That's surprising. Would anyone mind explaining why that's the case? YC has no such restriction, and it seems to work out ok. Why wouldn't later stage investors do deals with multiple startups in the same space?
In terms of why YC can get away with this, companies are so early that it's not clear what they will ultimately be doing in the end. Many of the internal competitors ended up that way by accident. Moreover, YC doesn't make singular large bets, while later stage investors do. As a result, they aren't nearly as diversified as YC is and so a duplicate represents a much greater overall risk (essentially guaranteeing a loss before you even start).
Hmm. YC does, and it seems to work ok. Why wouldn't it work for later stage VCs? Or rather, why does it work for YC?
A later-stage VC, though, you might be having conversations with portfolio company A about how to entice sales people from portfolio company B, or portfolio company B might be getting ready to expand internationally and you're advising them on the best way to capture market share in Asia, where company A is stronger.
It's much more specific, targeted advice and consultation vs. "here's how we're going to help you get off the ground"
YC doesn't have to give up the best of its limited resources to competing startups. They just help get you to a Series A round. They can do this for competing companies without any issues.
VC's help assemble an executive team, provide critical business relationships, etc., and it's not in anyone's interests if they have to divide their best contacts and resources between two competing companies. They should want to go all-out in providing one company all their best resources.
At latter stages it might arise if a company pivots into another space. See this on how Andreessen-Horowitz dealt with a conflict (Instagram v. PicPlz):