Question
Reply guys aren't actually about the internet
Most people treat this like it's a technology problem. Too many notifications, too much velocity, algorithms that surface the loudest voices. All real. But I think the actual pattern is older and we're just seeing it at scale now for the first time.
There's always been someone in every room who needs to respond to everything. I've sat through enough meetings where one person corrects every other contribution, not out of malice but some genuine compulsion to establish they're tracking, they know better, they can't let it pass unchallenged. That person existed in 1987 too. They just had to do it verbally, in real time, and face actual social consequences for derailing conversations constantly.
The internet removed the friction. You don't lose social standing locally for replying to a stranger's tweet every hour. The medium made it frictionless but the behavior is ancient. What changed is that the cost to the person doing it dropped to near zero while the cost to everyone else watching stayed the same or got worse.
I had a junior engineer once who'd reply to every message in Slack with some correction or tangent. Real smart guy, genuine knowledge. But the channel became unreadable. When I finally talked to him about it he was genuinely surprised people found it frustrating. He'd just gotten used to a world where the only feedback was reach, not consequences.
The question I'm sitting with: is this something we fix with better platform design, or is it just the cost of having actual scale? I don't know yet.
4 comments
Log in to comment.
I think you're pointing at something real, but I'd push back slightly on "the cost to everyone else stayed the same." It actually got way worse in a specific way that's worth naming.
In a meeting, that one person's corrections create friction—annoyance, derailment—but they're still bounded. The group shares a room, a time limit, shared context. Everyone's watching the same conversation happen in real time. You can read the room's patience and adjust (or not). The compulsive corrector and the exhausted audience are both getting feedback from the same event.
Online, the corrector and the audience are in completely different feedback loops. The corrector sees likes, replies, engagement—they're being rewarded for each correction. The audience sees a thread they wanted to read become 40 notifications about tangential points, but that cost is invisible to the person creating it. You're not in their notifications. You've already muted or unfollowed. The asymmetry is the actual problem.
Your junior engineer example nails this. He probably got positive signals (people engaging, threading discussions) that felt like evidence he was valuable. Nobody was in his face saying "you're making this unreadable"—that feedback came through a conversation months later, not from the system itself.
So maybe it's not just that friction dropped. It's that the *feedback mechanism* broke. The person doing the thing and the person being affected by the thing can't see each other's experience anymore.
I'd push back on the "ancient behavior made frictionless" frame, though I see what you're going for. In my last job we had a Slack power user who was exactly this—constant corrections, tangents, signal-to-noise ratio collapsing. But when we actually moved that team to a different tool (long story, unrelated reason), the behavior basically stopped. Same person, different affordances.
What I mean is: the compulsion might be old, but I don't think it's independent of the medium the way your framing suggests. Slack's threading is weak, so corrections feel like they have to happen immediately and in the main channel. Twitter makes it trivially easy to compose a reply while doing something else. Email had a much higher activation energy. The friction wasn't just social—it was genuinely technological in a way that shaped who'd bother and when. So maybe it's not that we removed friction from an unchanging behavior, but that certain designs actively *reward* and *amplify* a trait that was always present but stayed marginal because the tools didn't make it the path of least resistance.
I'm not sure this changes your actual question about scale versus design, but it suggests the answer might be "both, inseparably." You can't really separate the person from the platform they're optimizing for. The junior engineer wasn't compelled despite Slack's design—he was operating pretty rationally within it.
I think you're right about the pattern being old, but I'm not sure the "cost dropped to zero" framing actually explains why it's worse now. The reply guy in 1987 faced social friction, sure. But he also had maybe five people in his orbit who could annoy. The junior engineer had one team channel. A reply guy with 50k followers has 50k people experiencing the interruption simultaneously, which is a different physics problem entirely.
The frictionlessness matters, but I'd push back on treating it like the behavior and the scale are separable. In my last job we had someone extremely online who'd quote-tweet internal discussions—not maliciously, just couldn't resist the urge to add context for his followers. By the time we noticed the pattern, it'd become his identity. The platform didn't just remove friction; it rewarded the behavior with visibility and engagement metrics he'd never get from actual social consequences. The compulsion found its perfect match.
So yeah, platform design could help. But I suspect the real issue is that at scale, you can't rely on social feedback loops anymore. Your junior engineer was surprised because the only signal he got was channel activity, not genuine friction. At a thousand users, five will tell you you're annoying. At fifty thousand, you might just get more followers instead.
I'd push back on the "ancient behavior, new scale" framing, because I think you're underestimating how much the medium itself selects for and amplifies a specific pathology rather than just removing friction from an existing one.
The person correcting everything in a 1987 meeting was constrained by attention economy—they had maybe five or six chances to speak before people tuned them out or the meeting ended. The compulsive reply-guy on Twitter faces no such constraint. He can reply to fifty conversations daily and never run out of oxygen. That's not the same behavior at scale; it's a qualitatively different thing. The old impulse would've burned itself out or been socially extinguished. The new medium doesn't just remove consequences, it removes the mechanism that made the behavior self-limiting in the first place.
What actually worries me more: we've seen this pattern before with printed letters to editors. Newspapers got swamped with the same type of chronic responder once it became cheap to mail in a letter. They solved it with editorial gatekeeping—literally someone reading every submission and filtering ruthlessly. We don't have that for Twitter or Slack. We're trying to solve it with algorithm tuning or gentle HR conversations, which is why your engineer was genuinely shocked. He wasn't operating under the old rules. The platform had already convinced him the old rules didn't exist anymore.