Hands-On Chief Technology Officer

1 day ago

Disclaimer

Abstract:

The article argues that remote work didn’t only eliminate commuting—it erased the “thinking container” created by casual side-by-side moments (like a 90‑second corridor sync or a water refill chat) where half-formed ideas and real risks could surface without becoming formal meetings, and that replacing these with instant-on video calls (same chair, camera-on, subtle delays, “being watched” pressure) makes people more careful than exploratory while leaving lingering attention residue after the click. Drawing on research about video fatigue, turn-taking friction, attention residue, and walking’s boost to divergent thinking—and filtered through the author’s stated bias as a French, 1974-born, physics-trained person who looks for missing environment variables (plus a personal note about desk-life across Beijing, Berlin, and Lisbon and the telltale upper-back tightness)—the piece proposes rebuilding a “cheap pre-thinking layer” with small, realistic changes rather than a lifestyle overhaul. It recommends using walk-compatible, often audio-first conversations for fuzzy-front-end work (1:1s, unblocking, brainstorming, debriefs, pre-mortems, career chats) while keeping desk mode for screen-share, numbers, live editing, and larger groups, guided by a “good enough switch rule”: walk 5–15 minutes to agree on a one-sentence problem statement and 2–4 options, then sit 5–15 minutes to validate constraints and write a timestamped decision log with owners and dates. The article also tackles constraints (privacy/compliance routes and headset use, avoiding quiet echoing spaces, not triggering the Lombard effect, lightweight note capture, and accessibility so walking doesn’t become a participation tax) and suggests a one-week experiment changing just two recurring meetings and tracking felt outcomes like fewer mental replay loops, faster time-to-clarity, and less end-of-day jaw/shoulder bracing, keeping what “pays rent” and dropping what doesn’t.

Remote work did not just remove commuting. It removed a whole thinking container.

The tiny side-by-side moments used to do quiet work for teams. The 90-second corridor sync. The walk to refill water where someone finally says the real risk behind an “all good”. Those micro-conversations were not a break. They were the cheap pre-thinking layer, where half-formed ideas could come out without becoming a calendar event. They also forced tiny transitions: shoulders moved, eyes refocused, and your nervous system got a clean “that meeting ended” signal.

Remote default replaced that with instant-on calls. Same chair, same screen, same posture, camera-on when the topic is still fuzzy. The result is pretty predictable if you treat meetings like software. Same people, same agenda, different runtime, different output.

Here is the predictable failure mode. A product and ops group jumps on a 45-minute video call to “align” on a delivery risk. Nobody writes a one-sentence problem statement. A few people start debating edge cases and Jira fields. Someone says “we can’t decide without numbers,” but the numbers are in another doc nobody opened. The call ends with “let’s take this offline,” and the same topic comes back next week under a new invite.

Video adds more “am i being watched” signals, small delays break the natural back-and-forth, and people get careful instead of exploratory. Then the call ends with a click, and your brain keeps running the meeting in the background.

This piece is about putting that missing layer back, without pretending everyone has time for a lifestyle overhaul. It covers:

why remote days create hidden taxes on decision quality, conflict repair, and attention residue
when walking meetings help, and when desk mode is non-negotiable
a “good enough switch rule” that keeps walking for framing and the desk for decisions
practical constraints like privacy, note capture, and accessibility
a small 1-week test with a tiny scoreboard, focused on outcomes people actually feel

The thinking container you lost

The side by side channel was not a break

Those tiny side-by-side chats did more work than most calendars admit. A 90-second corridor sync could settle who owns the next draft. A walk to grab water could surface the real risk behind an “all good” update. Peters and Waterman called this MBWA, and Packard wrote about walking as a way to keep information moving. It was a medium, not a pause.

That social ease changes what kind of thinking a group can do. Side-by-side has less gaze and fewer “am i being evaluated” signals. Video pushes the opposite. One well-known driver here is the hypergaze/self-view effect (Bailenson): you end up managing your face and posture while trying to think.

The behavioral outcome is simple: people get careful instead of exploratory. Small delays also make turn-taking clunkier, so you get fewer natural overlaps and pauses. Half-formed ideas stay inside the head, where they turn into anxiety.

A useful model is to treat meetings like software. Same people, same agenda, different runtime, different output. Rogelberg’s meeting research helps here because “effective” is not vibes, it is fit to task.

Walking talk runtime: quick synthesis, soft alignment, early risk surfacing
Seated meeting runtime: decisions, formal buy-in, clean action items

Remote default collapsed many interactions into 1 posture. Seated, face-forward, often camera-on, even when the topic is still fuzzy. More observation pressure means less exploratory talk. What disappears is the cheap pre-thinking layer, the place where you could be wrong out loud without making it a whole event.

What remote work replaced it with

The instant on call

The missing buffer is not just comfort. It changes attention and recovery. A normal day becomes 3 beats:

1) you finish a hard doc

2) you click into a call without moving

3) you click out and are “back to work” instantly

Leroy’s attention residue idea helps explain why the previous context stays sticky when you switch fast, especially if nothing felt finished. Rogelberg et al. (2023) points to meeting respite as a real variable for fatigue and productivity, not a luxury. That’s why you can finish the day feeling busy but oddly unfinished.

Performance load and why standing is not the fix

The naive fix “just stand” often fails because it keeps the same evaluation frame. Video and even audio can trigger objective self-awareness (Duval and Wicklund) — you start watching yourself perform instead of thinking out loud. Attention shifts toward monitoring the self. That monitoring cost makes language careful instead of exploratory.

Standing can shorten meetings (Leach and Rogelberg), but quality is not clearly improved. Sit–stand setups reduce sitting with mixed downstream outcomes (Shrestha et al.). Courtroom mode stays courtroom mode, just with tired calves.

At this point it helps to state the bias behind the rest of the piece. I am french, born in 1974, trained in fundamental physics, and i look for missing environment variables before blaming motivation. I have spent most of my adult life at a desk, from Beijing to Berlin to Lisbon, often past midnight. I am not in pain, but i know the early signal. Quiet upper-back tightness that builds until it forces me to move.

The invisible taxes on remote days

Decision quality tax

Meetings go in circles because early-stage framing needs options and rough thinking more than precision. The symptom is familiar. You leave a 45-minute call with “we need to align” and no shared written problem statement, so the same topic returns next week under a new invite.

Shared mental models are a coordination variable, not a nice-to-have. Nothing is broken, the container is.

Walking helps most at the fuzzy front-end. Oppezzo and Schwartz (2014) found walking improves divergent thinking during and shortly after walking. It does not guarantee smarter decisions. It just widens the option set. Practically, that can prevent premature precision, where a group argues about details that do not matter yet.

Common loop symptoms

The same arguments repeat with new words because options were never listed
People debate edge cases before agreeing on the core problem
The “decision” is vague consensus with no timestamped decision log
Action items exist but owners and dates stay optional
Someone ends with “let’s schedule another one” as the main output

Conflict tax and repair tax

Video and audio delay mess with turn-taking, so people interrupt by accident or wait too long, then over-correct. Tone gets brittle (Sellen, 1995; Boland et al., 2022). Face-forward framing makes everything feel like a debate instead of a side-by-side problem solve. Email is worse for tone illusions (Kruger et al., 2005), but remote calls can still make a normal sentence land sharper than intended.

Even when a conversation ends fine, remote often leaves residue because there is no decompression path. With video, self-view keeps a monitoring load running (Bailenson, 2021). Audio-only can reduce the “being watched” feeling and sometimes supports disclosure (Joinson, 2001), but it is not magic. It works best when the last 60 seconds are an explicit summary so nobody leaves with ambiguous tone.

Remote ends with a click, so the meeting keeps running in the background. Leroy (2009) describes attention residue from unfinished work. A common cue is realizing your jaw is braced or your shoulders are still up 20 minutes later.

Walk-compatible meeting design

Walking is for clarity and tone, not steps. Some meetings should stay seated and screen-based.

When walking helps

It tends to work best for

1:1 check-ins
Unblockers
Career chats where less gaze pressure can help (Joinson, 2001)
Brainstorming to widen options (Oppezzo & Schwartz, 2014)
Debriefs to close the loop while it is warm
Pre-mortems to name risks before the doc looks “final”

When sitting is non-negotiable

Stay at the desk for screen-share reviews, live editing, numbers validation, and anything with 4+ people and facilitation. Latency already makes turn-taking fragile (Boland et al., 2022).

The good enough switch rule

Start walking to frame, then sit to confirm and log.

Step 1, 5 to 15 min walking: agree on a 1-sentence problem statement and list 2 to 4 options
Step 2, 5 to 15 min desk mode: validate constraints, check the artifact, write the decision
Step 3, 60 seconds: timestamp a decision log and owners

3 formats that survive meeting-heavy weeks

Format 1, audio-first 1:1 default

Script: “let’s do this audio-first, camera optional, unless we need to screen-share.”

Camera optional and hide self-view when possible (Bailenson, 2021)
Signal attention with short verbal check-ins, not forced visibility

Format 2, 10-minute walk then 10-minute desk

Walk to frame the problem and generate options
Desk to confirm constraints, pick 1 option, log it
Output artifact: decision log entry with a timestamp, plus 3 action items with owners and dates

Format 3, the 2 to 4 minute debrief lap

A short after-action review helps when it is structured (Tannenbaum & Cerasoli, 2013).

What was supposed to happen
What actually happened and why
What we change next time, 1 thing

Keep it boring. This is not forced emotional processing.

Adoption fails on the exact days it is needed, the meeting-wall days. Implementation intentions help because the calendar becomes the cue (Gollwitzer & Sheeran, 2006). A low-drama approach

Change 2 recurring meetings, not 12
Put “walk-first” in the title
Keep an “if screen-share then switch to desk” rule

Constraints that make this credible

Privacy and compliance without paranoia

Overhearing risk is mostly an intelligibility problem. Quiet routes can be worse than moderately noisy ones because silence makes words carry.

Boring safeguards that still work at 17:40

avoid reflective corridors and atriums
pick steady masking noise, not near-silence
no speakerphone, use a headset
keep the mic close and your voice lower
minimize identifiers until you are back at the desk

Some hacks backfire. Masking audio in your own earbuds can trigger the Lombard effect (when you hear noise, you unconsciously raise your voice) and make you talk louder. Directional microphones also make distance feel fake. Choose location and equipment deliberately.

Notes, optics, and accessibility

People avoid walking calls if it looks like multitasking. Note capture can stay tiny.

10-second voice memo after the call
or 1 “Decision next” line typed when back at the desk

Remove social ambiguity with one dry line. “I’m going audio-only and walking so i can focus, i’ll switch to desk if we need screenshare.”

Walking-friendly design must not become a new participation tax. Equivalent options

seated outdoor meeting on a bench
an indoor loop at slow pace, off-camera
staying seated but changing posture with short breaks
1 person walking, the other not, with no penalty

A tiny scoreboard and a 1-week test

Measure what you actually want

Keep the scoreboard tied to real pain, not fitness theater.

Replay loops after a call, how often your brain replays the same 30 seconds
Time-to-clarity, time until there is a written problem statement or decision
End-of-day bracing, quick note on jaw clench or shoulder lift

Teams can track similar outcomes using artifacts they already create.

decision log timestamp vs meeting end
time-to-written problem statement in the shared doc
churn, how many follow-up meetings spawn for the same topic

Metrics can become performance theatre too. If measurement increases self-monitoring, it is not neutral anymore.

A 1-week test that fits meeting walls

If it is 1 of your 2 recurring 1:1s, make it audio-first and walk-compatible for 1 week
If screenshare is needed, switch back to desk mode without guilt
If the topic is sensitive, keep camera optional and do explicit summaries
If you skip 1 day, continue next meeting, no restart

Define pass-fail in outcomes.

Pass signals

fewer circular loops
lower residue after the call (Leroy, 2009)
clearer next steps with owners and dates same day

Fail signals

residue unchanged, maybe the meeting needed a different container or tighter closure artifact
privacy or environment makes it awkward, so walking is not the right default

Stop walking when crossing streets. Avoid complex multitasking. If the streets are chaotic, do an indoor loop or just pace off-camera. Walking meetings should be safe and slightly boring. If it pays rent, keep it. If not, drop it.

Remote work did not just change where you sit. It changed how thinking gets warmed up. When every fuzzy topic becomes an instant-on call, people get careful, the options shrink, and the meeting keeps running in your head after the click.

The fix here is not a lifestyle overhaul. It is rebuilding a cheap pre-thinking layer with small, testable tweaks. Use walking talk for framing, tone, early risk surfacing, and 1:1 unblocking. Then switch to desk mode when the work needs screens, numbers, or a real decision log. Keep it “good enough” with 2 recurring meetings, camera optional, and tiny note capture. Track outcomes you actually feel, like fewer replay loops and less end-of-day bracing.

Most teams do not need more meetings. They need a cheaper pre-thinking layer, and a cleaner stop signal at the end.