If I’m understanding him correctly, he seems to suggest that much of success can be attributed to luck. That is, being at the right place at the right time.

But if being lucky means all the stars align and the market takes off, such as storage cost drops to enable Dropbox, or processing power per dollar drops to enable more expensive compute, really all players in the industry are able to take advantage of it. Therefore it doesn’t explain why Facebook was the one and not MySpace.

In my opinion, being a category killers comes down to having an revolutionary product that truly separates you from the pack as well as be at the right place at the right time. Google pagerank probably was the 10x improvement AND they had good timing. On the same line, Dropbox just worked at a time storage and bandwidth cost was rapidly decreasing and was 10x better than the competition.

Of course, there is another category of category killers that didn’t have any competitors to come before it. And that is likely because from an investment perspective, it looked so unfavorable no one even bothered.

As such, one offs such as Musk who are willing to act irrationally by funding the likes of SpaceX (original goal is to increase public interest in space) and Tesla created those category killers.

It’s more than simply luck.

Just my two cents.

That's true, but you also have to be well positioned to capitalize on it quickly when it happens. Dropbox was, and so was Microsoft.

It's not necessarily the insights that are luck...it's all the missed equivalents that we've all never heard of. There may have been 10 companies trying to do more or less exactly what Microsoft was doing. They may all have been at the right place at the right time, but Microsoft had a little extra something that's hard to pin down. While i'm speculating with Microsoft, this is certainly true in the case of social networks. There were dozens of them at the time. Livejournal, Xanga, Friendster, MySpace, etc...They were all doing more or less what Facebook was doing. What did Facebook do so right that they didn't? There's certainly lots of ex-post just-so narratives about their exclusionary, college based early access model. But it's very hard to know if that was truly causal or not.

I will say that the one example of Gwern's that I really take issue with is Google. Google was very much not just another search engine. Google had a true, deeply differentiated technical insight that nobody else had. They solved search. The 'search engines' that existed before Google weren't even worthy of the name. They really innovated, and they deserve every ounce of credit for doing that. While luck is always a factor in these things, I think Google's success was as genuinely merit-based as any company's i'm aware of.

Had they not been fortunate enough to stumble across the right way to do ads, they would have been yet another casualty of the dot-com bubble - 'yeah, they did a nice job with $TECH, but everyone knows you can't simply scale and count on finding a revenue model later on, which is what all those dot-com bombs did. Eventually you run out of greater fools, you know. Too bad, I liked their site, among many others.'

I absolutely do. Before google it was mostly dumb link aggregating ; I remember going pages and pages in altavista and lycos and using "search helpers" like Copernic and going through various shady website link lists until google came ; I never looked back because I mostly never had to look beyond the first page to find what I was looking for with google.

> If they failed, google.com would remain a time capsule, the Yahoo deal would likely collapse, and the company would risk burning through its funding into oblivion.

PageRank was nice but couldn't save them from that.

> Google's innovation was discovering the ad auctions that turned their gushing red ink into gushing green ink. PageRank is a nice idea, but I don't remember switching from Dogpile to Google and going 'oh my god! who are these Google people? They solved search. All the search engines I used before were not even worthy of the name!' I wouldn't call it so extreme as to redefine the entire category; I wouldn't call early Google even 10x better than its competitors. Maybe 2 or 3x better.

I guess we may just have to agree to disagree here. I remember Google being a fairly revolutionary improvement over the competition. PageRank was definitely a fundamental, novel insight into the underlying structure of the internet. IMO it made a very big difference in the quality of their results, and their immunity from being gamed.

> Had they not been fortunate enough to stumble across the right way to do ads, they would have been yet another casualty of the dot-com bubble - 'yeah, they did a nice job with $TECH, but everyone knows you can't simply scale and count on finding a revenue model later on, which is what all those dot-com bombs did. Eventually you run out of greater fools, you know. Too bad, I liked their site, among many others.'

I do think you're right that their ad-auction model was also a deep fundamental insight and I maybe should have included that as well. But in my opinion both the ad-auction model and the PageRank algorithm are cut from the same cloth: Deep, fundamental technical insights into the underlying structure of the problem that solve it in a (metaphorically) closed-form way.

I find this to be in stark contrast to say, Facebook, Instagram, Twitter, etc...who had no deep insight into much of anything. Not to say that they weren't smart, and maybe had some insight into human behavior, but I think luck was a much, much bigger factor there than it was for Google.

I feel that 2007 was a massive shift in internet consumption habits/demographic.

Prior to this, it was still relatively young/geeky - Mum and Dad would use it for e-mail and work research, but not much else.

Then the iPhone came along, and suddenly, your average Joe and Jane were flocking to the internet on their mobile devices. It was no longer geeky, it was mainstream.

Facebook opened the gates to the (non-university) public right around the same time, which meant they were perfectly positioned to capitalize on this colossal new demographic of internet media consumers.

I'm not saying that they wouldn't have built a successful social network without the iPhone. But I do think it would have just been another Myspace.

Same goes for Twitter, mind you.

Facebook went public before figuring out mobile. Something in your timeline is off.

    > They were all doing more or less what Facebook was doing. What did Facebook do so right that they didn't?

They were the first to really hammer on the "use your real identity, as opposed to an avatar of some anime character or whatever" thing.

That made FB a useful tool for connecting with people you already know which turned out to be a thing that billions of people wanted.

That meant your mom or your dad could join Facebook and instantly have a social network. They didn't have to piece it together from scratch, and slowly piece it together with a bunch of bizarro internet avatar names like XXanimeloverXX or whatever that may or may not have corresponded to people they actually knew in real life.

If you think about it, things like LJ and Xanga really were vestiges of "the early Internet" where the main attraction was "creating a virtual avatar" that could be very separate from your "meatspace" identity.

Now, don't get me wrong! I actually really prefer that early internet and miss it dearly. But it's easy to see why it was utterly baffling to your mom or your Uncle Harry, and why Facebook's more straightforward identity model finally made social networking comprehensible and useful to regular folks -- you know, the 99% of people out there that didn't want the future to resemble Snow Crash.

A good thing that Google is too smart to crowd out organic search results with more and more paid results.</s>