A number of the ideas I've had for new disruptive companies aren't traditional web applications or services, but instead are hardware producers, new takes on systems software, and re-imagining infrastructure. More SpaceX and NeXT than Hipmonk and Posterous.
These markets are full of slow-moving entrenched players: a ripe target for disruption in a major way. However, it's probably a harder market to get into, with more capital outlay required to get started.
My question is: would YCombinator be interested in such a startup? I imagine that the level of funding YC gives would help a team of founders get a purified idea of what they intend to make, and maybe some very early stage prototypes of software, though probably not hardware. Being successful would entirely depend on getting a funding round once we produce the plans so we can hire talent, purchase materials and start manufacturing.
"27. Hardware/software hybrids. Most hackers find hardware projects alarming. You have to deal with messy, expensive physical stuff. But Meraki shows what you can do if you're willing to venture even a little way into hardware. There's a lot of low-hanging fruit in hardware; you can often do dramatically new things by making comparatively small tweaks to existing stuff.
Hardware is already mostly software. What I mean by a hardware/software hybrid is one in which software plays a very visible role. If you work on an idea of this type you'll tend to have the field to yourself, because most hackers are afraid of hardware, and most hardware companies can't write good software. (One reason your iPod isn't made by Sony is that Sony can't write iTunes.)"