Startup School: Gmail, FriendFeed creator Paul Buchheit on winging it

(I’m live-blogging from Startup School, a daylong program from startup incubator YCombinator held at Berkeley today. You can watch it here live or update as we go. This is paraphrased.)
Paul Buchheit, who created Gmail and co-founded FriendFeed, is giving a talk entitled “They told me to wing it and that it would be cool.” He has a few slides and is basically making up his talk has he goes along — which he says is a metaphor for founding a company. (So meta.)
He goes along, saying this is a good analogy because “You could be a success or a disaster or worst of all, just waste people’s time. People who’ve flown from all over the world here could end up saying, ‘I hate that guy so much,” he said.
So here’s a few points - “There is no substitute for experience.”
“Limited life experience + overgeneralization = Advice.” Everyone gets up here on stage and says don’t take funding or take funding. I’m careful to not to tell people to do anything one way or another because my experience will be totally different from yours.
He went on to talk about his life - from school to his first job at Intel, and then to Google at age 22. He “retired” from 29 to 30, and then started FriendFeed. Now he’s at Facebook at 32, after the company acquired his startup FriendFeed.
On high school: I really hated programming but I’m really obstinate. By the point that I got really good at C, I started to tolerate it. I really think looking back at that time period. The thing I got the most value from is from me trying to do things, or sell things or defeat the computer.
On college: I really recommend doing internships. I really liked my work at Microsoft — I learned I liked doing programming every day. I worked 20-hour days and realized — hey, this is fun. One summer I decided to write a webmail program. So I designed it all. I wrote a few widgets. Then I got bored. So I started watching Dukes of Hazzard. (This is a pretty good show.) And riding my bicycle around. On the surface, that might seem like a failure. But I actually learned a few good lessons - I learned how not to write software.
On moving to California: I started working at Intel. I started hanging out with friends and they started about their retirement plans. You could be retired at 57 — and wouldn’t that be awesome? And then I got really depressed. I felt like I was going to throw away the next 30 years of my life. I realized you could have spent your life doing nothing of consequence. So I then started looking around. I looked on Slashdot, saw a post about Google. They were smart, so I thought it would be fun.
On Google: This was the best education I could have ever had. It’s hard to pick which companies will be successful, but you can choose to work with smart people. That’s probably the best heuristic.
On Gmail: On the first day, I actually wrote the whole thing. But it was really bad. When I first created as Gmail, so it only indexed my own mail. Then he put it out to the company and everyone else gave me really good feedback like: “I want to read my e-mail, not yours.”
On quitting Google: For some reason, one day I quit. It was a great place, and I wasn’t fully vested. It just came down to — it seemed like a good idea.
On retiring: It’s not as great as it seems. Like our whole society is based around people working together. It’s really easy to end up doing nothing. I see a lot of my friends doing this. Nothing is not satisfying. And then I went back to that fear of not doing anything of consequence.
On Friendfeed: We started this new company with some friends. And then for some reason, we sold it to Facebook. It seemed like a good idea.
On Facebook: The most interesting part of Facebook is that it’s not Google. The risk of working in one place for too long is that you begin to believe — inside a company are all these shared delusions, they develop all these corporate superstitions. In Google, there are all these rules. And they keep using them because they’ve made them successful before.
Even if you’re not going to start a startup right away. If you’ve been in your job for awhile, you should quit. Google was really comfortable. I knew all the people. It’s important to do things that will make you uncomfortable. So there’s a couple of other points.