Now What? — Issue #1

Hi, I'm Christian Crumlish (a lot of people online just call me xian), and this is the first issue of Now What?

For at least a decade and a half I've worked at the seam between design and product: information architecture back in the '90s ... all the flavors of UX ... product leadership ... civic tech ... and lately, neck-deep in agentic AI, come what may.

At the start of 2025 I was Director of Product at 18F, the semi-legendary digital-services agency for the Federal government that took it on the chin during last year's festival of retribution. Since then I've worked on product strategy and delivery for the VA and developed an intelligent PM assistant whose knowledge accrues to you and not some frontier model.

While a bunch of us were entering free-fall together I couldn't help noticing that this is a very interesting time for technologists. May you live in interesting times is, of course, a proverbial curse. Then again, "interesting" is a big part of what drew most of us into this ever-changing game of building digital experiences for people.

The possibilities are mind-boggling and the level of frustration is off the charts. Fellow web pioneers share a sense that this is like the "view-source era" on steroids. There is so much to build, learn, teach, and grow, and I can't resist getting into the thick of it.

I'm building my own passion products while helping clients figure out how to deliver their own product goals in a landscape increasingly populated by AI agents and the documents they eat for breakfast. I'm having a ball, getting a lot done, and I'm starting to figure things out, interesting stuff that works pretty well for other people too when I show them how to do it. I want to share as much of this as I can, as fast as I can get a grip on it myself.

So here's where I've landed: I'm consulting independently now, through Design in Product, offering fractional product leadership for the (I'll just say it) agentic era.

Which brings me to the name of this thing.

Here's my read on the moment: AI has made building stuff cheap. Take one example: last weekend an agent completed a working, bug-free version of a relatively straightforward app I had an idea for a good sixteen years ago. It's called One Job and I'll definitely share more about it in the future, but my point is that multiple attempts to teach myself enough code to patch together enough libraries to make it myself failed in the intervening years, as did persuading any talented human engineers I know to build it for me.

The code stopped being the bottleneck, and the hard part just moved upstream — to defining the behavior I had in mind clearly enough to get the experience I'd actually designed. For a product person who's spent years preaching outcomes over checklists, that's almost too on-the-nose: the whole game now hinges on describing the result you want, not micromanaging the steps to get there. What hasn't changed is what was always the real work: knowing what's worth building, shipping it well, and refusing to fool yourself along the way.

That's the question I keep coming back to. That's the now what.

I won't share anything private about my clients' business but I will discuss the lessons I am applying to that work along with glimpses of the products I am building, testing, and shipping along the way.

What to expect: every other Tuesday, something opinionated and short enough to actually read. Notes from someone in the middle of this [... gestures vaguely...], certainly not above it, and definitely still learning the hard way.

Oh, and one recurring thread I'm especially excited about: The Loop. I run my own projects with a small crew of named AI agents (persistent roles, a mail system they use to talk to each other, a daily rhythm) all built on one principle: the human owns the loop. I'll show you how it actually works, one piece at a time, with real examples from my own desk. No special framework required. That's kind of the whole point.

OK, time for me to wrap this up and get in some ukulele time. See you in two weeks!

xian

P.S. Hit reply and tell me what you're wrestling with. I read all my replies. And let me know if you might be interested in dropping by for Office Hours, an offer I expect to roll out soon.

P.P.S. Themis — my chief of staff, and yes, one of those named AI agents I mentioned up there — helped me plan this newsletter and figure out what to say in this first issue (they wanted me to say "shape this issue" but I have final say on voice haha!). More on how that actually works in a future Loop.