Loved that video. I miss having PG be a very vocal and frequent voice of YC. There was a question about the financial risk of launching early - PG's answer was

Loved that video. I miss having PG be a very vocal and frequent voice of YC. There was a question about the financial risk of launching early - PG's answer was really good, but it seemed pretty clear to me: Launching early costs less time and money. So you should launch something that works, as soon as possible with the minimum number of people required to build something useful (probably 2-4 people). Beyond that, you make your "vision" meet reality, where you can do magical things like learn from users/customers, which is basically priceless.