This is too confused to follow as a human, and much too confused to program an AI with.
Also ambiguity aside, (2) is just bad. I'm having trouble imagining a concrete interpretation of "don't over-optimize" that doesn't reduce to "fail to improve things that should be improved". And while short-sightedness is a problem for humans who have trouble modelling the future, I don't think AIs have that problem, and there are some interesting failure modes (of the destroys-humanity variety) that arise when an AI takes too much of a long view.
Got it. Believe it or not, I am trying to figure out the rules (which are radically different than a number of my initial assumptions) and not trying solely to be a pain in the ass.
I'll cool it on the top level posts.
Admittedly, a lot of my problem is that there is either a really huge double standard or I'm missing something critical. To illustrate . . . . Kingfisher's comment "Something is clear if it is easily understood by those with the necessary baseline knowledge." My posts are, elsewhere, considered very clear by people with less baseline knowledge. If my post was logically incorrect to someone with higher knowledge, then they should be able to dissect it and get to the root of the problem. Instead, what I'm seeing is tremendous numbers of strawmen. The lesson seems to be "If you don't go slow and you fail to rule out every single strawman that I can possibly raise, I will refuse to let you go further (and I will do it by insisting that you have actively embraced the strawman). Am I starting to get it or am I way off base?
Note: I am never trying to insult (except one ill-chosen all caps response). But the community seems to be acting against its own goals as I perceive they have stated them. Would it be fair to say that your expectations (and apparently even goals) are not clear to new posters (not newcomers, I have read and believe I grok all of the sequences, etc. to the extent that virtually any link that is pointed to, I've already seen).
Another, last comment. At the top of discussion posts, it says "This part of the site is for the discussion of topics not yet ready or not suitable for normal top-level posts." That is what led me to believe that posting a couple of posts that I obviously considered ready for normal prime-time (i.e. not LessWrong) wouldn't be a problem. I am now being told that it is a problem and I will abide. But can you make any clarification?
Thanks.
Kingfisher's definition of clarity is actually not quite right. In order to be clear, you have to carve reality at the joints. That's what the problem was with the Intelligence vs. Wisdom post; there wasn't anything obviously false, at least that I noticed, but it seemed to be dividing up concept space in an unnatural way. Similarly with this post. For example, "selfish" is a natural concept for humans, who have a basic set of self-centered goals by default, which they balance against non-self-centered goals like improving their community. But if...
In the spirit of Asimov’s 3 Laws of Robotics
It is my contention that Yudkowsky’s CEV converges to the following 3 points:
I further contend that, if this CEV is translated to the 3 Goals above and implemented in a Yudkowskian Benevolent Goal Architecture (BGA), that the result would be a Friendly AI.
It should be noted that evolution and history say that cooperation and ethics are stable attractors while submitting to slavery (when you don’t have to) is not. This formulation expands Singer’s Circles of Morality as far as they’ll go and tries to eliminate irrational Us-Them distinctions based on anything other than optimizing goals for everyone — the same direction that humanity seems headed in and exactly where current SIAI proposals come up short.
Once again, cross-posted here on my blog (unlike my last article, I have no idea whether this will be karma'd out of existence or not ;-)