[015a]

> I can’t imagine staying home all day everyday

I can imagine it, because I've been doing it every day for the past year. It sucks. It sucks a lot.

I can't even count the number of people in my tech circle who went from (pre-covid) "ugh, so-and-so works from home at an all-remote company, that sounds fantastic" to (early-covid) "wow, working from home is fantastic, I can start a bit later, no boring commute, this is amazing" to (late-covid) "holy crap fuck working from home, i never want to do this again, I'm over it, get me back into an office".

COVID itself had something to do with this, in the sense that never leaving your home, for work or play, is exhausting, and makes work-from-home worse.

But, I don't think most people will disentangle that from the reality of how they feel. I think many, many people want to get back to the normalcy of a commute and office life. And, a few years from now, we'll all be tired of it again. And that's fine.

This whole "remote-first" wave we're seeing is almost entirely isolated to tech jobs in tier-1 tech cities (SF, Seattle, NY). That's definitely a situation where, many many people wanted to get out of the city pre-COVID, due to rising costs, and COVID-fueled remote work became the excuse they needed. Then, companies corroborated by going all-remote because corporate real estate is expensive, and if its what employees (and especially c-suite) want, lets do it.

I fundamentally believe that its not the wide-spread movement that's so easy to conclude it to be. First, because many people want to go back to an office, especially outside of tier-1 metro areas. Second, because going remote is fucking hard on your business, and executives generally aren't dumb, they know its hard. Collaborating asynchronously? No one can do it. Human beings don't have the patience. Scheduling a 60 minute meeting because a topic isn't resolved in ten minutes of Slack chat? So fun, much asynchronous. Documenting everything? Ha. Nope. We can hire anyone! I mean, in our timezone, maybe +/-1, because timezones and latency are a fundamental, physical, law of nature bitch and a half. Sorry, my microphone was muted, and my dog is barking (and if you think we're going to "solve" online video chat in our lifetimes, go buy a printer and realize we haven't even solved that).

Also, to some degree I hold the controversial opinion that many of these "pioneering" remote-first companies (Gitlab, Coinbase, Spotify (to a lesser degree), etc) are generally involved in product domains which require, lets say, less creativity. I love their products; they're absolutely fantastic, and I don't mean that disparagingly in the slightest. I simply mean that their biggest problems are generally operational in nature, or have relatively well-defined edges. If you're at a company where you can't even imagine what your product/project will look like in 12 months, where you need zero latency high creative output from your team, remote is even harder.

I have a friend who works on Azure, and she said they were seriously considering full remote. Its like, duh, that'll work great for Azure, because they're highly operational and a ton of their product development is just "what did AWS do last year". It works for some companies, and it definitely won't work for others, but maybe they need to go through the pain to realize it.

[sverhagen]

I agree that a lot of people will want to go back to 60% or 80% in the office, once again they can. But I feel you're doing a lot of looking into other people's business and being dismissive of their complexities. Or, at least, it sounds like you're using "highly operational" as a proxy for that. Yes, we're not all in the top tier of, whatever is there, maybe rocket science. But there is a very large middle tier of still very complex businesses, which may look simple from the outside, but often only deceivingly so.

[aboodman]

I have to know what you do for a living that building and running one of the largest clouds in the world is "highly operational".

[numair]

I would say that building and running a cloud — multi-tenant outsourced digital infrastructure — seems to be the very definition of “highly operational.” Especially when you’re working at Microsoft, which continues its unregulated tradition of cloning and/or embracing-and-extending its competitors.

[aboodman]

I don't think there's any way to build something like CosmosDB - or even heck, s3 - at scale, without an extremely high degree of creativity on the team.

[015a]

A useful proxy for organizational creativity oftentimes comes down to "how well-defined are the edges".

Every human who interacts with (externally) and works on (internally) S3 has an intrinsic understanding of what S3 is. It has well-defined edges.

Higher definition of edges within a product domain naturally correlates with a decrease in necessary creativity and an increase in operational expertise. Its more important that someone who works With and On S3 understand what S3 is, rather than what it could be. That's just the nature of the beast.

That isn't to say that it requires zero creativity; that isn't remotely what I'm saying. There is creativity involved in operational expertise, and beyond even that, S3 gets awesome new features all the time which requires creativity in thinking about how to solve customers' problems in a generic, customer-agnostic way, how to implement those solutions at scale, etc.

Though, even developing out these features are more operational, because you have the defined edges to anchor yourself against. You have customers who are reporting their problems to you. You have a profitable operational model which guides investment in the product.