There was a dev I worked with for years, a few hours ahead of me, never met face to face, who would not say a word in any meeting we shared.
Sprint planning, design discussions, retros. The PM would speak, the leads would speak, QA would speak, EM would speak. I would speak. The dev would listen, take their notes, and stay quiet. Not because they had nothing to say. While everyone else was shaping the discussion, they were a few steps ahead, working it out in their head.
And yet the implementation was theirs. The lead trusted them, the EM trusted them, and I did too, before we had any reason to give it in the room. The reason was on GitHub.
Their PRs were the conversation we did not have out loud. Some first cuts handled edge cases before I named them. Other times my review made them think harder, or their pushback made me rewrite mine. A few rounds in, the diff would settle. Reviewing their code felt like reading someone who had already read mine.
I would approve, merge, sometimes leave a question. They would respond inline, change the diff, sometimes push back. I reviewed theirs. They reviewed mine. A Slack thread now and then, mostly the diff itself.
One round, I argued for a refactor I thought the contract wanted. Their two-line counter pointed at a caller I had not traced; the refactor would have broken it. The next diff was the answer. Another time, their first cut already handled an edge I would have asked about in review – a null path on a slow dependency, written before I noticed it was missing. Either way, the patch already had what the meeting was for.
Their PRs showed both sides at once. You can see how someone reads in the comments they leave and the questions they ask. Their first cuts show how they think. Most channels at work show one or the other. A meeting shows thinking, live, but you cannot tell from a five-minute opinion how the speaker would have received yours. A PR exchange shows both, iterating against each other, over months.
This is the part remote work changes.
The meeting rewards the fastest mouth in the room. The PR rewards the clearest mind. In an office, those two things blur – you see someone in both modes through the day, and the impression they leave is some average of the two. At a distance, with code as the medium, they separate. The dev was not choosing the PR over the meeting because they could not speak in the room. They were choosing the channel that measured what they were good at – careful reading, careful response, the next diff better than the last. The meeting was measuring something else. Sometimes that lined up. Often it did not.
I would nudge them in meetings sometimes, ask for their opinion to bring it into the room. Few words, mostly in the affirmative. The actual answer would land in the next PR.
This was the first thing remote made obvious. It was not the only one.
The room never heard them. The codebase did.
I have not worked with that dev in years. I still know how they think.