy on each other.
I think it’s because I found myself on a webpage about old operating systems, but a very very long time ago, when I was younger than 10 years old, I got to use an early Apricot Computer. One of the things I remember about it was that the keyboard and mouse used “light-pipe” cables, which transmitted IR light. Oh, and they used a GEM-type interface.
via Tim Maughan, Sara Yasin’s tweet about a grocery store throwing a 1st birthday party for a security robot, and Adrian Short’s cutting question about whether the store ever holds a birthday party for the actual security guard. The poster for the birthday party promises “a coloring activity, birthday cake [and] fun giveaways” which, you know, isn’t exactly appropriate for a 1st birthday party. Hilariously, the grocery chain is called Stop & Shop, which only faintly produces connotations of beloved policing strategy Stop & Frisk. Marty at Stop and Shop has its own website, which is exactly what you’d expect it to be. There are over 300 Martys in operation according to that website.
Christopher Noessel of Sci-Fi interfaces is running The Fritzes, a set of awards for the best interfaces in a movie. I don’t think I need to tell you why this caught my attention.
Phew. Nearly 2,500 words and I’ve still got a bunch of other things in the queue that I’d like to write about: a thing about the Untitled Goose Game, game design and, uh, gamification; a thing about how we talk about AI/machine learning predictions and detection; and a thing about the Files app on iPadOS (ugh, isn’t the last one boring).
It is cold here in Portland. We had a powercut yesterday that lasted nearly 12 hours and let me say it was an experience with our young children, but also a nice preview of our post-apocalyptic life.
I hope you’re well and if today didn’t go so good, tomorrow is another day.
Best,
Dan