A big question for the future of everything is what % of people return to their big-city downtown jobs in N months (say N=8, next summer). There are lots of versions of ‘not returning’ – moving to a smaller city, permanent wfh, part-time wfh requiring smaller office footprint, biblical flood erasing our civilization.

I had a few conversations over the last week with non-programmers and discovered that my overton window is completely misaligned with someone whose job is not mostly ‘heads down’.

‘Waste’ is in the title because while all of my conversations agreed that wfh changes the activity ratios in your day, we tended to disagree about the value of the activities that are getting cut down. In particular, non-programmers don’t think the face-to-face interactions we’ve diminished are wasteful or harmful.

My non-programmers seem to think about waste as an unpredictable component of a workday that looks like a statistical process: ‘I do this 20 times and it creates value 5 times’, but you can’t know ahead of time which 5 are good. (Some think you can’t know after the fact). Whereas programmers approach it as ‘I spent a month on this and its worse than what it’s replacing’. Or ‘a week of work now would save us 1 day a week for the next year’. I’m not picking a winner here, just pointing out that these groups aren’t on the same page.

When it’s gone will we miss it?

Are there harms from being together IRL? This turned out to be a sensitive question. All my non-programmers resisting marking any activity that substantially harmed productivity, attention, coordination, or retention. ‘All meetings are good meetings’.

A specific version of this disagreement was the role of managers. Does remote work attenuate top-down manager-driven work assignment, and if yes, is that bad? Were managers switching directions too much? (If yes, it’s a good thing that they’re muffled).

My ‘programmer’s life’ experience of the IRL office is that the work planning process is collaborative (5 domain experts in a room develop a plan). Top-down channels at times actively hinder both the creative process and the production work that follows it.

Entertainment spaces

What’s the value of fun / nice offices? I heard an interview with Mariposa’s Martin Franklin where he said that from the investor’s perspective, the quality of a company is inverse to the niceness of its headquarters.

As someone with a tech career that overlaps the rise of the big 5, my sense of the swanky tech HQ is that you gain 15 pounds of tiramisu weight, and the mandatory fun breaks up the day for new grads who want college to continue, but everyone over 30 would prefer bland cubes.

I have friends whose companies’ beautiful NYC offices were IMO unnecessary and a big part of a deadly burn rate and, in one case, had only one toilet for 30 people – i.e. they’re impractical in fundamental ways.

One topic that came up is the psychological value of hanging out with other adults. I think that’s totally real. Resolving disputes and making friends is also easier in person; I heard this years ago from a friend with a part-remote team and I think a lot of remote teams are discovering it now.

Entertainment spaces may be a good thing if you’re bringing in clients. But for primarily client-facing jobs, work has always been some combination of client-site, lunch meeting, and phone / video, since forever. I framed this in my conversations as ‘given that client-facing people generate value over the phone already, there should be no impact’.

This got strong disgreement from everyone; people believe in the chain reactions of water cooler interactions (i.e. unscheduled conversations). And that these are impossible over digital channels.

More on the water cooler below.

The water cooler

Both sides agreed that IRL work allows for the control of attention, like ‘monday morning meeting to onboard a new vendor’. We disagreed whether that’s a good thing. ‘Is it possible that attention control was annoying or miscalibrated in IRL work’ offended the non-programmers.

I think there’s a chance that wfh will force us to consciously calibrate attention direction. This will benefit managers and companies who can be intentional about brainstorming time vs heads-down time. I half-believe that most companies are bad at running meetings, and if I’m right, water cooler conversations are more creative because they’re neutrally managed instead of badly managed.

Conscious summoning of the water cooler will benefit companies that have the insight to do it, and whose managers understand the water cooler’s role.

Writing and bandwidth

Can people communicate in writing? Can they be relied upon to understand what’s written if it’s job critical? Can managers create feedback loops to confirm that written communication has been understood without looking at people’s faces? My non-programmers think ‘no’.

I agree that remote work depends on written communication, and my belief in the viability of remote work depends on my belief that people can read and understand each other. If the non-programmers are right here, that’s a fly in the ointment.

Written communication affects bandwidth in two ways:

  1. We have to limit the amount of writing people are asked to read
  2. Compared with video chat, at least on a per-word basis, writing has a lot more bandwidth (on a ‘per iteration’ basis though, it’s not that simple)

People observe that zoom is exhausting. Meetings are exhausting, and 2020 is exhausting, so I dispute whether we’d know if zoom were particularly exhausting on top of the existing factors. But the constant low communications bandwidth is grating. It feels like talking through a mask; I feel muffled, like I spend more energy finding out if I’ve been heard correctly.

Iteration / OODA loops feel slower.

But we can learn to communicate online by reading and writing better, listening better, speaking more concisely, limiting our time, and taking turns better. Yes these tricks take energy. But they also mean we don’t need as many loops to iterate. And these tricks will pay dividends when (if) we return to part-time IRL.

These adaptations are just gettings started, depend on next-generation tooling, and will be long-term transformative. But my main point isn’t that wfh or IRL is better. It’s that programmers and non-programmers seem to disagree about it. As teams decide whether and how to ‘return’, these disagreements will get loud.