As for how you attract them - same way you attract anyone else. Give them something they want. For these people it’s agency, freedom and high level guideposts. For a productive big corp employee it’s clear guidance, timelines and a hierarchy to adhere to.
You have to become a superstar first.
I had the good luck of living down the street from the AirBedAndBreakfast founders in San Francisco in 2008 and we'd commute together to our Tuesday night YC dinners in Mountainview.
I picked up early that Nate was already a superstar coder, Joe was a superstar designer, and Brian was a superstar designer and salesman.
I knew I was not, and had to double down and practice, practice, practice.
Craftsmanship comes first. It took those guys decades of practice at their crafts before they founded AirBedAndBreakfast.
Think of Paul Graham. Before he created HackerNews and YC, he had already written and published a book on Lisp! A master craftsman.
Craftsmanship comes first.
If you are not a master at your craft, don't even waste your time trying to recruit superstars. Instead, spend your time on practice.
I’m starting to think paying attention to what is happening is undervalued.
Heavily agree. But paying attention to what is happening _in nature_.
Pay little attention to what is happening in symbolia.
I got rid of my cell phone 2.5 years ago after my daughter kept saying "no phone dadda". Forces me to pay attention to nature and real world every moment I'm not at my computer. Huge life improvement in every way.
many superstars aren't even good at doing the expected when it would be a good idea, often because of drug addictions
this happens for two reasons. one is that if you're selecting people along two axes, the more harshly you select on one axis, the less candidates remain to select along the other, unless the axes are perfectly correlated. the other is that, whether you're a candidate selecting strategies or a judge selecting candidates, two axes along which you can select are mean and variance. in any event where you take the best of multiple trials, the top performers will almost always have high variance, not just a high mean
The ability to pay attention is where the unexpected comes from in my judgement. The whole world looks at something and tunes out early because they think they understand. An individual that is truly paying attention notices all the subtleties that that can be used for a fresh solution.
My point is that an inconsistent superstar is not as good as an individual with a consistently fresh solution.
Going out one more layer I thought the topic was how to have a stable organization with consistently fresh thinking. I responded to a topic discussing how to attract superstars. I believe anyone has the capability of being a superstar in the right environment.
And it's also easy to speculate in a forum about super stars and "10x coders" and what team you should have.
But in reality, you have to play with the cards you have been dealt with. You can search for better folks, but until you find them, the show must go on.
In many card games, being dealt 8 Aces means losing the game, and the same is true for startup teams with all-alpha animal "leaders" who are all super smart but cannot agree on anything.
Realistically tell yourself that 10x developers are a silly myth and the people you were able to hire for pennies from the local demographic are the superstars, and try not to hang your business model on them actually being the best in their field. As long as you don't fact check that against reality it'll be OK.
Thankfully almost no one is the best in their field :) You probably can't afford the best, you probably don't need the best, you probably can't keep the best happy, you definitely don't have problems that need the best.
That being said if you pay dollars instead of pennies you can find some decent folks and maybe save some money in the long run :)
You don't need the single best, you just should aspire as high a concentration of n * x for n>1 (2x, 10x, ...) as your talent to raiase funding permits.
But don't forget it's not all about "best coding"; you want people best in different things (PM, UX, architecture, security, databases, algorithms, compilers, documentation, networking etc. - also depending on what the product is about) - a diverse team of rock stars.
But companies should be honest with themselves about what they need. Everyone still seems obsessed with being able to pretend they somehow got the absolute best deal on the employment market instead of interviewing for the position itself.
/s
