> unless we're simply talking about people who are "productive"

That is literally what the grandparent poster said, and matches the first two definitions I can find on Google.

Personally, I often think of this example: https://news.ycombinator.com/item?id=3033446

It's really interesting to see how both sides approach the problem (without rushing it), but only one of them solves it. That certainly counts as a 10x difference?

> "10x engineer" is a marketing gimmick used to gain leverage in various ways, IMO.

I've been following the discussion, and I think that "10x is a lie" has itself become a dangerous meme. It's already hard to tell some of my clients that developer quality matters, that hiring cheap developers en masse is pointless if they can't reliably fizzbuzz. Now Developer Twitter decides to proudly proclaim that everyone is equally capable as a developer, except some of them are toxic jerks. Sigh.

I don't believe that the majority of comments on Twitter are trying to say that every developer is as as equal as every other. Many of them are arguing that the "10x developer" idea is unrealistic in the sense that management wants to hire someone 10x more productive than everyone else (this story agrees, 10x is a big exaggeration). Some people have worked with so-called 10x developers, who have been slavishly defended from criticism by management, who are churning out piles of code that will be difficult to maintain, it seems that some of those people also have some toxic behaviors.

In terms of the question "do 10x developers exist", I wonder if it really matters. An argument has been made that there are people far better at chess or football or piano than everyone else, so there must be developers who are vastly better than the average in the field. I'm not sure that software development is really the same kind of thing... But even so, we do not have the systems in place that would allow us to recognize such people. All we have, for the most part, are anecdotes from recruiters and managers and many of these stories sound self-serving and exploitative.

It could be the 10x developer exists in the same sense that the 10x CEO exists: every so often one appears and seems to meet the standard. But the closer we look, the more circumstance and the people they surrounded themselves with seemed to be even more important. Often it looks like maybe the particular period of time was a factor as well. And the argument goes on and on, so on and so forth.

That's a more generous reading of the mood on Twitter. I also think that many people have good intentions, and are only trying to reduce the general levels of impostor syndrome.

> Some people have worked with so-called 10x developers, who have been slavishly defended from criticism by management, who are churning out piles of code that will be difficult to maintain, ...

And I have experienced this anti-pattern several times. But I worry that we're throwing out the child with the toxic genius bathwater.

> 10x is a big exaggeration

The actual actual factor is a function of your work environment. If HR hired "nice" people who failed all the trivial interview questions, would that number still seem unbelievably high? That's what some small shops do when developers are hard to find :|

Fair enough, once there's more than a handful of replies a Twitter thread turns into a trash fire. The replies and posts that I have read, insofar as I had patience to read them, struck me this way.

I don't think any organization has the ability to hire exclusively 10x developers, even if they did exist. Large places with deep pockets can afford to hire many people and then promote those they perceive as meeting their own idea of what a 10x developer might be.

On average, I think everyone should try to hire the best people that they can. Companies should try to retain their best people and encourage everyone to improve. Over time, the productive people will emerge and should be valued by the organization. Are they going to be the mythical 10x developers? Maybe, maybe not.

This idea that you can hire a 10x developer to guarantee the success of your project is, in my opinion, a waste of time and money. If you think it's working then I bet technical debt is piling up and it will catch up with the organization eventually.

> But I have never seen one that can do that _and_ produce quality.

I've never seen a mediocre dev produce quality so it isn't needed in order to be 10 times as productive as a mediocre dev.

What I think usually happens is that the 10x dev produces 10x as much technical debt as everyone else, so the team thinks that this guy is blocking all of them when it would have been just as bad if they put 10 mediocre people to do it instead, just that then the blame would have gotten spread out among 10 persons.

Or he produces 10x the quality and 1x the code of everyone else, but his mediocre colleagues doesn't notice that his code is better since everyone else added so much crap so he is seen as just a decent 1x dev.

My first team lead was a 10x developer; I still remember searching for the j2ME code that would have taken us two weeks to complete that he wrote overnight -- I remember how dismayed I was to find it all directly in the keyboard event method instead of being spread over to various places it should have been.

So his initial code was a "10x" achievement, accomplished in literally 1/10 (or less) of the time it would have taken the team.

....but the legacy of his "10x achievement" was an increased maintenance burden. So the overall gain was less than 10x. Perhaps it wasn't even a net gain in the long run.

This is a good example of my main beef with the whole "10x" thing. So many superstar developers are lauded for banging things out quickly, while everybody else is stuck cleaning up 10x the mess.

I truly believe that there are 10x developers, but identifying them is nearly impossible unless the person doing the identification is (a) a pretty savvy developer themselves (b) is intimately familiar with the choices made by the "10x" dev and their long-term ramifications.

I don't know about 10x, but the most productive devs I meet are usually the TDD kind, because they don't create as much technical debt, they are faster.