The "there's no such thing as a 10x engineer" crowd forgets or dismisses that there's a distribution of engineering talent and capability.
You can even find examples. Fabrice Bellard immediately comes to mind.
You can even find examples. Fabrice Bellard immediately comes to mind.
It's not that the 10x engineer is a myth - it's a fallacy. You'll find that people in the same profession (any profession) will have variance in their abilities and their return to a project, but it's a big leap to assume that someone that appears to perform much better than their peers is actually doing so in any objective sense.
The 10x engineer is more likely than not the person most familiar with the project and the codebase, knows the real reason why Mark didn't just put everything on Kubernetes or why Tanya didn't want to switch from React to Vue before leaving the team.
Or.. it's an engineer that has the ear and the faith of higher-ups in the organisation to deliver results because they're good at convincingly taking credit for them.
Generally I find that 10x engineer on here means "knows more than most about [tech stack X]" and can work quickly. That's cool if all other things are equal, but I've had to clean up more than a couple of catastrophic messes left in the wake of people that worked quickly.
Just to pre-empt someone telling me I must never have met a real 10x engineer or some other such dismissive response - I'm an average fish in an average pond. I've had people tell me how good I am in the past, both reports and managers, and have in both cases corrected them politely that I've just had more time with the problem/project/codebase than others have.
Anytime I hear someone talk about 10X engineers, let alone describing themselves as that, all it makes me think is that there's a failure of humility.
None of this belittles Fabrice Bellard. Or John Carmack. Or Chris Lattner. Or [known and respected programmer]. Their achievements speak for themselves.
You could give names of people in the field all day. There's a lot of us around. How many of the self-described or colleague-described "10X engineers" floating around are Linus Torvalds?
Their solutions are elegant, performant and survive design changes. Their code is easier to maintain. Paying 2x the average to these people is a steal,as they are a LOT better than that. Not sure how much, but I'd rather have one of them on my team than 3 average engineers (probably more).
I'm not saying average engineers aren't useful workers. They are, they get things done if they aren't incompetent. But they are nowhere near a 10x engineer.
10x engineers aren't perfect. But they are amazing nonetheless.
If you have never worked with one, maybe it's because you had bad luck, haven't been around long enough or haven't worked in great teams. Whatever the reason,once you meet a 10x, you will know it.
That should save us some time.
> Whatever the reason,once you meet a 10x, you will know it.
Guess not.
* You've never met a 10x engineer, and you can reason clearly and cogently as to why they don't exist, so you conclude that they don't exist;
* Someone ... indeed, several someones ... tell you they they have met, worked with, and/or managed people who really are 10x engineers;
* You say that these people must be wrong.
This feels a lot like the biologists who wrote - convincingly - that kangaroos can't exist. They'd never seen one, and gave completely water-tight arguments that the method of locomotion was inherently impossible.
This computer-aided leap in output actually does separate software engineering from every other discipline except things like pure math.
The use of specialist tools to increase output happens in hundreds of other careers and I’m sure it just falls under learned knowledge of the craft.
But I'm not sure there are any other disciplines where a single person does it regularly.
I think the difference is we're not dealing with anything physical. A structural engineer never has the leeway to invent new ways to build bridges while under deadline. Maybe some ingenuity and process improvement along the way, but nothing like 10x on any single project.
Nope.
To whatever extent they actually exist, the 10x engineer is the one who starts without a codebase, and creates one that is elegant, flexible, extensible and performant, avoiding mistakes and working faster than most of their peers would. They build a platform for their successors to work with that enables more than it impedes.
If you've never seen this, then fine, but that's most likely because you've not been around during the prehistory phase of a given codebase.
"There's no such thing as a 10x engineer" people probably haven't seen a normal distribution before.
And yet there aren't 17 meters tall people (avg adult male height back home is 1.7m)
Mind you, I consider the whole discussion stupid since it's an intellectual/creative job
Every. Single. Instance. Of trying to measure such things has failed at producing consistent results or results with predictive capability.
Talking about X times engineer makes no sense when you don't even know what a 1x engineer is.
Personal anecdote: I've worked with 10x, 1x, 0x, -1x and -10x engineers, and everything in between.
I think the debate is over the properties of that distribution, and whether it needs to be plotted on a logarithmic scale