I've tried explaining LW stuff to my high school friends, and had the most success explaining map and territory, cached thoughts, and what is meant by truth/evidence. But don't try to cram too much into one hour.
A lot of people will think (due to hindsight bias, I assume) that this stuff is obvious once you've explained it. This means they won't take it as seriously and are less likely to check out follow-up materials. I would start by asking them (if the format makes this possible) what they think "truth" and "evidence" mean. Usually they run out of ideas within 30 seconds, and then are more interested in what you have to say about it.
I suspect truth could take long enough to unravel all by itself if you first ask the audience what they think it means, but on the other hand I think that what evidence is is a major issue on which most people are tremendously confused, and if you clarify that, a sense of what truth is will probably accompany it.
Important points to cover:
Conservation of evidence; people intuitively tend to raise their confidence, or at least not lower it, in the face of any evidence consistent with their beliefs.
Absence of evidence is evidence of absence; addresses such a common mistake that its negation is a popular saying.
Positive bias; one of the most pervasive biases preventing people from usefully testing their beliefs.
With just an hour, I think I'd focus on evidence and map vs. territory, or maybe even entirely on evidence.
You can't fit very much into just one hour. To maximize your effect, make the hour entertaining, and leave them with a handout and links that will make some of them spend much more time following up.
Here is a presentation that was used in a similar setting before.
I recommend trying to cover less than you currently plan. Just one or two big ideas should be more than enough.
Second on 4 topics in an hour being too much. That only leaves 15 minutes for a topic. I'd have trouble fitting a good explanation of any of those in to 15 minutes, and high school kids will probably have follow-up questions that make it more time consuming.
Biases is wonderful, but I think it also needs the biggest block of time to really be effective - teaching one or two biases doesn't help a lot, and risks giving them Fully General Counterarguments. No student should be trusted with that sort of armament ;)
I'd personally pick two of your topics to focus on, then have a few related biases you can include to stretch things out if everything goes well - I've found that having flexible plans that can stretch/compress is very good for dealing with groups that are likely to ask a lot of questions, or for an informal context where people will get distracted.
My personally, I'd be biased towards "make beliefs pay rent" (as that was most useful to me) and "map and territory" :)
Map and territory, reductionism and what having a belief should mean (paying rent) are all much, much too complicated topics to cover in a one hour class for kids (it's not ever close), even if you just picked one of these topics. (I've been teaching college students for 14 years.)
Consider teaching simpler topics such as ignore sunk costs, tradeoffs are everywhere, "measure twice and cut once" and even "think before you act".
I have no idea how to teach "think before you act." It's the sort of nonsense that adults always say to kids (I myself have said it to a kid and when I did I died a little inside). If you could teach kids that, you would be a millionaire.
"Think before you act" is a simple way of saying that there are significant costs to not exercising impulse control. You teach it to kids through repetition and by pointing out the correlation between positive life outcomes and impulse control. "Think before you act" also means that most people devote sub-optimal amounts of time to thinking about major life decisions.
I wonder if instead, kids simply learn it as their brains mature. That is, maybe the repetition does nothing but annoy everyone. There are all sorts of studies that show that brains continue to mature even into the twenties. When, as a kid, I heard the phrase "Think before you act", it was usually not in the context of major life decisions -- it was in the context of measure-twice-cut-once situations. To be fair, some of these could turn out to be major life decisions (i.e. diving head-first into an insufficiently deep pool), but they aren't intended to be so.
I am going to do Splash! Fall 2011 at the University of Chicago, where I chose a subject to teach kids. I'm teaching a 1 hour class on the basics on rationality, and I am outlining the topics I want to cover.
Right now, I'm planning on teaching map and territory, reductionism, a basic introduction to biases, and what having a belief should mean (paying rent).
What other ideas would be useful to teach high school students? And does anyone have suggestions on interesting ways to teach these concepts?
Edit: All comments say four is too many topics, I will focus on map and territory and beliefs paying rent.