An option might be to send RMS posts from users with low karma into a slashdot-esque crowdsourced moderation. Randomly show it to 10 users with a high enough karma on any of the list pages (frontpage, newest, classic, etc.). It's then only made public if a high percentage of such "moderators" approve of it (not necessarily counting "approval" as an upvote, just meaning "this is legit"). The percentage could be pretty high, say 80%, and it could optionally be log(karma)-weighted or similar. 100% might be risky; I have no idea.
I have to say I'm not a huge fan of the slashdot moderation system in general, but it could work within this sort of limited scope. Allow people to opt-out altogether via a setting if necessary.
Tangent: I was amazed at the first spammy RMS submission I encountered. It had actually been upvoted. Spammers self-upvoting, or people wholesale upvoting any RMS they see without looking at it first?
Rating without looking at the underlying content is horrible. Perhaps it could be mitigated by tracking if the link was followed, weighting authenticated upvotes more heavily than unauthenticated ones (where "authenticated" means a click-through to the content was observed).
You could track a link click either by setting a cookie client-side with javascript, or by linking to a trampoline page on HN that tracks the click, then redirects to the real URL.
I have, on occasion, voted up a link on HN without clicking through iff I have already read the article in question on, say, reddit. (I mouseover to make sure it's the same URL, though, and not just title and domain.)
I have to say I'm not a huge fan of the slashdot moderation system in general, but it could work within this sort of limited scope. Allow people to opt-out altogether via a setting if necessary.
Tangent: I was amazed at the first spammy RMS submission I encountered. It had actually been upvoted. Spammers self-upvoting, or people wholesale upvoting any RMS they see without looking at it first?