All of jknapka's Comments + Replies

jknapka100

Survey taken. I hope I didn't break it - I am a committed atheist, but also an active member of a Unitarian Universalist congregation, and I indicated that in spite of the explicit request for atheists not to answer the denomination question. (Atheist UUs are very common, and people on the "agnostic or less religious" side of the spectrum probably make up around 40% of the UU congregations I'm familiar with.)

jknapka00

(I am in the midst of reading the EY-RH "FOOM" debate, so some of the following may be less informed than would be ideal.)

From a purely technical standpoint, one problem is that if you permit self-modification, and give the baby AI enough insight into its own structure to make self-modification remotely a useful thing to do (as opposed to making baby repeatedly crash, burn, and restore from backup), then you cannot guarantee that utility() won't be modified in arbitrary ways. Even if you store the actual code implementing utility() in ROM, baby ... (read more)

0wgd
I think the important insight you may be missing is that the AI, if intelligent enough to recursively self-improve, can predict what the modifications it makes will do (and if it can't, then it doesn't make that modification because creating an unpredictable child AI would be a bad move according to almost any utility function, even that of a paperclipper). And it evaluates the suitability of these modifications using its utility function. So assuming the seed AI is build with a sufficiently solid understanding of self-modification and what its own code is doing, it will more or less automatically work to create more powerful AIs whose actions will also be expected to fulfill the original utility function, no "fixed points" required. There is a hypothetical danger region where an AI has sufficient intelligence to create a more powerful child AI, isn't clever enough to predict the actions of AIs with modified utility functions, and isn't self-aware enough to realize this and compensate by, say, not modifying the utility function itself. Obviously the space of possible minds is sufficiently large that there exist minds with this problem, but it probably doesn't even make it into the top 10 most likely AI failure modes at the moment.
0Transfuturist
I'm not so sure about that particular claim for volatile utility. I thought intelligence-utility orthogonality would mean that improvements from seed AI would not EDIT: endanger its utility function.
jknapka70

I agree that basic probability and statistics is more practically useful than basic calculus, and should be taught at the high-school level or even earlier. Probability is fun and could usefully be introduced to elementary-school children, IMO.

However, more advanced probability and stats stuff often requires calculus. I have a BS in math and many years of experience in software development (IOW, not much math since college). I am in a graduate program in computational biology, which involves more advanced statistical methods than I'd been exposed to be... (read more)

-2JRMayne
"Computational biology," sounds really cool. Or made up. But I'm betting heavily on "really cool." (Reads Wikipedia entry.) Outstanding! Anyway, I concede that you are right that calculus has uses in advanced statistics. Calculus does make some problems easier; I'd like calculus to be used as a fuel for statistics rather than almost pure signaling. I actually know people who ended up having real uses for some calculus, and I've tried to stay fluent in high school calculus partly for its rare use and partly for the small satisfaction of not losing the skill. And probably partly for reasons my brain has declined to inform me of. I nonetheless generally stand by my statement that we're wasting one hell of a lot of time teaching way too much calculus. So we basically agree on all of this; I appreciate your points.
jknapka50

Noted. I'll try to update accordingly.

jknapka30

Raemon, this is really great. As a lay leader of a Unitarian Universalist congregation, I love what you say about the importance of ritual -- it can be strongly affecting, and can motivate people to action they might not otherwise take. If we can construct rituals that inspire and invigorate, without misleading, then that is a win.

I'd suggest that when doing this kind of ritual, we should invite guests who are almost-but-not-quite in the rationalist camp. It can be a tool to attract new minds.

I will try to do a similar event at my church next year. We h... (read more)

8Vladimir_Nesov
I expect this is wrong both as a model of the community and of Raemon's model of the community. Appearances don't matter, and psychological effects do.
jknapka00

I mean, eternal torture is pretty frickin' bad. I think in the end, I'd convert. And I'd also try to convert as many other people as possible, because I suspect I'd need to be cruel to fewer people if fewer people went against Christianity.

This is a very good point, and I believe I'll point it out to my rather fundamentalist sibling when next we talk about this: if I really, truly believed that every non-Christian was doomed to eternal damnation, you can bet I'd be an evangelist!

Extreme Altruism

While I might not donate all my money to save 10, I think

... (read more)
1DanielLC
If I really, truly believed that every non-Christian was doomed to eternal damnation, I'd donate to a charity that distributes condoms to people in Africa. The key here is to minimize the number of non-Christians, not to make more people Christian.
jknapka10

I seem to be succeeding in helping to convince my graduate program in bioinformatics to ditch Perl in favor of Python. I'm very happy about this! When you don't have a programming background, and you're going into a field with heavy programming, Perl will hurt you -- it's likely to make you dislike programming. Python OTOH is like the fuzzy kitten of programming languages -- but it still has claws! (By which I mean, you can do serious stuff with it, despite its apparent adorableness.)

Also I've just started juggling again after a longish hiatus. I just ... (read more)

jknapka00

I was over 100 years off, but in the opposite direction.

jknapka30

I took the survey, sometime last week I think. EDIT: I think I may also have messed up the "two-digit probabilities" formatting requirement. I can't recall specifically any answer that might have violated it, but I also don't recall paying attention to that requirement while answering the survey.

jknapka140

Hello, all. I'm Joe. I'm 43, currently a graduate student in computational biology (in which I am discovering that a lot of inference techniques in biology are based on Bayes's Theorem). I'm also a professional software developer, and have been writing software for most of my life (since about age 10). In the early 1990's I was a graduate student at the AI lab at the University of Georgia, and though I didn't finish that degree, I learned a lot of stuff that was of great utility in my career in software development -- among other things, I learned abou... (read more)

0shokwave
If you have the time and inclination to test this, you can use this site to discover your level of under- or over-confidence, and adjust appropriately. In any case, welcome to LessWrong! I look forward especially to hearing about the process of unschooling; there is (very rightly) an impression here on LessWrong that raising a child is one of the hardest tasks; it seems like also taking responsibility for their education is even more daunting!
jknapka90

Thanks, Raemon, this is inspiring. It reflects my experience learning to draw as an undergraduate, many years ago. I have not drawn much since college, but I do recall vividly the experience of, "Holy crap, is that what a person really looks like?!?" upon first producing a half-decent quick figure drawing. I eventually developed a pretty decent drawing ability, which has atrophied quite severely in the intervening years. The experience definitely influenced my overall thinking though -- I'm very aware that my brain is not telling me the actua... (read more)

jknapka100

"We DO NOT WANT lukeprog's How To Be Happy to sound authoritative. The reason for that is if it turns out to be 'more wrong' it will be that much easier to let go of."

This.

Whenever you give a collection of concepts a name, you almost automatically start to create a conceptual "immune system" to defend it, keep it intact in the face of criticism. This sort of getting-attached-to-names strikes me as approximately the opposite of Rationalist Taboo. (Hey, did someone just dis Rationalist Taboo? Lemme at 'em!)

4JoshuaZ
I suspect that giving a name to a hypothesis can cause you to defend it but it might be able to do the opposite also if it is already a hypothesis you dislike. I suspect that it is more likely to move one's emotional attachment towards extremes rather than move one's attitude in any specific direction. I also suspect this is more likely to be a problem for extended hypotheses that are more networks of interlocking ideas than simple hypotheses (so e.g. NLP would be a name in this sense, but I suspect that "Rationalist Taboo" would be too simple to have much of an actual impact.) Shorthand hypothesis names are generally helpful. I suspect that for most purposes naming hypotheses will provide more help (in terms of efficient communication and in terms of one's own mental shortcuts and processing) than it will harm.
1Solvent
The idea that giving a name to a hypothesis causes you to defend it is an interesting one. That's the most meta concept I've heard in a while.
jknapka00

So the best strategy would be to maximize, and then when you feel dissatisfied, remind yourself that this feeling is misplaced, since you've probably achieved a situation that is objectively better than the one you would have achieved via satisficing. Will that actually work to de-fuse the feeling of dissatisfaction, I wonder? (Personally, I am a habitual satisficer, and feel pretty happy about most things in my life, while recognizing that there are many ways I could have done better.)

3[anonymous]
Here's a more viable strategy. Ask a friend to pick your house for you, maximizing to his/her heart's content, and narrowing it down to 2-3 choices for you to personally pick from. This negates any dissatisfaction you might feel about maximizing, because you didn't have to.
jknapka20

Murder is the most common cause of death today for some groups (young African American males, for example).

I don't believe it is correct in general that intentional killing was the most common cause of death in primitive tribes; and if it was the case in specific groups, they were exceptional. The citation that occurs to me immediately is "Sex at Dawn" (Ryan & Jetha), which goes to some trouble to debunk the Hobbesian view that primitive life was "nasty, brutish, and short". (Also, my partner is a professional anthropologist with ... (read more)

3TheOtherDave
Yeah, it seems moderately plausible to me that in primitive tribes the killing of out-group individuals as part of inter-group violence would be a lot like war.
jknapka30

I'm a member of my local Unitarian Universalist church (in El Paso, just down the street from Waco by SW standards), and it is very friendly to atheists and skeptics -- I would say 15% to 20% of the membership would identify as "agnostic" or more skeptical. However, it is also friendly to an array of other, much less evidence-based views. I'd say a UU church would definitely be worth a look, and would almost certainly be a better fit for a LW denizen than a "non-denominational Christian" one. But one might need to be tolerant of some rather silly beliefs. OTOH, I'm starting to take it as an opportunity to learn to "evangelize" (gently).