I maintain a reading list on Goodreads. I have a personal website with some blog posts, mostly technical stuff about math research. I am also on github, twitter, and mastodon.
Eigil Rischel(Eigil Fjeldgren Rischel)
I think I managed to avoid the Inbox Zero thing by not reading my emails, if the little bit of text that Gmail displays is enough for be to be confident that I don’t need to read or respond to the mail. This means that I have a huge, constantly growing number of unread mails in my inbox, so the idea of getting it down to zero isn’t really attractive.
I still check my email unnecessarily often, but I don’t feel a compulsion to read any new mails immediately.
I think I don’t understand the Löb’s theorem example.
If is provable, then , so it is true (because the statement about is vacuously true). Hence by Löb’s theorem, it’s provable, so we get .
If is provable, then it’s true, for the dual reason. So by Löb, it’s provable, so .
The broader point about being unable to reason yourself out of a bad decision if your prior for your own decisions doesn’t contain a “grain of truth” makes sense, but it’s not clear we can show that the agent in this example will definitely get stuck on the bad decision—if anything, the above argument seems to show that the system has to be inconsistent! If that’s true, I would guess that the source of this inconsistency is assuming the agent has sufficient reflective capacity to prove “If I can prove , then . Which would suggest learning the lesson that it’s hard for agents to reason about their own behaviour with logical consistency.
(There was a LaTeX error in my comment, which made it totally illegible. But I think you managed to resolve my confusion anyway).
I see! It’s not provable that Provable() implies . It seems like it should be provable, but the obvious argument relies on the assumption that, if * is provable, then it’s not also provable that - in other words, that the proof system is consistent! Which may be true, but is not provable.
The asymmetry between 5 and 10 is that, to choose 5, we only need a proof that 5 is optimal, but to choose 10, we need to not find a proof that 5 is optimal. Which seems easier than finding a proof that 10 is optimal, but is not provably easier.
“If you’ve never missed a flight, you spend too much time hanging around in airports” ~ “If you’ve never been publicly proven wrong, you don’t state your beliefs enough” ?
I added some clarification, but you are right.
(Since has the root , it’s clearly not true that all fifth-degree polynomials have this property)
This is great. A point which helped me understand number 6: If you ask someone “why do you believe X”, since you’re presumably going to update your probability of X upwards if they give a reason, you should update downwards if they don’t give a reason. But you probably already updated upwards as soon as they said “I believe X”, and there is no theorem which says this update has to be smaller than the latter update. So you can still end up with a higher or equal probability of X compared to where you were at the beginning of the conversation.
Could one approach to detecting biases be to look for “dominated strategies”? For instance, suppose the human model is observed making various trades, exchanging sets of tokens for other sets of tokens, and the objective of the machine is to infer “intrinsic values” for each type of token.
(Maybe conditional on certain factors, i.e “An A is valuable, but only if you have a B”, or “a C is only valuable on Tuesday”).
Then if the human trades an A and an E for a B, a B for a C, and a C for an A, but then trades an A for ten Es, we can infer that the human has some form of bias, maybe neglecting tokens with small value (not realizing that the value of an E matters until you have ten of them), or maybe an “eagerness” to make trades.
This clearly relies on some “Strong assumptions” (for instance, that tokens are only valuable in themselves—that executing a trade has no inherent value).
“If a tree falls in the woods, but no one is around to hear it, does it make a sound?” doesn’t sound like an argument, but a question. “Yes, because the presence of a person with ears doesn’t affect the physical behavior of the air” or “No, because air waves shouldn’t be considered sound until they interact with a mind” are arguments.
Or do you mean “argument” in the sense of a debate or discussion (as in “we’re having an argument about X”)?
If “such techniques usually give a boost for some time before dropping back towards baseline”, the obvious way to use this information would seem to be starting a new note-taking system every so often. That way you can keep on taking advantage of the boost, at least as long as you can keep finding new systems (which may eventually become a problem, but even so doesn’t leave you worse off than before). Of course, this does suggest a bound on how many resources you should invest in these new systems.
I’ve managed to implement this for computer monitors, but not for glasses. But my glasses seem to get smudged frequently enough that I need to wipe them about every day anyways. I guess I fidget with them much more than you?
I’ve noticed a sort of tradeoff in how I use planning/todo systems (having experimented with several such systems recently). This mainly applies to planning things with no immediate deadline, where it’s more about how to split a large amount of available time between a large number of tasks, rather than about remembering which things to do when. For instance, think of a personal reading list—there is no hurry to read any particular things on it, but you do want to be spending your reading time effectively.
On one extreme, I make a commitment to myself to do all the things on the list eventually. At first, this has the desired effect of making me get things done. But eventually, things that I don’t want to do start to accumulate. I procrastinate on these things by working on more attractive items on the list. This makes the list much less useful from a planning perspective, since it’s cluttered with a bunch of old things I no longer want to spend time on (which make me feel bad about not doing them whenever I’m looking at the list).
On the other extreme, I make no commitment like that, and remove things from the list whenever I feel like it. This avoids the problem of accumulating things I don’t want to do, but makes the list completely useless as a tool for getting me to do boring tasks.
I have a hard time balancing these issues. I’m currently trying an approach to my academic reading list where I keep a mostly unsorted list, and whenever I look at it to find something to read, I have to work on the top item, or remove it from the list. This is hardly ideal, but it mitigates the “stale items” problem, and still manages to provide some motivation, since it feels bad to take items off the list.
I’m curious about the remaining 3% of people in the 97% program, who apparently both managed to smuggle some booze into rehab, and then admitted this to the staff while they were checking out. Lizardman’s constant?
A simple remark: we don’t have access to all of , only up until the current time. So we have to make sure that we don’t get a degenerate pair which diverges wildly from the actual universe at some point in the future.
Maybe this is similar to the fact that we don’t want AIs to diverge from human values once we go off-distribution? But you’re definitely right that there’s a difference: we do want AIs to diverge from human behaviour (even in common situations).
Should there be an arrow going from n*f(n-1) to f (around n==0?) ? The output of the system also depends on n*f(n-1), not just on whether or not n is zero.
The example of associativity seems a little strange, I’m note sure what’s going on there. What are the three functions that are being composed?
This is a great list.
The main criticism I have is that this list overlaps way too much with my own internal list of high-quality sites, making it not very useful.
A thought about productivity systems/workflow optimization:
One principle of good design is “make the thing you want people to do, the easy thing to do”. However, this idea is susceptible to the following form of Goodhart: often a lot of the value in some desirable action comes from the things that make it difficult.
For instance, sometimes I decide to migrate some notes from one note-taking system to another. This is usually extremely useful, because it forces me to review the notes and think about how they relate to each other and to the new system. If I make this easier for myself by writing a script to do the work (as I have sometimes done), this important value is lost.
Or think about spaced repetition cards: You can save a ton of time by reusing cards made by other people covering the same material—but the mental work of breaking the material down into chunks that can go into the spaced-repetition system, which is usually very important, is lost.
There’s some recent work in the statistics literature exploring similar ideas. I don’t know if you’re aware of this, or if it’s really relevant to what you’re doing (I haven’t thought a lot about the comparisons yet), but here are some papers.
CFAR must have a lot of information about the efficacy of various rationality techniques and training methods (compared to any other org, at least). Is this information, or recommendations based on it, available somewhere? Say, as a list of techniques currently taught at CFAR—which are presumably the best ones in this sense. Or does one have to attend a workshop to find out?
Belief: There is no amount of computing power which would make AlphaGo Zero(AGZ) turn the world into computronium in order to make the best possible Go moves (even if we assume there is some strategy which would let the system achieve this, like manipulating humans with cleverly chosen Go moves).
My reasoning is that AGZ is trained by recursively approximating a Monte Carlo Tree Search guided by its current model (very rough explanation which is probably missing something important). And it seems the “attractor” in this system is “perfect Go play”, not “whatever Go play leads to better Go play in the future”. There is no way for a system like this to learn that humans exist, or that it’s running on a computer of a certain type, or even to conceptualize that certain moves may alter certain parameters of the system, because these things aren’t captured in the MCTS, only the rules of Go.
This isn’t an argument against dangerous AGI in general—I’m trying to clarify my thinking about the whole “Tool AI vs Agent AI” thing, before I read reframing superintelligence.
Am I right? And is this a sound argument?