Dear Dave,
A few weeks ago I wrote about speed vs momentum and the perceived perils of process in tech. We do not trust the process. We often loathe it, in fact. The process slows us down and must be disrupted! (Cue pitchforks.)
I also wrote a while back about solving problems in our AI solutions-focused era, noting that:
“The design process has always been adaptable done right — grabbing the pieces we need, leaving behind the ones we don’t, moving at the speed demanded by the problem at hand.”
As you can likely glean, I am something of a process nerd. I love seeing how people adapt and remix processes to solve problems in new ways and pull humans together to do hard things well. There’s delight in the “product-y” way of looking at the world — always looking for opportunities to shift left, try things differently, and see what learning comes out, not just in what we ship, but how we ship it.
I’ve always been conscious of the Feelings process can inspire in others — whether negative (“boo meetings!”), unproductive (“more meetings!”), or dismissive (“don’t talk to me about meetings!”).
But process is how things happen. It reflects the deeply human nature of product work: the thinking, the doing, the dot-connecting we do together as a team to make something other humans find genuinely useful.
And lately I’ve noticed a fascinating development: suddenly, everyone seems to care about it.
Specifically, there’s a new emphasis on process improvement through AI — in particular, a fixation on personal software, tools you build for yourself at work whose only purpose is making you happier and more efficient. We are being more prescriptive about process than we’ve been since the days of waterfall and scrum, and the prescription is simple: use AI.
Dave recently wrote about AI learning and development, and I’d add: it’s a marvel how much space we’re being afforded to learn these tools on the job, often without any specific metric besides usage (yet).
It’s a curious phenomenon, and one many find rightly frustrating. Adopt AI tools — why? For what? We might respond: tell me what problem you think this can solve, and let’s see whether it actually makes sense.
But tucked inside even a blunt mandate is an invitation: to look at how we work, interrogate the defaults, and ask whether what we’ve been doing is the best way to do it. That’s what I’m embracing — the focus not only on building, but on how we build.
Process, at its best, is just that question, made repeatable.
From shuddering at a distant memory of the job title “Scrum Master,”
Danielle
This is great!
I feel AI is shrinking the “perceived” need for collaborative process. People can do more without feeling like they need help. Sometimes that’s good, and sometimes that means things get battle-tested less, or flimsier assumptions survive with a coat of polish.