Cyan2

Karma: 375

Cyan2 Jun 29, 2008, 4:38 AM
0 points
on: [Error communicating with LW2 server]
In reply to Q1, I’d want to introduce new terminology like “implicit understanding” and “explicit understanding” (paralleling the use of that terminology in reference to memory). I’d say Sarah had explicit understanding (akin to a good student’s understanding of biology from a lecture course), but not implicit understanding (akin to a biology laboratory researcher’s understanding).

In reply to Q2, it really depends on what you (generic “you”) are trying to accomplish. It’s totally dependent on the person and the situation—there’s no one-size-fits-all answer here. Total cop-out, I know.

Cyan2 Jun 29, 2008, 1:44 AM
1 point
on: [Error communicating with LW2 server]
Angel: “If you’d like an example of something that offends me, putting words in my mouth is a good place to start.”

Eliezer: “Angel, I’m not arguing that you so generalized—let us both be careful not to put words into the other’s mouth.”

I believe this confusion is over the referent of “we” and “us” in the Eliezer’s statement, “So it is not that the truth is laid out plainly, and you see it, but we are blind. That is treating us as defective versions of yourself.” Eliezer meant Robin and himself; Angel read it to refer to all men.

Cyan2 Jun 28, 2008, 2:30 AM
0 points
on: 2-Place and 1-Place Words
A good compromise might be for the site editors to ruthlessly deport sufficiently tangential threads to the forum (the existence of which I was recently reminded). Although that’s easy for me to suggest: I’m not a site editor, so there’s no extra work for me.

Cyan2 Jun 26, 2008, 3:17 PM
7 points
on: No Universally Compelling Arguments
What’s the difference between calling Bayesian reasoning an “engine of accuracy” because of its information-theoretic properties as you’ve done in the past and saying that any argument based on it ought to be universally compelling?

Bayesian reasoning is an “engine of accuracy” in the same why that classical logic is an engine of accuracy. Both are conditional on accepting some initial state of information. In classical logic, conclusions follow from premises; in Bayesian reasoning, posterior probability assignments follow from prior probability assignments. An argument in classical logic need not be universally compelling: you can always deny the premises. Likewise, Bayesian reasoning doesn’t tell you which prior probabilities to adopt.

Cyan2 Jun 25, 2008, 6:58 PM
0 points
on: The Design Space of Minds-In-General
Unknown, okay, I see it now. Thanks.

The word “the” is something every English document has in common.

… except for Gadsby, A Void, and other lipograms. ;-)

Cyan2 Jun 25, 2008, 5:18 PM
0 points
on: The Design Space of Minds-In-General
Phil, I’m not really sure what your criticism has to do with what Eliezer wrote. He’s saying that evolution is contingent—bits that work can get locked into place because other bits rely on them. Eliezer asserts that AI design is not contingent in this manner, so the space of possible AI designs does not form a natural class, unlike the space of realized Earth-based lifeforms. Your objection is… what, precisely?

Cyan2 Jun 23, 2008, 5:08 AM
0 points
on: Surface Analogies and Deep Causes
Unknown, MDL works for me. (I think your 100-property example works better in terms of MDL than your 2-property example.)

It seems to me that the justification for taking an outside view is some kind of probabilistic exchangeability. In fact, exchangeability seems to me to be the key concept grounding an outside view, and I’m surprised no one else has brought that concept up.

When the events being predicted are rare and complicated, I have a hard time seeing an outside view as justified. I don’t see any justification for an inside view, either.

Cyan2 Jun 23, 2008, 1:10 AM
0 points
on: Surface Analogies and Deep Causes
Yeesh, that last comment of mine was poorly written. Let me try again.

It’s highly debatable whether it matters which measure of simplicity is in use. It definitely matters that the one you use is not precisely specified. As it is, I have nothing more to go on than, “Unknown said so”—and I can’t generalize to see if its use makes sense (satisfies some optimality criterion) beyond the limited case you’re arguing for.

Cyan2 Jun 22, 2008, 10:23 PM
0 points
on: Surface Analogies and Deep Causes
I doubt it particularly matters which precise measure of simplicity I use, probably any reasonable measure will do.

You’re too easy on yourself.

It’s highly debatable whether it matters which measure of simplicity is in use, and it definitely matters that the one you use is precisely specified. Otherwise, I have nothing more to go on than, “Unknown said so”—and I can’t generalize to see if its use makes sense (satisfies some optimality criterion) beyond the limited case you’re arguing for.

Cyan2 Jun 22, 2008, 7:14 PM
1 point
on: Surface Analogies and Deep Causes
Unknown, what is the function you are using that takes pairs of logical statements as arguments and outputs their simplicity ordering?

Cyan2 Jun 20, 2008, 10:58 PM
1 point
on: How An Algorithm Feels From Inside
Silas,

The essential idea is that network 1 can be trained on a target pattern, and after training, it will converge to the target when initialized with a partial or distorted version of the target. Wikipedia’s article on Hopfield networks has more.

Both types of networks can be used to predict observables given other observables. Network 1, being totally connected, is slower than network 2. But network 2 has a node which corresponds to no observable thing. It can leave one with the feeling that some question has not been completely answered even though all the observables have known states.

Cyan2 Jun 20, 2008, 10:58 PM
0 points
on: Heading Toward Morality
Silas, I’ve replied on the comment thread of “How An Algorithm Feels From Inside”.

Cyan2 Jun 20, 2008, 3:38 PM
1 point
on: Possibility and Could-ness
Joseph Knecht, where you go wrong is here:

...when we… ask what it means for a state to be reachable, the answer circularly depends on the concept of possibility, which is what we are supposedly explaining. A reachable state is just a state that it is possible to reach.

A state labelled “reachable” by a human may in fact be impossible as a matter of physics. Many accidental injuries and deaths follow from this fact. It is as a result of evolution that a given human’s label “reachable” has any relation to reality at all.

Cyan2 Jun 20, 2008, 4:57 AM
0 points
on: Possibility and Could-ness
Joseph Knecht, you call it a quibble, but it seems to me that it is the whole key to dissolving the question of free will and could-ness. If Eliezer really did subtly violate the Taboo in a way I didn’t notice, then what I have in my head is actually a mysterious answer to a mysterious question; I’m at a loss to reconcile my feeling of “question dissolved!” with your assertion that the Taboo was violated.

Cyan2 Jun 20, 2008, 1:38 AM
1 point
on: Possibility and Could-ness
Joseph Knecht, Brandon put “realizability” in the territory, whereas Eliezer put “reachability” in the map. Compare and contrast:

Brandon: When we speak of “could” we speak of the set of realizable worlds [A’] that follows from an initial starting world A operated on by a set of physical laws f.

Eliezer: We’ve previously discussed how probability is in the mind… possibility is also in the mind.

It is not the case that Eliezer is criticizing Brandon for the same mistake that Eliezer made himself.

Cyan2 Jun 18, 2008, 9:10 PM
0 points
on: The Ultimate Source
Phillip Huggan, the wikipedia article on tachyons answers your question. Extremely short version (granting tachyons exist): for a tachyon, there is no distinction between the processes of emission and absorption. The attempt to detect a tachyon from the future (and violate causality) would actually create the same tachyon and send it forward in time.

Cyan2 17 Jun 2008 13:52 UTC
10 points
on: Grasping Slippery Things
Alexandre passos, Unknown, (and Caledonian too, from a previous thread,)

Eliezer has already stated that he’s taking a deterministic many worlds interpretation of reality as a premise (and explained at some length why he does so in the QM series). If you disagree with that premise, of course the conclusions do not necessarily follow.

I’m not defending the assumption of determinism—but I am saying that a criticism of the argument that flows from Eliezer’s premises would be more apposite and interesting than essentially posting over and over again, “Nuh uh! What if the universe isn’t deterministic, huh?”

I’m sorry, I’m probably just cranky in the morning. I’ll go drink some coffee and then start regretting posting this.

Cyan2 16 Jun 2008 3:08 UTC
4 points
on: The Ultimate Source
But human minds are too complex for us to do this—so we attribute its operations, which we cannot comprehend, to chance. Ignorance is the key.

I’d agree that ignorance is the key, but not that people in general attribute the operations of the human mind to chance. Rather, it seems we attribute these operations to a noumenal essence that is somehow neither deterministic nor random. We make this attribution so readily (and not just to other human minds) because that is our internal experience—the feeling of our algorithm from the inside.

Cyan2 16 Jun 2008 2:51 UTC
0 points
on: The Ultimate Source
No. [<-- Useless single-line flat contradiction that can be deleted without affecting any of the actual arguments. EY.]

Caledonian,

“Wrong. [etc.]”

would probably have gotten the same edit, but I suspect

“On the contrary, [etc.]” “I disagree. [etc.]” “I see a flaw in your reasoning: [etc.]”

would have passed without comment. If Eliezer’s edits are bothering you, you might consider trying these or other formulations for your thesis statement.

Cyan2 14 Jun 2008 20:09 UTC
0 points
on: Causality and Moral Responsibility
Surely the place to look for the effect (if any) of moral philosophy on the real world is in jurisprudence?