OK, so I know it’s totally uncool by now to write about this year’s Startup School (which I’ll refer to as SUS from here on out). But I’m going to do it anyways. My excuse is I already wrote most of this and I moved the day after SUS. Oh and I was really sick and swamped with work for almost two weeks. Plus I haven’t seen many articles that I liked about what people took away from it. So here goes…

Preschool

I’m guessing not many people knew about this, but there was “a reception for speakers, press, and some of the attendees” the night before SUS. Someone made a mistake and invited me. Sadly they made a few other mistakes as there was no press and only one speaker there, Paul Graham himself. Basically it was just a bunch of random people with a large number of the currently funded Y Combinator startups. I was surprised by the abnormally high concentration of people from the UK. I was not surprised by the abnormally high concentration of iPhones. Everyone had one. And if they didn’t, they had a Blackberry. Although a couple people had both a Blackberry and an iPhone.

After a few minutes at the very homey, warm, and intimate Y Combinator office, I finally found out what the invitation had meant by “there will also be large metal robots!”. A robot named Monty rolled out on a pair of segway-like wheels. Monty started shaking people’s hands and throwing the horns. It was really impressive, but not nearly as astounding as the second robot in the Anybot workspace. Dexter was a biped robot, learning to walk. He can walk, jump, and catch his balance if pushed. Just watch:

I found it really interesting the mix of people who showed up. I shouldn’t have been surprised to find a few of them knew Mike Speiser. That guy knows everyone. It was mostly 20-something nerds. I was impressed by the attitude and energy and drive of everyone I met. They all wanted to “build something people wanted” as Paul Graham puts it.

Saw a bunch of impressive startup demos during the reception; there were a lot of friggin’ smart people. The most impressive startup I saw was called 280 North, and their product, 280 Slides, was almost literally Apple’s Keynote but on the web (not surprising since I believe they all used to work at Apple). I’m not talking about some simple Powerpoint web clone. This was the shit. I had a screenshot but they requested I take it down until they launch.

280 Slides felt almost exactly like a desktop application. I suspect the reason for this is the startup wrote the whole thing in something they called Objective-J, which as they explained it was essentially Objective-C style Javascript. There’s some sort of interpreter that runs client-side—but trust me—it doesn’t feel slow. Their very brief explanation of the whole thing only made me more convinced they were some of the smartest engineers I’ve ever met.

Someone put on loud music early on, which combined with the small space and packed room, caused me to all but lose my voice. I had been there for almost six hours and my feet were aching. I was tired and eager to get to sleep so I could get to SUS on time in just a few hours.

Startup School at Stanford

There really no way to understate this, so I’ll just say it. Parking at Stanford is a bitch, especially for guests. I walked into Kresge Auditorium, 20 minutes late, hoping I wouldn’t get a ticket. It turns out it was OK, I wasn’t really interested in the first two speakers anyways.

Paul Graham really surprised me when he took the stage. The last words I expected to come out of his mouth would be “stop using the Internet”, but that’s precisely what he told us. It turns out he was using 280 Slides and with all the attendees trying to use the wifi, Paul couldn’t load his presentation. His talk was good but just wasn’t as inspiring as I had hoped.

David Heinemeier Hansson gave my favorite talk of the day. He was definitely the rockstar of everyone there. He went right on the heels of Greg McAdoo, a VC at Sequoia Capital (who has funded just about every major tech company). DHH had lots of little pokes at Greg’s main points, which boiled down to the need to turn your company into a billion dollar company and some sort of surfing metaphor for catching the next big wave. I dunno, he wasn’t really saying much other than get lucky and make me hundreds of millions, which was a lame message.

DHH was the complete opposite of the preceding talk. His title was “The secret of making money online”. First he presented the classic Slashdot comment:

1. Great Application
2. ???
3. Profit!

What could step two possibly be? It’s having a price DHH told us. You have an application, you set a price, and people pay you money. “I’ve heard, over time, hundreds of years actually, that has been how most businesses made their money,” David stated plainly. For some reason this notion of having a price doesn’t seem to have made it to the Internet. Everyone goes for free with ads, or anything but actually putting a price on their product.

DHH put up some basic math trying to remove the idea that making a million dollars a year isn’t as hard as you might think. If you built a great application and sold subscriptions for $40/month, all you would need is ~2,000 users. That’s not unthinkable. Build something people want. Focus on what DHH calls the Fortune 5,000,000. Businesses tend to focus on the consumer or the enterprise; there’s no middle ground.

I loved and was relieved to hear him talk about how it’s OK to just make millions, and you don’t have to rush it. DHH explained how the patterns 37signals set, stuck with them. They don’t work crazy hours, in fact they just moved to four day work weeks. As he said if you could just get actual work done for two hours a day you’d be in good shape. His message of calling your own shots and running at your own pace hit welcome ears. I would love to have a job akin 37signals.

One thing that really comes through with DHH in person is how sarcastic and hilarious he is. So just trust me, you seriously have to watch his presentation, then come back and finish reading.

Jeff Bezos came to SUS to talk about Amazon Web Services, which I assumed would be quite boring. I’m sure many if not most tech people know of AWS (S3, EC2, SimpleDB, SQS, FPS, and Mechanical Turk), but most people on this planet haven’t heard of it and I believe this quote summarizes that in an unexpected way:

“Amazon on Monday announced persistent storage for its EC2 service and what’s notable is how quickly the e-tailer is running ahead of the competition. In fact, Amazon’s real business down the line will be its cloud services. Amazon will be like a book store that sells cocaine out the back door. Books will be just a front to sell storage and cloud computing.”
Larry Dignan of ZDNet

Bezos explained the basic AWS principles, easy to use, fast, elastic, highly available, and pay by the drink, were centered around a story I had not heard. Back in the 19th century (I think) breweries started to build their own generators because electricity allowed them to make beer better. Better meaning more efficiently as there was no difference in taste. The reason they had to build and manage their own generator was the electrical grid as we know it today did not exist yet. AWS aims to “make electricity, so you don’t have to”.

Bezos was charismatic and brilliantly diplomatic. I hugely respect him now. One question asked about a frustrating limit around SimpleDB and domains. Bezos didn’t know enough technically to answer the question so he asked a colleague of his to field the question. The colleague started giving a fluffy non-answer, totally avoiding the question. Bezos interrupted his colleague and said, “so basically he’s not answering your question. Come find us after the talk and we’ll work something out.” I’ve never seen someone handle a situation with a frustrated customer as well as that one.

Peter Norvig gave an interesting talk but it really wasn’t what I was hoping to hear for some reason. He did say something really funny though when someone asked a really lame question about semantic web and the future. Norvig dropped this golden nugget after a slight pause, “the semantic web will always be the future of the web”.

Check out my Flickr set from the weekend. I highly recommend all the other SUS talks on Omnisio’s site.