[prawn]

The new YC fascination (recent RFS) appears to be in solving big problems without building the big product right away. You can build an add-on that works you towards some level of driving automation well before you build an entire robotic car, for example.

PG's line there is a reinforcing hint for those that missed the Boosted Boards example previously or weren't too optimistic in response to that RFS. Hopefully it helps more people think about how they could approach big problems in an initially accessible way.