= 27d567bc2d48d9f40e9ccc20ea370b15)
DeepMind isn't doing safety engineering; they're doing standard AI. It doesn't matter if Elon Musk is interested in AI safety, if, after his deliberations, he invests in efforts to develop unsafe AI. Good intentions don't leak value into the consequences of your acts.
In response to
Rationality Quotes February 2014
You are never going to catch up, and neither is anyone else.
-- Gian-Carlo Rota
In response to
Tricky Bets and Truth-Tracking Fields
Bob will accept that some phrase X is meaningful if there is a test that can be performed whose outcome value depends on truth value of X. If there is such a test, then we can construct a further test of asking someone who has performed the original test what the outcome of the test was. Since the people who set up tests are usually honest, this test would also be a test of X (provided the original test exists).
If I ask an honest peasant how long the emperor's nose is, but I also suspect no one has ever seen the emperor, how much do I learn from her statement? What if she says, "I have never seen the emperor, but other people tell me his nose is 5cm"? How many people has she talked to? Has any of them seen the emperor?
I don't know how to answer those questions, and yet your example is even less clear. You think no one has seen the emperor and you're not sure if he can be seen. 5cm? Well she is honest.
I would phrase this more along the lines of "If nothing MIRI does works, or for that matter if everything works but it's still not enough, CFAR tries to get a fully generic bonus on paths unseen in advance."
Do you choose that rephrasing because you don't see how MIRI's work could be harmful or because there is nothing CFAR can do in that case?
In response to
Above-Average AI Scientists
Mirror of the Bonobo Conspiracy webcomic: #569: Easy once you know
Nope!
Richard Jeffrey!
Karma points to whoever identifies them correctly!
The guy on the right is more accurately represented than the guy on the left, but the guy on the left may still be guessable based on the subject matter. Hint: the guy on the left is a philosopher known, among other things, for his post-1950 contributions to theory of causality.
(The guy in the middle is Judea Pearl, Ilya's advisor at UCLA.)
Update: dougclow correctly identified the man on the right, and vallinder correctly identified the man on the left.
The man on the left is Hans Reichenbach.
In response to
comment
by
on
Harry Potter and the Methods of Rationality discussion thread, part 26, chapter 97
Conscious control over social presentation is a learned skill; it doesn't come in the same box with intelligence or rationality, although either or both might make it easier to pick up. I suspect it's prominent in Methods mainly because it serves the particular type of wheels-within-wheels plotting that Eliezer seems fond of.
We could have a conversation at this point about whether constructed social presentation is unethical or "creepy", but I don't think it'd get us anywhere. Some people have the squick response, some don't.
(Incidentally, I don't feel like Snape's got this in-story. He's certainly got a facade, but it's the sort you build semi-involuntarily when you hate parts of yourself and desperately want to hide them, not the kind you consciously build to optimize social outcomes. Harry does have it, but shouldn't have had the opportunity to develop it; it may be part of his Mysterious Dark Side/possible Harrymort package.)
In response to
comment
by
on
Harry Potter and the Methods of Rationality discussion thread, part 26, chapter 97
For Snape, I was specifically thinking of the scene in Dumbledore's office where Harry reveals that he knows about the prophecy and Snape reacts without hesitation as though he hadn't heard of it. Snape was also a double agent during the war, and continues to maintain close relationships with Dumbledore and Lucius Malfoy. His actions do seem crude, awkward, uncontrolled or mostly defensive in other scenes such as in the bullying arc or his conversation with Quirrell in the forbidden forest in Chapter 77. But then, one can act with false impulsiveness too.
I suppose the characters are in a cold war and in the shadow of a hot war. That circumstance makes "offensive" deception in one's social presentation more useful.
Quirrell, Dumbledore, Snape, Harry, and (increasingly) Draco have something in common. They are all creepy. These characters are intentionally inauthentic - acting as though they posses the specific beliefs, preferences, and abilities that they want others to attribute to them.
I feel unusually strong revulsion about this kind of deception - more than toward someone hiding their faults to manage their appearance, much more than toward someone being tactful and withholding or biasing sensitive claims to avoid conflict.
When I try to unpack "creepy", my mind suggests it has components of outrage at violations of close interpersonal social norms, distrust of unfamiliar thought patterns, fear of people with motivations that need to be hidden, and a special kind of disgust related to fears of idols, photographs, glassy eyed dolls, humanoid robots, and other simulacra. - the disgust toward an exemplar that doesn't fall clearly in or out of the human-mind category, toward a soul that has been captured in the depiction of a face and deprived of its intelligence and agency.
Are very intelligent people generally creepy like that? If I were a standard deviation smarter, would my peer group consist of people strategically concealing their identities and mutually modelling their mutual modelling up to the nth order of meta? Or is that inauthenticity just an abnormal personality type that doesn't correlate much with intelligence, but does fit nicely into a rationalist literary drama?
View more: Next
= b715d02e85f7b3b4ae65ca3d3e450868)
Subscribe to RSS Feed
= f037147d6e6c911a85753b9abdedda8d){kind=logo}
General case:
When someone posts links from webpage X, which can be refuted from webpage Y (or vice versa), and so on, without adding anything themselves to the discussion.
Motivating example:
I've often seen things posted on climate change, lifted directly from http://wattsupwiththat.com/ , that can be refuted from http://www.skepticalscience.com/ , which can often be re-refuted from the original website, and so on. Since we're just letting the websites talk to each other, and neither poster has any relevant expertise, this seems a pointless waste.
An argument that halts in disagreement (or fails to halt in agreement) because the interlocutors are each waiting for another to provide a skillful assessment of their own inexpertly-referenced media sounds a lot like a software process deadlock condition in computer science. Maybe there's a more specific type of deadlock, livelock, resource starvation, ..., in the semantic neighborhood of your identified pattern.
Dropping references, while failing to disclaim your ability to evaluate the quality and relevance of topical media, could be called a violation of pragmatic expectations of rational discourse, like Grice's prescriptive maxims.
Maybe a telecommunications analogy would work, making reference to amplifiers \ repeaters \ broadcast stations that degrade a received signal if they fail to filter \ shape it to the characteristics of the retransmission channel.
"Rhetorical reenactment" sounds like "historical reenactment" and hints at the unproductive, not-directly-participatory role in the debate of the people sharing links.