I think that the lesson you'd draw from many of these successes is something like this:
1. Find a specialty or niche that you care about.
2. Excel in it.
3. ????
4. Profit!
First, look at #1. For Microsoft, this was putting a modern (by the standard of the time) language on microcomputers as the PC era began. For Facebook, it was improving the way people manage and form weak ties. For Google, it was improving search.
These ideas seemed "lame" for businesses because they were hobbyist projects, but done to excellence because of the passion that was behind them. Still, hobby + excellence doesn't always make a business. You need a lot of help, and that help can be hard to summon.
Of course, there are also counterexamples. If Evan Spiegel excels at anything, it's writing creepy emails to his frat brethren. He's clearly not a technologist, and that applies to most of the VC darlings in this round, which is why it's going to be horrible for a lot of people when the chickens come home to roost and these P&D fake-tech companies have no buyers. But enough about that...
Where there's confusion, I'd say, is at step #3. To what extent is excellence (even in something that seems "lame", insofar as being impossible to build a business on, like fixing web search) going to get noticed and draw enough backing to overcome the extreme "pointiness" (as opposed to well-roundedness) that comes with narrowly-focused excellence (which seems, to most, to be a hobby project at best and obsessive at worst). And what are the odds? I don't know what the answers are.
I did #1 and #2 when I optimized hand-luck out of a trick-taking game, Ambition. ( https://docs.google.com/document/d/1S7lsZKzHuuhoTb2Wj_L3zrhH...
) It's been published in a few gaming magazines (mostly abroad, as in Hungary and Japan) but generally hasn't caught on, and I doubt it'll ever make me rich. I guess you can say that step #3 isn't automatic. But I'm not sure that that's a bad thing. Excellence doesn't need to produce wealth to be valuable, and doing something to make money makes excellence almost impossible in practice, anyhow. (Excellent products can make money, but products that are optimized for money-making are rarely excellent, because value-creation and value-capture are separate disciplines and antagonistic toward one another at the highest levels of performance.)
One thing I admire about Paul Graham (despite differences) is that he does place a value on excellence, as you can see in his work on Lisp. He understands excellence as a virtue (which is admirable) and believes that the market will reward it (which is debatable; we agree that there is correlation but would disagree on its strength). For his part, he tried to make an excellent Lisp (Arc) and showed a strong aesthetic sense. It didn't catch on. Rich Hickey made a Lisp that was, at the time, uglier but because it ran on the butt-ugly JVM, and you had a lot of S-expressions that looked more like
(.addEventListener thingy other-thingy)
than
(map (comp f g) list).
Eight years later, most of Clojure is beautiful (unless you peer into the innards) because of the work it has attracted. All I'm saying is that it's hard to predict who'll win. If it was 2006 and someone told you that in half a decade, the best Lisp would be on the JVM, you'd laugh in that person's face. And yet... that's what happened.
I think we can agree that excellence, because of the narrowness of focus it often (and not always, but often) requires, will lead a person to be thought "lame". (There is also the extremely slow process through which true excellence is recognized.) In 2006, many would've written Rich Hickey off as a barely-employed ex-musician trying to do the impossible (write a beautiful Lisp on top of the JVM) and unsellable (who'd buy it)? And yet Cognitect has shown that Clojure's excellence is sellable. From the statically typed world we seem similar slow uptake: Haskell's over 25 years old and just starting to be recognized as a practical language for "real programming". Full agreement exists on #1 and #2.
What about #3: the "????" term that leads from focused, diligent excellence to freely flowing money? Before Silicon Valley, I feel like Americans were more honest about the lack of a strong relationship between excellence and financial success. People understood the "starving artist" concept, and that writers of trash novels (e.g. Fifty Shades of Gray) would economically outperform people writing genuine literature. (That said, don't write off Tolkien, King, or Rowling just 'cause they do "genre lit"; they're fantastically skilled.) The problem with Silicon Valley exceptionalism is that these people believe that they've transcended the nonexistence of a causal arrow between excellence and financial success. That leads to a naive optimism at first, and later to a self-serving sense of "meritocracy" that's divorced from reality. The truth about item #3 in the above is that it's very random and we have almost no insight into how it actually works, and that's upsetting but I don't know what to do about it.
These ideas seemed "lame" for businesses because they were hobbyist projects, but done to excellence because of the passion that was behind them. Still, hobby + excellence doesn't always make a business. You need a lot of help, and that help can be hard to summon.
Of course, there are also counterexamples. If Evan Spiegel excels at anything, it's writing creepy emails to his frat brethren. He's clearly not a technologist, and that applies to most of the VC darlings in this round, which is why it's going to be horrible for a lot of people when the chickens come home to roost and these P&D fake-tech companies have no buyers. But enough about that...
Where there's confusion, I'd say, is at step #3. To what extent is excellence (even in something that seems "lame", insofar as being impossible to build a business on, like fixing web search) going to get noticed and draw enough backing to overcome the extreme "pointiness" (as opposed to well-roundedness) that comes with narrowly-focused excellence (which seems, to most, to be a hobby project at best and obsessive at worst). And what are the odds? I don't know what the answers are.
I did #1 and #2 when I optimized hand-luck out of a trick-taking game, Ambition. ( https://docs.google.com/document/d/1S7lsZKzHuuhoTb2Wj_L3zrhH... ) It's been published in a few gaming magazines (mostly abroad, as in Hungary and Japan) but generally hasn't caught on, and I doubt it'll ever make me rich. I guess you can say that step #3 isn't automatic. But I'm not sure that that's a bad thing. Excellence doesn't need to produce wealth to be valuable, and doing something to make money makes excellence almost impossible in practice, anyhow. (Excellent products can make money, but products that are optimized for money-making are rarely excellent, because value-creation and value-capture are separate disciplines and antagonistic toward one another at the highest levels of performance.)
One thing I admire about Paul Graham (despite differences) is that he does place a value on excellence, as you can see in his work on Lisp. He understands excellence as a virtue (which is admirable) and believes that the market will reward it (which is debatable; we agree that there is correlation but would disagree on its strength). For his part, he tried to make an excellent Lisp (Arc) and showed a strong aesthetic sense. It didn't catch on. Rich Hickey made a Lisp that was, at the time, uglier but because it ran on the butt-ugly JVM, and you had a lot of S-expressions that looked more like (.addEventListener thingy other-thingy) than (map (comp f g) list).
Eight years later, most of Clojure is beautiful (unless you peer into the innards) because of the work it has attracted. All I'm saying is that it's hard to predict who'll win. If it was 2006 and someone told you that in half a decade, the best Lisp would be on the JVM, you'd laugh in that person's face. And yet... that's what happened.
I think we can agree that excellence, because of the narrowness of focus it often (and not always, but often) requires, will lead a person to be thought "lame". (There is also the extremely slow process through which true excellence is recognized.) In 2006, many would've written Rich Hickey off as a barely-employed ex-musician trying to do the impossible (write a beautiful Lisp on top of the JVM) and unsellable (who'd buy it)? And yet Cognitect has shown that Clojure's excellence is sellable. From the statically typed world we seem similar slow uptake: Haskell's over 25 years old and just starting to be recognized as a practical language for "real programming". Full agreement exists on #1 and #2.
What about #3: the "????" term that leads from focused, diligent excellence to freely flowing money? Before Silicon Valley, I feel like Americans were more honest about the lack of a strong relationship between excellence and financial success. People understood the "starving artist" concept, and that writers of trash novels (e.g. Fifty Shades of Gray) would economically outperform people writing genuine literature. (That said, don't write off Tolkien, King, or Rowling just 'cause they do "genre lit"; they're fantastically skilled.) The problem with Silicon Valley exceptionalism is that these people believe that they've transcended the nonexistence of a causal arrow between excellence and financial success. That leads to a naive optimism at first, and later to a self-serving sense of "meritocracy" that's divorced from reality. The truth about item #3 in the above is that it's very random and we have almost no insight into how it actually works, and that's upsetting but I don't know what to do about it.