[DocSavage]

Sadly, the attitude evident in the last two paragraphs of your comment runs deep in young people. Zuckerberg at YC Startup last year, age discrimination suit at Google, etc. Logan's Run in Silicon Valley.

It's also upside down compared to other fields like medicine and writing. In medicine, you make your "mark on the world" after you go through significant training, which ends around 30 or later.

As far as making it entrepreneurially:

1) If it depends purely on intelligence or hacking ability, PlentyOfFish.com and other success stories wouldn't have made it. Even if it's somewhat true, consider that the decrease of your ability with age should be less than the interpersonal variability of ability. (50 yo Einstein > 20 yo Homer Simpson)

2) If it depends on luck, don't worry about age.

3) If it depends on connections, you get more with age.

4) If it depends on others' perception of your youth.... you may be in trouble.

[nostrademons]

Markus Frind is quite a bit more intelligent than many people give him credit for, eg. he discovered the then-longest arithmetic progression of prime numbers in 2004. Same goes for many other entrepreneurs that people assume were just lucky - James Hong, Marc Andreesen, Brad Fitzpatrick, etc. And the variation due to age may be much less than the variation due to person, but remember that there are many more smart people than there are millionaires. Empirically, it seems like you need to be smart, young, and lucky.

Also - I doubt few people would dispute that succeeding with a tech business requires the ability to concentrate intensely for long periods of time. (Something I'm obviously failing at, having been back to news.YC 3 times so far today.) Certainly many young people lack that, but for a given person, it seems likely that they're better able to do that at 20 or 25 before things like family, business connections, or reputation intercedes.

It definitely is upside down compared to medicine or writing, but that's because computer programming is nearly in unique among human fields in being subject to lots of chaotic effects. Almost every other body of knowledge is additive, but in software, small changes to requirements at the base layer can have ripple effects that invalid your whole system's design. That's why the software industry seems to undergo major tectonic shifts every 10 years or so, each of which creates lots of opportunities for young entrepreneurs who haven't learned the now-obsolete ways of "how things are done".

When I was a kid, the hot new field was microcomputers, the programming language was assembly (no wait...Pascal. no wait...C), and the largest platform was Windows. When I was a teenager, the hot new field was the web, the hot programming languages were Perl and PHP and Java, and the largest platform was Netscape. Now that I'm in my twenties, the hot new field is rich Internet applications, the hot programming languages are Flash, JavaScript, Python, and Ruby, and the largest market is Internet Explorer. I can almost guarantee that the hot field, language, and platform will be different in 10 years, and all the technical skills I've learned so far will be obsolete.

Edit: On reflection, it's not really software as a whole that's chaotic so much as the leading edge of software, i.e. the edge where you can still have a viable startup. If you're making a desktop app, the tools and skillsets are only slightly different from making a desktop app in 1995. If you're making a webapp that's just a sequence of forms, the tools from 1998 will work fine. But those are mostly solved problems: either there's an off-the-shelf solution that's already been built, or you can hire one of an army of consultants that knows how to build one. The leading edge, by definition, involves tackling problems that haven't been solved before, and so there's no advantage to experience in solving them.

[DocSavage]

I didn't know about Markus Frind's intelligence (interesting info), but in PlentyOfFish's case (and successes like MySpace), it's no masterpiece of hacking.

Writing prose can be similar in intensity to writing code. John Scalzi, a recent sci-fi award winner, wrote a book for writers called "You're Not Fooling Anyone When You Take Your Laptop to a Coffeeshop." Many writers block out pure concentration time. Nonetheless, I'll concede that long stretches of time are more available to the young, unless they are out partying :)

I wonder how many big successes depend on bleeding edge software? For every Google, there are plenty more like MySpace, Ebay, Amazon, and Facebook.

I usually live off specialized applications, where you build software that focuses on a particular need for a niche population. Knowing the domain (say radiotherapy or creative writing) is really useful because of insights into how you can fashion new tools using advancing technology (e.g., faster 3D graphics or a better web framework). So while I agree that creating totally innovative approaches might be easier for the young, I think experience helps in creating domain-specific software.

My main point, though, is even if you believe you have a limited shelf-life, you might consider it's your shelf-life in only one of many possible avenues to make your mark.