No matter how early and rough, show your product to potential customers and listen to what they say. You'll learn more from an hour doing this than a month of brainstorming with your team about what features you should someday maybe add to make your product compelling.

Over many cycles of sharing many prototypes with customers, you'll figure out what your future users actually want your product to do. This will almost always be way different than your original vision.

And stop worrying about whether you're too early: many of our best startups get rejected before applying again with progress and getting accepted. Pretty much everybody thinks they're too early or late for YC, and they're all wrong;)

(Disclaimer: I work for YC advising hardware startups.)

Rigetti computing was accepted with no traction but the founder had significant market expertise.

That said, keen to hear what kind of milestones are expected of early stage hardware startups.