= 27d567bc2d48d9f40e9ccc20ea370b15)
Related: The Math of When to Self-Improve.
In response to
Toy problem: increase production or use production?
Oh, haha, yup! Now I'm anticipating a flood of negative comments on my post as well.
Toy problem: increase production or use production?
There is a class of problems that I noticed comes up again and again in various scenarios. Abstractly, you can formulate it like this: given a time limit, how much time should you spend increasing your production capacity, and then how much time should you use your production capacity to produce utility? Let's take a look at two version of this problem:
Version 1:
You have N days. You start with a production capacity C=0 and accumulated utility U=0. Each day you can either: 1) increase your production capacity (C=C+1) or 2) use your current production capacity to produce utility (U=U+C).
Question: On what days should you increase your production, and on what days should you produce utility to maximize total accumulated utility at the end of the N days?
It's trivial to prove that the optimal solution looks like increasing capacity for T days, and then switching to producing utility for N-T days. What is T? In this case it's really straight-forward to figure it out. We can compute final utility as U(T)=(N-T)*T. The maximum is at T=N/2. So, you should spend the first half increasing your production and the second half producing utility. Interesting...
Version 2:
You have N days. You start with a production capacity C=1 and accumulated utility U=0. Each day you can either: 1) increase your production capacity by a factor F (C=C*F), where F>1 or 2) use your current production capacity to produce utility (U=U+C).
Same question. Now the final utility is U(T)=F^T*(N-T). Doing basic calculus, we find the optimal T=max(0, N-1/ln(F)). A few interesting points you can take a way from this solution:
1) If your growth factor F is not large enough, you might have to stick with your original production capacity of 1 and never increase it. E.g. F=1.01 and N=100, where optimal T=max(0,-0.499171).
2) The bigger the N, the lower growth factor you can accept as being useful, i.e. T>0.
3) For most scenarios, you should spend 80-90% of the time increasing the production. Example. With larger F, T will approach N. This reminds me of Reducing Astronomical Waste post.
Questions for you: where have you seen these types of problems come up in your life? Is this a known class of problems?
In response to
Some alternatives to “Friendly AI”
Either of the first two options sounds good to me.
In response to
Is there a way to stop liking sugar?
I recall the claim on LessWrong that you can train yourself to hate chocolate by getting a pound of M&Ms, putting each one in your mouth and chewing it up, then spitting it out rather than swallowing it. After a while, all you'll feel is the crappy taste and not the feeling of having eaten it.
Dunno how that would go for sugar in general.
I did a low-carb diet for a while (I have given it up and am revelling in sugar and starch) and anything sugary was like OH MY GOD CRACK on my cheat day.
I've tried this with chocolate chip cookies and it worked, but only for that specific brand of cookies. I think it's because my mind remembered that box very clearly, so a lot of the disgust became associated with it and now the actual cookies. If someone wants to try it, try to avoid this mistake.
Also, spitting it out is not enough. You have to actively feel disgust as you are looking, smelling, touching, and tasting the cookie. "Oh my god, looks like squishy it is. Looks like shit. Argh, that nauseating fake-sugary smell is so disgusting. Do I want a bite? Oh god, no, how revolting. I'm about to puke." I'd go through that for each cookie in the box (about 2 dozen), then crumble it in my fingers, and throw it in the trashcan.
In response to
Quantum Decisions
On Monday, I have to decide whether to eat ice cream or not. After a little thought, I decide it's all right to eat ice cream because <reason>. The reason is a lame reason that my mind came up with because it really wants ice cream.
On Tuesday I only eat ice cream with 50% probability.
Nisan, please see the revised version of the question.
On Tuesday, I'm still picking my threshold "manually". Let's say the quantum random number generator I use can give me a random 1-6 number, just like a 6-sided die, so I'm not stuck with 50/50.
In response to
Quantum Decisions
I guess part of what you are asking about is the difference between using a quantum random number source, and using a deterministic pseudorandom number source (which happens to be inside your brain). Stuart Armstrong wrote Quantum versus Logical Bombs about one difference, suggesting that maybe risk-averse actors should prefer the quantum version (because then even if you make a bad decision, at least there will exist some multiple worlds in which you didn't...).
Yes! Thank you, Pfft, that's exactly what I was trying to get at. I'll go read the post right now.
That's the point of this question.
I don't think it is the point at all, since the difference between quantum random and pseudorandom has nothing to do with making decisions. I think you are simply using a cool-sounding word "quantum" where "probabilistic" is what you really mean.
Sigh. No, I mean precisely quantum and not probabilistic. I'm wondering how it affects the "reality fluid", for example. I'll go and edit the question to make it more clear.
In response to
Quantum Decisions
Why are you using the word "quantum" here? Do you expect any difference if you use a classical pseudo-random number generator... say, on Wednesday?
May be. That's the point of this question.
In response to
Quantum Decisions
Two questions:
What kind of decisions do you have in mind? Presumably I'm not rolling each time I take a step (or not) while walking down the street, right?
Is there any reason not to say 100% for all my decisions on Tuesday? (Assuming I don't happen to find myself in any adversarial interactions with Omega.)
See my answer to gjm.
Let's say I'm lazy, and I can't decide for sure if I want vanilla or chocolate. Rolling the quantum die seems easier.
= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
= f037147d6e6c911a85753b9abdedda8d){kind=logo}
Most of the time I'm sitting in front of a computer (modern equivalent of pen and paper), so talking out loud seems to be rather unnecessary.