But surely you don’t think people will generally read all your work in the order you’d prefer, right? So now that you’ve realized some things about how status works, wouldn’t it be better to write in some way that doesn’t lead a sizable fraction of the people you wish were helping your cause to conclude that you’re a “colossal prick”?
Because when I talk to academics who might be good collaborators with MIRI, and they’ve read Eliezer’s stuff, they pretty often have had negative reactions to Eliezer’s writing. I try to re-explain MIRI’s work in less off-putting ways, but it’s hard to overcome initial impressions.
What percentage, would you say, of technical academics (who’ve read Eliezer’s writtings and conversed with you on the subject) have been turned off by it?
Wow. This really is a bubble. And here I was thinking the 3 sigma crowd might be less likely to be turned off by Eliezer. Do they state their reasons; if they do, could you list the compelling ones?
Probably all sorts of subtle things in the writing voice that people who trained to be scientists in academia learn by the time they get to PhD, and they can tell if those are missing and get the subconscious crackpot signal. Same as what happens every time when an outsider tries to influence people in a subculture and keeps getting the subtle subculture communication patterns wrong.
Might be easier if EY went in saying “here’s some science I did, take a look” rather than “you’re doing it all wrong, you should be working on this stuff like this instead”.
I have emailed roughly a dozen phds I know from grad school and work. I’ve asked if they remember what pushed them away, and if they don’t to take a second look. Hopefully some of them will register to give their own responses (and so you can converse with them), but for the ones who don’t register but email me, I’ll add them here myself.
My own response- I’m a physics phd, and for the first example that springs to mind, I find the quantum physics sequence very off putting. It opens by saying quantum physics isn’t mysterious and by the time it gets to the born probabilities, it admits that quantum physics is very much mysterious. I’m sure you can find previous posts where I’ve argued against many worlds as the one true interpretation.
First outsider response back;
James R. physics phd- “I’m pretty familiar with LessWrong, and have previously tried to read the first sequence. It seemed a lot like what you’d find on any atheist blog (a bit on evidence and sagan’s dragon, etc). I stopped when I got to a discussion of emergence that was appallingly ignorant. It was especially vexing given that the author, while deriding the idea, clearly believes intelligence is an emergent phenomenon, not dependent on the underlying neurons, or else he wouldn’t advocate you can reproduce it in silicon. Especially ironic given that a the whole next sequence (I didn’t read it) seems to be about the fact that words mean things. I won’t bother reading an author who won’t do the bare minimum of due diligence to investigate a term before writing a whole blog post about it.”
It seems a bit sad to entirely stop posting things to LessWrong, but I suppose that if only Facebook routes things to people who will want to read it, I should post any possibly-offensive or controversial material to only Facebook.
I was thinking more that you could try your hand at a writing style that communicates the same stuff but doesn’t annoy so many people. E.g. Bostrom writes about the same topics but seems to annoy people less often.
The primary failure mode of writing is that nobody reads it. I don’t know how to write like Bostrom in a way that people will read. I’m already worried about things like the tiling agents paper dropping off the radar.
Lots of people read Bostrom. And he gets listed in the FP 100 Global Thinkers list. And his works are widely translated. And he’s done hundreds of interviews in popular media. Lots of people read Robin Hanson, too.
I’m not saying you should drop all current projects to learn this additional writing skill of being fun to read while also not pissing people off, I’m just saying that I think the lesson to be drawn from lots of smart people being annoyed by your tone is a bit deeper than “Just don’t use this article as an introduction to LW.”
The steps you could take to avoid the nobody-reads-it failure mode seem to me to be orthogonal to the steps you could take to avoid the author-is-a-colossal-prick failure mode. Given that you started this whole damn web site and community as insurance against the possibility that there might be someone out there with more innate talent for FAI, lukeprog’s suggestion that you take steps to mitigate the author-is-a-colossal-prick failure mode in furtherance of that mission seems like a pretty small ask to me. And I say this as one who has always enjoyed your writing.
I’ve had the experience of finding your writing very annoying, but then coming around. On the basis of this, I have a suggestion: I don’t think you’re going to get the signaling thing right. Bostrom is good at that because he’s an academic and that involves years and years of signal training, and heavy selection on that basis.
I ceased to be annoyed by your writing when it occurred to me that no one in the world, nor all of us together, could subject you to a greater hell of ridicule than you will if you don’t make some significant progress on this FAI thing in your lifetime. My urge to adjust your sense of status evaporated when I realized I don’t need to put a sword over your head, because the sword you put there is bigger than any I could come up with.
Your rhetorical and personal assets are sincerity and passion. These are undermined by glibness and irony, and appeals to status. So I suggest avoiding these things, rather than working on a more professional style.
It’s interesting to note that all this would be largely mediated if Eliezer hadn’t made himself a character in the story. A case for land value tax, easily moveable housing, the value of rationality, his belief in the incompetence of this dimension– all this could have done in a story that doesn’t star Eliezer– and the reader wouldn’t, then, have to deal with the dissonance between the status it looks like Eliezer assigns himself (that of a player from a higher league slumming here) and that which the naive reader of the piece assigns him; so long as the character is not an obvious stand in.
As of now the story is a super stimulus for the improper-status-assignment emotion he recently discovered—which I suppose could be the true prank.
There’s something to learn in Feynman’s writings. Surely You’re Joking, Mr. Feynman! is a book about how a highly-intelligent man gets by in a world full of fools. And yet the readers never feel like the author thinks they are fools, even though from Feynman’s perspective they almost certainly are.
But surely you don’t think people will generally read all your work in the order you’d prefer, right? So now that you’ve realized some things about how status works, wouldn’t it be better to write in some way that doesn’t lead a sizable fraction of the people you wish were helping your cause to conclude that you’re a “colossal prick”?
Why do you believe there is “a sizable fraction” of such people? The above mentioned anecdote is really the only instance that comes to (my) mind.
Because when I talk to academics who might be good collaborators with MIRI, and they’ve read Eliezer’s stuff, they pretty often have had negative reactions to Eliezer’s writing. I try to re-explain MIRI’s work in less off-putting ways, but it’s hard to overcome initial impressions.
What percentage, would you say, of technical academics (who’ve read Eliezer’s writtings and conversed with you on the subject) have been turned off by it?
So in my limited experience with physics and statistics phds I’ve sent here, its north of 80%.
Wow. This really is a bubble. And here I was thinking the 3 sigma crowd might be less likely to be turned off by Eliezer. Do they state their reasons; if they do, could you list the compelling ones?
Probably all sorts of subtle things in the writing voice that people who trained to be scientists in academia learn by the time they get to PhD, and they can tell if those are missing and get the subconscious crackpot signal. Same as what happens every time when an outsider tries to influence people in a subculture and keeps getting the subtle subculture communication patterns wrong.
Might be easier if EY went in saying “here’s some science I did, take a look” rather than “you’re doing it all wrong, you should be working on this stuff like this instead”.
I have emailed roughly a dozen phds I know from grad school and work. I’ve asked if they remember what pushed them away, and if they don’t to take a second look. Hopefully some of them will register to give their own responses (and so you can converse with them), but for the ones who don’t register but email me, I’ll add them here myself.
My own response- I’m a physics phd, and for the first example that springs to mind, I find the quantum physics sequence very off putting. It opens by saying quantum physics isn’t mysterious and by the time it gets to the born probabilities, it admits that quantum physics is very much mysterious. I’m sure you can find previous posts where I’ve argued against many worlds as the one true interpretation.
First outsider response back; James R. physics phd- “I’m pretty familiar with LessWrong, and have previously tried to read the first sequence. It seemed a lot like what you’d find on any atheist blog (a bit on evidence and sagan’s dragon, etc). I stopped when I got to a discussion of emergence that was appallingly ignorant. It was especially vexing given that the author, while deriding the idea, clearly believes intelligence is an emergent phenomenon, not dependent on the underlying neurons, or else he wouldn’t advocate you can reproduce it in silicon. Especially ironic given that a the whole next sequence (I didn’t read it) seems to be about the fact that words mean things. I won’t bother reading an author who won’t do the bare minimum of due diligence to investigate a term before writing a whole blog post about it.”
Ouch. (The statistics PhD thing is saddening to me but not surprising, alas.)
It seems a bit sad to entirely stop posting things to LessWrong, but I suppose that if only Facebook routes things to people who will want to read it, I should post any possibly-offensive or controversial material to only Facebook.
I was thinking more that you could try your hand at a writing style that communicates the same stuff but doesn’t annoy so many people. E.g. Bostrom writes about the same topics but seems to annoy people less often.
The primary failure mode of writing is that nobody reads it. I don’t know how to write like Bostrom in a way that people will read. I’m already worried about things like the tiling agents paper dropping off the radar.
Lots of people read Bostrom. And he gets listed in the FP 100 Global Thinkers list. And his works are widely translated. And he’s done hundreds of interviews in popular media. Lots of people read Robin Hanson, too.
I’m not saying you should drop all current projects to learn this additional writing skill of being fun to read while also not pissing people off, I’m just saying that I think the lesson to be drawn from lots of smart people being annoyed by your tone is a bit deeper than “Just don’t use this article as an introduction to LW.”
I don’t know how to do that. I don’t know how to learn to do it either.
The steps you could take to avoid the nobody-reads-it failure mode seem to me to be orthogonal to the steps you could take to avoid the author-is-a-colossal-prick failure mode. Given that you started this whole damn web site and community as insurance against the possibility that there might be someone out there with more innate talent for FAI, lukeprog’s suggestion that you take steps to mitigate the author-is-a-colossal-prick failure mode in furtherance of that mission seems like a pretty small ask to me. And I say this as one who has always enjoyed your writing.
Do you know how to learn to do this?
I would miss his current writing style.
If you’ll forgive unasked for advice:
I’ve had the experience of finding your writing very annoying, but then coming around. On the basis of this, I have a suggestion: I don’t think you’re going to get the signaling thing right. Bostrom is good at that because he’s an academic and that involves years and years of signal training, and heavy selection on that basis.
I ceased to be annoyed by your writing when it occurred to me that no one in the world, nor all of us together, could subject you to a greater hell of ridicule than you will if you don’t make some significant progress on this FAI thing in your lifetime. My urge to adjust your sense of status evaporated when I realized I don’t need to put a sword over your head, because the sword you put there is bigger than any I could come up with.
Your rhetorical and personal assets are sincerity and passion. These are undermined by glibness and irony, and appeals to status. So I suggest avoiding these things, rather than working on a more professional style.
It’s interesting to note that all this would be largely mediated if Eliezer hadn’t made himself a character in the story. A case for land value tax, easily moveable housing, the value of rationality, his belief in the incompetence of this dimension– all this could have done in a story that doesn’t star Eliezer– and the reader wouldn’t, then, have to deal with the dissonance between the status it looks like Eliezer assigns himself (that of a player from a higher league slumming here) and that which the naive reader of the piece assigns him; so long as the character is not an obvious stand in.
As of now the story is a super stimulus for the improper-status-assignment emotion he recently discovered—which I suppose could be the true prank.
There’s something to learn in Feynman’s writings. Surely You’re Joking, Mr. Feynman! is a book about how a highly-intelligent man gets by in a world full of fools. And yet the readers never feel like the author thinks they are fools, even though from Feynman’s perspective they almost certainly are.