If you've already pointed them to Three Worlds and HPMR and they are interested, they will likely start poking around themselves. So, unless they've asked specifically for suggestions, it isn't clear to me that such suggestions are a good idea.
Assuming they have shown that they are amenable to such suggestions, one slightly Dark Artish but potentially useful idea is to send them first to things they are more likely to agree with. So for example, if they are an agnostic or atheist, sending them to the religion related posts (like the Mysterious Answers sequence) will get them more likely to feel that the website is worth reading. This is because they will be exposed to a combination of arguments they've seen versions of before and ones they have not seen before and all those arguments will likely support their pre-existing viewpoint. This is a very effective way of making people sympathetic and willing to read more.
I recently introduced a friend to HPMR and she went on to discover Less Wrong entirely of her own accord. She has explicitly cited it as sparking her interest in things like Bayesian inference, which she would never have considered learning about before.
I can give a data point in favor of this approach: I've gotten at least one person to view this site (and even link to it on a few occasions) by showing them pages of interest; specifically of interest to them were "37 Ways That Words Can Be Wrong", the "The Neglected Virtue of Scholarship" and "The Best Textbooks on Every Subject". This person was an atheist with a predilection for autodidactism.
You're probably correct in your first point, I should introspect on why my first instinct was to manage their experience myself. I am concerned that a randomly selected article might put them off for the reasons I mentioned above.
Your second suggestion is likely to be extremely effective from the sounds of it. I wouldn't worry about the dark arts elements, if we are finding areas they are interested in already, and they are already rationalist-ish they will likely agree with the main points anyway, and be impressed by the overall structure.
Wouldn't it be best to simply link them to posts relevant to the current topic, whatever that is? (Politics, pundit's predictions, how much information evolution can maintain, the usefulness of knowing simple equations for Fermi calculations, etc.) At least, that's what I tend to do on the DNB ML or Reddit or my own site, and it seems to generally work well.
I agree. I do not link people to LW unless there is an LW article that strongly relates to the topic of discussion; but there is quite often a LW article that strongly relates to the topic of discussion. Earlier today, for instance, a friend complained on Facebook that the nation is obsessed with a murder trial and all but ignores several larger-scale tragedies both at home and abroad. I talked about scope insensitivity for a bit, and linked to the LW article on it.
Yeah, I recently noticed that myself - this Anthony trial or whatever. I was thinking to myself what LW article would be most insightful - perhaps the Bayes' theorem article showing that proving a faked robbery was equivalent to proving Amanda Knox & Sollecito guilty? - but didn't come up with anything solid.
I suggest Truly Part of You, as it has the highest ratio of insight to length. My only concern is that the jargon is uses may violate your first criterion. However, the jargon is explained sufficiently where it's used in TPoY; it's only a problem if seeing such terms is intimidating to your friend. For example, there's the passage
So, McDermott says, "A good test for the disciplined programmer is to try using gensyms in key places and see if he still admires his system. For example, if STATE-OF-MIND is renamed G1073..." then we would have IS-A(HAPPINESS, G1073) "which looks much more dubious."
I didn't know about gensyms at the time (and still don't have mastery of the concept), but you can see that in context it makes the point clear.
Maybe link them to an article at the start of one of the more fun sequences of posts, and also to the sequence list page. Reductionism or A Human's Guide to Words would be my picks.
EDIT: actually, the newb-friendliness of those may be hindsight bias - the Mysterious Answers to Mysterious Questions sequence underlies both of those sequences, and so is probably better to start with,
I've recently been trying to introduce some of my friends to less wrong. As a starting point I've linked them to HP:MOR and 'Three worlds collide' on the basis that they are both entertaining and accessible. But I'm not sure where to go next. Sending them to a main index page would likely be overwhelming, so could you suggest some sample articles to give them a general flavour of less wrong and envourage them to read more?
Qualities for such articles would probably include being:
What would you suggest?
[As I'm still a relative novice myself, apologies if a similar discussion has been had in the past and I haven't stumbled across it.]