Posts
Comments
Yes, I'm aware it's not universal. I can't really explain why it's so bothersome - it's similar to occasional words being in a bright colour, or someone poking me every so often while I'm trying to read. It's probably a pity because, combined with my laziness, it just means that I avoid reading writing with lots of italics. If I'm feeling particularly motivated I'll modify the text to remove all the italics before reading it.
I strongly agree. I find Eliezer's italics so off-putting that I avoid reading his writing in formatted text. I don't know why, and I'm sure not everyone has the same reaction, but excess italics just make me twitch.
I am not from Mudd, but I went to a talk by Maria Klawe on this very topic. I feel suddenly potentially useful. Warning: this is all only from memory. I thought I had the slides somewhere, but cannot find them. I'll email Maria, and if I hear back from her, I'll pass it on.
First off, here's the abstract for her talk:
Begin abstract
In 2006, much like at many other institutions, about 10% of HMC’s CS majors were female. At that time only a third of HMC’s students were female, but CS was an aberration. About 20% of the Physics majors and close to 30% of the engineering majors were female. Four years later 42% of HMC’s CS majors were female, exactly the same percentage as the whole HMC student body. This talk describes how the CS department accomplished this change.
End abstract
She emphasised that they tried to change how their first year program was run, with more cooperative group work. She was big on how they did a lot of work to try and get rid of the "macho" attitude among their undergrads in early courses - by "macho" she seemed to mean some sort of arrogant hackerish programmer attitude. She mentioned a bunch of mentorship programs for female undergrads, and programs to help undergrads get to conferences like Grace Hopper.
But! A couple friends and I were bothered by something else she said they did: they changed their undergrad admissions so as to admit more women to computer science in first year. Because they are a small, elite college they were able to do this without affecting the quality of their students, she felt. I thought that probably this was what made most of the difference, but that's only my opinion.
As an academicish person, I suggest a few questions that bothered me at first: Why aren't more artificial intelligence research groups at universities working on FAI? Why doesn't the Singularity Institute publish all of its literature reviews and other work?
I'm curious: if you're a person interested in "benevolence training", why do you want to have more benevolence or empathy for others? I generally want to be less empathetic, and I'd love to be convinced that I'm wrong.
If you had lots of end states, and lots of non-end states, and we want to assume the game ends when someone's won, and that a player only moves into an end state if he's won (neither of these last two are necessarily true even in nice pretty games), then you could treat it like a classification problem. In that case, you could throw your favourite classifier learning algorithm at it. I can't think of any publications on someone machine learning a winning condition, but that doesn't mean it's not out there.
Dr. David Silver used temporal difference learning to learn some important spatial patterns for Go play, using self-play. Self play is basically like watching yourself play lots of games with another copy of yourself, so I can imagine similar ideas being used to watching someone else play. If you're interested in that, I suggest http://www.aaai.org/Papers/IJCAI/2007/IJCAI07-170.pdf
On a sadly less published (and therefore mostly unreliable) but slightly more related note, we did have a project once in which we were trying to teach bots to play a Mortal Kombat style game only by observing logs of human play. We didn't tell one of the bots the goal, we just told it when someone had won, and who had won. It seemed to get along ok.
Your first paragraph rings true to me: the complaints I've heard are basically those you mentioned.
My friends are mostly fairly contrarian late-twenties male engineering, computing science and math people. I think that apart from not enjoying Methods, they're pretty much the usual LW demographic. That's part of the reason I was surprised when they didn't like Methods. There are lots of possible reasons for this (to me) surprising result. Maybe they thought I didn't like it, and wanted to mirror that back. Maybe they're a group already biased against LW. Maybe they actually just dislike the writing style. Who knows? If they don't enjoy Eliezer's writing style, then maybe LW is not a good place for them to hang out, so it doesn't matter that it didn't work as advertising on them.
Do you think that LW doesn't need other methods of marketing?
I've had mostly negative reactions to Methods of Rationality from 20-something males (and a few females) who are nerdy and geeky and mostly already like GEB, so I agree that this community needs other methods of marketing.