The Pocket Test

(Last week's Broadcast was the love letter to a small genre of consumer software that wants you to leave home. The replies were generous, and several of you wrote back asking variations of the same question, which was, roughly: "isn't the phone still in your hand though?")

I tried. I lost.

The audio drops out of my head the second the screen wants to be interpreted. My attention has a hand, my hand is on the screen, and the podcast keeps talking into a room I have already left.

That single failure cracked a question open. What else is the same shape?

Five walks over a few days after work, same neighborhood, different attention stack each time.

One: Ingress + podcast. The original failure. Walking with an audiobook does not measurably wreck your gait, but walking with an audiobook while also doing a spatially interpretive task is a different story. The second your walking becomes hunting, the audio stops being background and becomes static. (Niantic seems to know this. They sell an official Bluetooth accessory whose entire purpose is to let you play Pokémon GO without looking at the app. Their own hardware is a workaround for the attention cost their own software creates.)

Two: Ingress + person. Last weekend, L and I walked a few miles through a stretch of Back Bay I had not really seen before. I stopped every couple of blocks to grab an Ingress portal. The walk was nominally shared. The portal was, as it turned out, solo work I happened to be doing next to her. It pulled me into the place and pushed me away from the person walking in it. The partner-phubbing literature has a name for this. Co-present phone use becomes harmful mostly when it reads as withdrawal, and pause-to-portal reads as withdrawal even when it doesn't look like one. Groups containing a phone user walk measurably slower than groups without, and the choreography itself changes even before anyone says anything.

Three: solo, music only, no game. A different attention state. The music made a floor, and thinking had room to roam above it. A pleasant baseline I had forgotten was a baseline.

Four: solo, podcast only, no game. Different from the others. Sometimes I was deeply paying attention. Sometimes a host said something and my mind drifted into problem-solving: a question, a connection, a phrase worth keeping. Walking roughly doubles creative output over sitting in the same room, and the podcast doubled as a Signal-generator on top of that. The cost is that the Signals tended to evaporate before I got home. Capturing them mid-walk turns out to be its own design problem, and I do not yet have a good answer for it.

Five: forced by weather. A stretch of May rain wiped out my normal walking practice and forced me into longer make-up walks to hit my step target. The marathon walks exposed what the daily ones had been hiding. What I had been calling "walking" was actually four different things stacked into the same word. Exercise. Thinking. Escape. Companionship. Each of the variations above did some of those well and others not at all. It took two weeks of rain to spot that.

Six: no phone, no audio, no artifact. Solo or with a companion. The baseline against which every variant above registers. Ninety minutes in nature significantly decreases rumination and stress-linked neural activity compared with ninety minutes in an urban setting. A separate study found that even a brief greenspace walk measurably improves directed-attention performance by similar margins. The unplugged walk is upstream of every variant above, and reaching for any of the others is reaching past it.

But the most common version of "no-tech walk" in the wild isn't really no-tech. It's scrolling mindlessly while the dog bounds and the kid runs and the stroller rolls, while the actual companion has the walk you are not having. The pocket-leavers at least have a game-loop justifying the attention cost. Mindless scroll has no loop, and the dog watches.

The phone moves your attention rather than multiplying it, trading one slice for another: the street under your feet for the street on the screen, the person walking with you for the destination on the map. Smartphone walking is a layered negotiation among "streets as experienced through the body," "streets on the screen," and "streets in the mind." Three streets, one attention, and you are picking between them whether you notice or not.

Last week's newsletter said the screen should leave your hand. This week's: it matters which hand picks it up, and who is standing in the room when you do.

Before you open the app on your next walk, ask one question.

Does this app, used the way I actually use it, stay in my pocket?

That is the test, with five answers and five kinds of tool. Most of the tech you carry is somewhere on this list, and some of it crosses categories depending on how you hold it.

Pocket-killers. Turn-by-turn directions and highly automated driving live here. Your body arrives but does not learn the route. Convenient, and with a real cost. These do not really compete with the companion next to you, but they thin the experience of the place itself to almost nothing.

Pocket-leavers. Ingress, Pokémon GO, the map app you check every block in an unfamiliar city. The phone keeps coming out and going back in and coming out again. This is the category most of the genre I wrote about last week actually lives in, even the ones whose marketing implies otherwise.

Pocket-friendly. Last week's audio walks (ECHOES, Detour, Cardiff's Forest Walk, Halsey Burgund's Roundware installations) all live here, plus museum-specific guides like SFMOMA's audio-first app. You set up the walk, you put the phone in your pocket, headphones in, the app does the rest. The screen starts the experience and then disappears.

Pocket-honest. Last week's pocket-honest catalog lives here: eki stamps, National Park passports, scratch cards, letterboxing on Dartmoor, the geocache log on the state line, the orienteering merit badge. No screen at all. The artifact picks the destination, scaffolds the choice, and then the entire walk is between you and the place.

Pocket-free. No phone, no card, no app, no artifact. The walk that needs nothing to be a walk.

Before opening any app on a walk, name the pocket it lives in for you, this week.

A new Probe in progress. Inspired by last week's scratch cards, I am building a tiny tool that picks an activity for L and me by text message when we cannot decide between options. SMS, not cardstock. The interesting design question is which tier of the Pocket Test it actually lands in: the screen delivers a single prompt, then ideally the walk happens without further screen contact. Pocket-friendly if the message stays put. Pocket-leaver if I cheat and keep checking. More once the Probe runs. Text-message-to-audio prompts… or even phone calls.

India just banned phones at four pilgrimage sites. Uttarakhand authorities cited "smoother, distraction-free darshan" as the reason. This is the Pocket Test written into administrative policy: an institution deciding, for an entire holy site, that the device and the desired presence are not compatible.

Blast Theory's Can You See Me Now? (2001). A locative game built explicitly around the tension between presence and absence, twenty-five years before consumer AR got around to remembering the question mattered. The art world was here first. The art world is almost always here first.

The weekend-warrior exercise studies. The American Heart Association has confirmed that marathon make-up walks deliver roughly the cardiovascular benefits of distributed daily walking. The research is silent about whether they deliver the other three uses from earlier, companionship most of all. The literature has the same gap my rainy May did.

Spark capture mid-walk is its own design problem. When a podcast spawns a Signal, or a corner does, the Signal usually evaporates before I am home. Voice memos work, but they make you say it out loud in public and change the privacy of the thought. Apps with screens break the pocket-honest principle. Pen-on-a-clip works, but only at a stop. Open design space, no good answer yet.

Talking to AI on a walk is weird in a different way. Voice-AI in the pocket is now a category, with hundreds of millions of users and growing. It does not fit cleanly into any of the tiers of the Pocket Test. The eyes are free, the body keeps moving, the conversation is real-time, and yet you are out loud with no one there.

The question to leave you with: which presences are your apps quietly choosing for you?

The pocket is the test. The pocket is where the app's marketing meets the actual room you are walking through and the actual person walking next to you.

If you take only one thing from this issue, run the test on whichever app you opened most this week. The next time you reach for it on a walk, you will know what you are about to spend, and on what.

/ David

On repeat this week: The Long and Winding Road, The Beatles, 1970. 🎶 "The long and winding road that leads to your door..." 🎶 (Yes. The door is still the metaphor.)

Reading: A Philosophy of Walking by Frédéric Gros, 2008. Gros pulls the different uses of walking apart better than the research does, and with more sentences worth underlining.

Building: A text-message activity picker for L, inspired by last week's scratch cards. SMS delivers the prompt, and the walk runs without further screen contact. Forthcoming.