I started programming at 8. I gave up on it at 16. Picked it up again in my mid 20's. Responsibilities are an issue, (and this is especially an issue for women, who, if they want to have a family can't wait as long as men can), but it isn't the only one.

One thing that I think is pretty clear is that as we age, our thought processes change. This is something we don't like to talk bout because because it brings up spectres like age discrimination, but it is true.

The thing is, this isn't a neutral tradeoff. While purely abstract manipulation of data may be easier in one's twenties, solving real-world problems is much easier at the age of 40, not only because we understand the problems we are solving better but also because we understand problem solving better, and tend to be better at it.

Can someone with a liberal arts degree pick up coding on her spare time and make real, significant contributions to the world by so doing? Absolutely. Will she be interchangeable with someone who has been coding since age of 6? Probably not. But interchangeability is overrated and the ability to find our own problems to solve is not.

This is a great comment and has changed my perspective somewhat. Thank you.

BTW, I am pretty sure my grandmother didn't pick up coding until at least in her 30's. Could she have written a device driver? Probably not. Did she write nuclear physics simulations in Fortran? Yep.

Well, maybe I am an outlier, but I didn't start programming until I was 20, and didn't seriously get into it until I was 24 (i.e. actually shipping something).

And I sure as hell don't think Zuck is "better" at tech or being a hacker than I am. Luckier, perhaps, but not better. I guess we'll see.

Also, I was married at 19, and a year later was a father. By 24, I had two children with a third shortly thereafter. So I definitely had the whole "lots of responsibility" thing dynamic, and while I do agree it slowed me down relative to what I could have done if I was single, it wasn't fatal and didn't stop me from learning, or becoming good at what I'm doing.

But again, I'm probably an outlier.

Since we're on the topic of unpopular-but-true, I'll go ahead and post another: It's not about whether we can program so much as what we can accomplish by learning to program while we're young. You and I may make fine and capable employees, but some want more than that. It takes more than just ambition: it takes positioning and luck. I hope it won't be too controversial to say that younger people can take more risks due to the fact that they're young. Therefore they will take advantage of the lion's share of the luck.

In other words, if we were willing to sacrifice our dependents to take more shots at the type of success which attracts people to YC, then we'd be in the same position as someone who was young. But we aren't willing, so we're not in that position.

You've moved the goalposts. Neither ancestor comments nor the article were talking about founding a startup. But having conceded that programmers who start late are not necessarily at a disadvantage in terms of technical skill, you've changed the question to whether programmers are "fine and capable employees" or "more than that," apparently meaning success in the tech industry as a startup founder.

Younger people have advantages entrepreneurially speaking. But we shouldn't mistake that for technical skill by redefining hacking as "being a technical startup founder."

It really depends onto what one has been doing these 22 years ;) There's an effect of transfer learning. And I was not talking about the very top. You are right, getting to the very top, like Woz or Carmack probably requires that early start.