This comes off very strongly as the typical bureaucratic protectiveness—a business doesn’t want to share raw data, because raw data is a valuable resource. If you came out and said this was the reason, I’d be more understanding, but it would still feel like a major violation of community norms to be so secretive.
If simple secrecy is indeed the case, I would urge you, please, be honest about this motive and say so explicitly! At least then we are having an honest discussion, and the rest of this comment can be disregarded.
We collected lots of data before and during minicamp
In short, what is the reason you can’t share this RAW data, which you state you collected, and which you’ve presumably found sufficient for your own preliminary conclusions? I don’t think Silas is asking for or expecting an elegant power-point presentation or a concise statistical analysis—I know I would personally love to simply see raw data.
Is there truly not a single spreadsheet or writeup that you could drop up for us to study while you collect the rest of the data?
Good grief, people. There are conspiracies that need ferreting out, but they do not revolve around generating fake data about the effectiveness of an alpha version of a rationality training camp that was offered for free to a grateful public.
I went to the minicamp, I had a great time, I learned a lot, and I saw shedloads of anecdotal evidence that the teachers are striving to become as effective as possible. I’m sure they will publish their data if and when they have something to say.
Meanwhile, consider re-directing your laudable passion for transparency toward a publicly traded company or a medium-sized city or a research university. Fighting conspiracies is an inherently high-risk activity, both because you might be wrong about the conspiracies’ existence, and because even if the conspiracy exists, you might be defeated by its shadowy and awful powers. Try to make sure the risks you run are justified by an even bigger payoff at the end of the tunnel.
There are conspiracies that need ferreting out, but they do not revolve around generating fake data about the effectiveness of an alpha version of a rationality training camp that was offered for free to a grateful public.
I don’t think anybody is accusing the minicamp folks of anything of the kind. But public criticism and analysis of conclusions is the only reliable way to defend against overconfidence and wishful thinking.
When I ended my term as an SIAI Visiting Fellow, I too felt like the experience would really change my life. In reality, most of the effects faded away within some months, though a number of factors combined to permanently increase my average long-term happiness level.
Back then the rationality exercises were still being worked out and Luke wasn’t around, so it’s very plausible that the minicamp is a lot more effective than the Visiting Fellow program was for me. But the prior for any given self-help program having a permanent effect is small, even if participants give glowing self-reports at first, so deep skepticism is warranted. No conspiracies are necessary, just standard wishful thinking biases.
Though I think this was the third time that Silas raised the question before finally getting a reply, despite his comment being highly upvoted each time. If some people are harboring suspicions of SIAI covering up information, well, I can’t really say I’d blame them after that.
I find that unlikely. That would mean you never followed up after trumpeting your success—you just posted the topic, and never bothered to come back and see what people had to say. And that you didn’t see the top comment on the 125k fundraiser thread. Then again, this is consistent with what komponisto said about your “Mt Olympus” mentality: just say stuff ex cathedra and expect everyone to fall in line or otherwise swoon.
I don’t understand this bit about my ‘My Olympus’ mentality. Until very recently I wasn’t on SI’s full-time staff. And as far as I can tell, I’ve spent vastly more time substantiating what I say by citing the relevant scientific literature (rather than relying on whatever personal authority I’m supposed to have, which I don’t think is much at all) than anyone else on Less Wrong.
And no, I don’t expect everyone to “fall in line or otherwise swoon.” It’s just that I don’t have time to write up a 20-citation research article supporting every sentence I write. If the reasons that led me to write a certain sentence aren’t available to you, as is usually the case, then you should only be updating your beliefs as much as you should given the evidence of my testimony, which in many cases should be very little.
As for you not believing me when I say that I don’t recall reading your earlier comments calling for evidence about minicamp’s success, well… the only evidence I have for you besides my testimony is that I hadn’t replied to any of your earlier comments on the matter. If you don’t believe me, well, so be it: that’s all I’ve got.
I don’t understand this bit about my ‘My Olympus’ mentality. Until very recently I wasn’t on SI’s full-time staff. And as far as I can tell, I’ve spent vastly more time substantiating what I say by citing the relevant scientific literature (rather than relying on whatever personal authority I’m supposed to have, which I don’t think is much at all) than anyone else on Less Wrong.
Indeed you have, and you’ve been well rewarded, with 24,000+ karma points and a full-time position at SI (with EY himself begging for money to pay you in a promoted LW post—I’ll bet that felt good!). What you haven’t earned, however, is the right to ignore people without their being offended. (The only person who might conceivably have that level of status is EY, and I think even that is debatable.)
The impression I think you give is one of writing all this great LW material, but then being “too busy” with your high-status SI work to read people’s comments on it. Surely you can see how that comes across.
I reply to many comments but certainly not all. I can’t respond to all my critics, and it’s probably unwise to do so. I’m also sympathetic to thomblake’s comment on this discussion:
I appreciate your efforts to decode this ‘Olympus mentality’ nonsense, and in general to make sure you’re not making communication errors. But at this point I believe you’re just wasting your time. You’ve documented your research methods better than I’ve ever seen someone do, and they certainly don’t need defending here.
On behalf of those who believe your work can positively impact the future of humanity and your time can be better spent elsewhere, I humbly request that you please file what you’ve been responding to under ‘trolling’ and move on.
To be clear, my comment pertained to reading others’ comments, not necessarily replying to them.
I would also like to stress that, while I am not sympathetic to thomblake’s comment (I could hardly be expected to be, since he effectively labeled a comment of mine “nonsense”), I have not voiced any complaint about the underlying subject (proof of minicamp effectiveness), and should not be confused with Silas. (This isn’t to say that I don’t think he has a point also; but I emphatically do not consider myself to have participated in any “trolling”.)
I entered the discussion because of the “Olympus” issue, which I had noticed in other contexts and had considered bringing up before. (Evidently I am not the only one, because my comment on the matter has—against all expectations of mine—been voted up into the 20s.)
“The impression I think you give is one of writing all this great LW material, but then being “too busy” with your high-status SI work to read people’s comments on it. Surely you can see how that comes across.”
LOL. If one does form that impression from Luke’s posts I would suggest that you:
1.) Don’t be so sensitive and get a life beyond obssessing over blog comments.
2.) Realize that people who actually get things done usually don’t make responding to blog posts a high or even medium priority in life. Seriously, most people with productive lives don’t have time to obssess over blogs.
Appreciate the good content, criticize the points worth criticizing and try not to cry about the minor points and the fact that someone doesn’t respond.
As for you not believing me when I say that I don’t recall reading your earlier comments calling for evidence about minicamp’s success, well… the only evidence I have for you besides my testimony is that I hadn’t replied to any of your earlier comments on the matter.
My point was that either a) you didn’t do what most people do when they start topics—look at them a second time—or b) you’re lying. Neither is likely, but a) is at least consistent with the Mt Olympus mentality—why else would you never follow up on a thread with a major announcement? What other mentality would lead to that behavior?
My request for evidence was also (for most of the time it was up) the top comment on the 125k fundraising thread, which does give someone reason to be skeptical you weren’t aware anyone had asked for any substantiation.
It’s just that I don’t have time to write up a 20-citation research article supporting every sentence I write.
Sure, that should be resolved for the more improbable claims.
And in the case where your justifications come from the literature, I thought you could just google the term, read the abstract, and copy-paste the citation?
Being busy with research and writing? Seriously, I don’t spend all day re-checking the comment threads on old LW posts. If you want to reach me, please contact me directly or reply to one of my comments, so it shows up in my LW inbox.
You were aware he was replying to a strawman and that I never expected that he should “spend all day rechecking …”, right? (Guess not, and ditto for those who voted you up.)
I can see the general point you were making with your criticisms and so could the other people who voted you up on your early comments on this issue. That being said I suggest you quit while you are (or were) ahead. It is too easy to let others reclaim the moral high ground if you stay on the same topic for too long and let them put you on the defensive.
This would allow you to maintain context specific credibility for use the next time the same, or similar claims were made about success that you don’t believe are justified.
Would I be unreasonable or unrealistic if I expressed a desire to not see this any of this SIAI inside baseball on Less Wrong… ever? Insofar as these rationality minicamps are something we think people who want to be more rational should take part it, obviously data on their effectiveness is very important. But insofar as the ‘success’ of the minicamp figured into SIAI’s decision to hire lukeprog (and that seems to be the issue for the moment) I could care less. I realize of course Less Wrong and SIAI are intimately connected and I’m personally at the low end on a spectrum of interest and involvement in SIAI. And I’m fine with seeing the occasional fund-raising post or strategy discussions in Discussion… after all I’m not paying for the pleasure of posting and reading here. But an ongoing flame war about a random criteria in an SIAI hiring dominates the recent comments section and is of no interest to me (and one assumes, others in my position). In the interest of keeping SIAI and Less Wrong somewhat separate shouldn’t SIAI have some other avenue donors can use to voice concerns and criticisms so that it doesn’t interfere with the interesting content here?
(And if Silas isn’t a donor… Thomas %&$@! Bayes why does anyone care?!)
Would I be unreasonable or unrealistic if I expressed a desire to not see this any of this SIAI inside baseball on Less Wrong… ever?
I don’t care for flame wars either. But what I do care about is that if it is permitted to make a declaration of fact on lesswrong it is permitted to to refute it. The details of what you suggest in the parent violate this. You advocating the lesswrong equivalent of true evil!
I just don’t feel like I have the standing to demand that SIAI fund-raising letters be left off the main page entirely- so I was looking for a compromise.
Downvotes imply net disapproval, particularly for someone whose comments get read as much as Eliezer’s. If you think of it as simply losing points, it seems trivial, but if you take it as a sign that “people seem not to like it when I do that,” it’s a meaningful consideration.
This is a pretty sad day for LW, to learn that you can just lie and strawman your way out of criticism, because whoever calls you on it is just “staying on the same topic for too long” :-/
(I hope you’ll pardon the digression into crude discussion of rational strategies for influence. This is a far more interesting topic than what serves for the object level at this point!)
This is a pretty sad day for LW, to learn that you can just lie and strawman your way out of criticism, because whoever calls you on it is just “staying on the same topic for too long” :-/
Speak more strategically. Don’t let the verbal signals you utter be tools you use to salve and release your own feelings. Not because it is virtuous, purely because that doesn’t get you what you want. Also note that not only am I someone who consistently voted you up and those insulting you down, my comment provided stronger support for your position than your most recent comments managed. Allow me to translate what I said into ‘fun’ rather than vaguely polite:
“You’re so right man. They’re full of shit. I mean WTF is with this claiming stuff with no evidence then ragging on you instead of answering you. That’s fucking pathetic. Ape Status Bullshit 101 - If someone asks critical questions don’t answer them, beat them with a fucking stick so nobody else thinks it’s ok to dissent. That’s what people with status and power do and people always let them get away with it. But here’s the thing: You’re making it easy for them by being a whiny little bitch. How’s that working out for you? Not working, huh? Yeah, no shit. What did you expect? Now stop crying ‘cos the world isn’t fair to you and start saying shit that works. Also, accept that you cannot change other people and instead work out what influence you can have and make it. In this case it would be ‘make all unjustified claims a net PR loss for Luke/SIAI by calling them on it effectively and moderately.’ And I’d have called that a real success. Heck, it even worked. Look at Anna’s replies from a few days ago. Now man the fuck up and stop being a pussy. Because I often agree with the complaints that you have about stuff but don’t want to look bad by association.”
Now, consider the difference between the above wording and what I actually said. Notice that it positions me as somewhat of an ally, assumes the criticisms you make of Luke are valid but at the same time doesn’t try to alienate me from the tribe. See why I chose to use the wording I did and, more importantly, which conceptual territory I chose to stake out and claim. A good rule of thumb is that if you are acting less savvy, constrained and strategic than the wedrifid persona then you are doing something wrong. Because I’m rather flippant and cavalier myself.
You had a good point, you made it, and you pointed out the problems with the responses to it. All of these comments were upvoted, many to double digits. But then your comments turned into personal attacks on Luke (suggesting he doesn’t understand the material he posts, suggesting he is lying about not seeing an earlier comment of yours). At that point, I felt (as, I’m guessing, did others), that your comments were actively counterproductive in trying to learn more about the minicamp, as well as promoting a community norm of insulting each other and assuming bad faith.
I also tend to get annoyed by, and downvote, comments to the effect of “The fact that I was downvoted reflects badly on all of you, who obviously downvoted me for [reason]” since I usually didn’t downvote for the reason mentioned and I don’t see them as a sincere attempt to understand the source of disagreement.
as well as promoting a community norm of insulting each other and assuming bad faith.
(The community norm should be to assume bad faith as much as is suggested by evidence. The extent to which bad faith is assumed shouldn’t be a product of a community norm. Insulting is rarely useful, of course.)
Given the human tendency to get emotionally involved in an argument, I think a rule of “assume bad faith as much as is suggested by evidence” qucikly devolves into “assume bad faith”. If you want to argue for a community norm of “assume bad faith as much as suggested by the evidence even after updating on all theevidence that people are really bad at evaluating other people’s motives”, I wouldn’t neccesarily disagree, but in practice, I think that looks a lot like “assume good faith”.
If there are known flaws in a method of inference, taking them into account should be part of what’s done when performing an inference, or what’s meant when suggesting to perform it. There should be no distinction between suggesting to look for a fact and suggesting to take into account possible flaws in the method of looking for that fact. This is simple exercise of the human power, something to encourage, not work around.
But, for instance, we know that flaws in our way of thinking about politics are so pervasive that we’ve decided to avoid it as much as possible. I would argue that flaws in our way of assessing whether other people in an online argument are arguing in good faith are nearly as pervasive, to the extent that assuming good faith is a better heuristic than assuming bad faith as much as is suggested by evidence.
And people who use an “assume good faith” model still change their mind once the evidence starts to accumulate; it’s about what your default assumption is, not whether it’s ever appropriate to say “You are arguing in bad faith.”
In order to see your earlier comments I would have had to return to that topic or specifically to the $125k post that I didn’t write, and I’d need to have done that sometime after your comments appeared on one of those posts, and I would have had to have comments sorted so as to see your comment before I gave up scrolling through all the comments there, and something about the first bit of your comments would need to have grabbed my attention so that I would have read it rather than continuing to scroll down.
My LW comment-reading behavior doesn’t ‘zoom in’ on comments from SilasBarta about minicamp. Either I would need to have gotten lucky or I would have to be doing that kind of thing with a broad selection of LW posts (and the comments made to them), and that’s just not the case, because I don’t “spend all day re-checking the comment threads on old LW posts.”
Asking generally—is there a compelling technical reason we don’t have an option to “subscribe” to a thread, or be emailed notifications of direct replies to our comments? Or even that there was a reply, if not the reply itself?
I am mildly irked that I have to go check my LW inbox for this; I think it reinforces my light tendencies towards online OCD and ADD to work to a pull model rather than a push.
(Or if there is one, please enlighten me, as I had thought I had searched sufficiently to find one if it existed.)
Even now, there are only 3 threads on your mini-camp result announcement topic, and mine is at the top, 16 comments in total. 90% of the discussion is about the request for evidence. No need to “zoom in” on anything, nor re-check frequently. Please, stop trying to come up with stories to account for not having seen it; it’s obvious you just never came back.
And really, it’s not some kind of mortal sin or anything—I don’t see why you’re goint to such lengths to justify it.
If I was a Confessor I would have tazed you by now.
I am alright with your original questions on this, but now you’re stretching. You seem to be going to unnecessary extremes to find fault with anything and everything that Luke has said on this. I judge this a violation of sanity.
If I were a Confessor I would have tazed you by now.
That’s probably why Confessors don’t exist. We’re not ready from them; we haven’t grown up enough to cope with even a single, tiny dissenting voice without resorting to threats of counterfactual violence.
If I were a Confessor I would have tazed you by now.
Thanks for pointing that out. Typing quickly on the go does not afford much spell/grammar checking.
And yes, by all means, I only meant that from reading (most of) the comments and discussion on this topic that I in my current state would have tazed him, had I the job description of a Confessor. I didn’t mean to imply that I was exceptionally good at judging sanity violations in any way, just a reference and a pithy statement of my view.
If I was a Confessor I would have tazed you by now.
I would have tazed you in turn. Not because you tazed Silas—I’d have done that too for his own sake. Rather, I’d have tazed you for the reasons you gave. You are observing two people bitching at each other each with their own (vastly different) kind of insanity and siding with the one with the most status and whose insane bitching is the most skilled (and socially typical). You are tazing the unsophisticated, lower status insane bitcher.
The evidence given suggests you are well suited to be a player in the social environment but not a confessor. In the future, when it matters, you can be expected to act as a social enforcer and not as a last bastion of sanity.
In the interest of full disclosure, I read the majority of this exchange in unordered chunks from the Recent Comments and mostly-backwards by going up context levels and trying to figure it out. And like I said to paper-machine I don’t mean to say I’m exceptionally good at judging sanity violations, just being pithy.
I’ll probably later on read them in some more-ordered fashion and see if I would taze luke too (even taking into account your claim you would).
Glad to know you’d be there to taze me if I started to go insane. It is appreciated. Not that I’m evaluating you as a fully superpowerful Confessor at the moment or anything. Here’s a question though… who would you have tazed first?
In the interest of full disclosure, I read the majority of this exchange in unordered chunks from the Recent Comments and mostly-backwards by going up context levels and trying to figure it out.
I can see why reading in that order/style would leave you just shooting Silas. :)
Here’s a question though… who would you have tazed first?
Chronologically Luke. He was insane way back when the claims were first made/not defended and Silas hadn’t gone insane yet. If I were to enter the room now after observing from outside I would shoot Silas first, pointedly shoot Luke as well and give everyone else in the room a stern look. Then I’d confiscate your tazer and turn in my confessors hood myself. Because I don’t want that kind of responsibility.
I’d keep the tazers. Because I have yet to meet anyone who I would trust to confessor at me, even though there are those whose advice I value. I would always take care to position myself with my back to the wall such that I could see the movements of any confessors and rely on my reflexes and laser tag prowess to protect me from any nosy interventionists. If necessary I’d take them all out in a massive confessor tazing spree.
Yes, but I’ve got the complicated issue of taking your interjection entirely truthfully. I don’t strongly believe you have any motivation to lie to me, but I may want to go through a few just to verify.
In any case, I’m not going to do it now, just when I have some spare time and am not browsing other comments.
This isn’t even an interesting thread relative to other flame wars we’ve had!
I only really started posting comments in March of this year. Reading the comments at all about a month or so before then, and have been reading LW itself for a little over a year. I may still be a little green for any of the more interesting flame wars.
And yet crap, I’m already doing what I said I wouldn’t. Shucks.
If Luke wants to say he just ignored comments when he trumpeted the success (like a good Olympian) -- fine.
If he wants to invent stories about how seeing my request would have required “spending all day re-checking the comment threads on old LW posts” or how it would have been difficult see my comment in the massive thicket of 3 threads with mine at the top, etc etc etc—then he’s making things up, which is not fine.
Yes, I can think of someone who warranted a Confessor zapping.
I think there’s a reasonable middle ground between fire-and-forget posting that you’re pushing, and the obsessive checking of posts that luke (very clearly) hyperbolized.
I think I remember coming back once soon after posting, but YES! This is what I’ve been trying to say! I never saw your earlier comments concerning the evidence for minicamp’s success, but you said you thought I was lying about that.
Lying or never following up because of the Olympus mentality, yes. Neither reflects well on you. In any case, I did contact you via other means and got no response that way either. (And other people have been made aware of this issue, who probably contacted you as well.)
Make sure to thank the folks who have modded you up for your strawman comment and your elaborate excuses for how you could have missed the entire discussion on a thread you started.
So are you going to settle on the “too busy” or “never saw it” position?
When did you contact me via other means? Was it by email? Who else do you think contacted me about the issue? As I’ve said, I don’t recall seeing your earlier comments on the matter. I’ve also said I don’t recall returning to the minicamp announcement discussion more than once after I originally posted it, but I don’t think this was because I have a ‘Mount Olympus’ mentality—more likely, it’s because I was busy doing other things. After all, it was an announcement post, not a ‘let’s discuss topic X’ post or a post that asked questions and expected replies.
I am curious what’s giving you the impression that I have an ‘Olympus mentality’, though, and whether others have gotten the same impression. The feeling ‘on the ground’ is quite different. I feel I (justly) have no authority at all because (1) I learned about the intelligence explosion less than a year ago and discuss it every day with people who have thought for much longer about the subject, (2) I have completed no degrees and published no papers (yet) on the subject, and (3) I am surrounded by math and programming geniuses who inadvertently cause me to feel insecure about my relative lack of training in those fields.
Moreover, I try to speak less “from personal authority” than everyone else, via bothering to cite the scientific papers supporting many of the claims I make—and even if all I did was track down the right papers, read the abstracts, and cite them, this would still be more work than other LWers usually do to ground their claims in the scientific literature. (Of course this isn’t always the case; I’m talking mostly about claims made in my articles about psychology and neuroscience.)
(Also, I don’t just start with personal claims and then do a Google scholar search for supporting evidence. I start with a question and then read textbook and review article excerpts to figure out which researchers are studying the topic, and then I read or skim their articles on the topic to figure out what we know, how we know it, and what we don’t know. And then I post my claims and cite the studies I found that guided me to make those claims.)
Back to your requests for evidence of minicamp’s success, and my impact upon it....
Anna took the time to write up some quantitative results from our exit survey. I haven’t seen you either thank her for answering your request or give a different reply yet.
She also included testimonials as to my own effectiveness during minicamp, and otherminicampers have given their own (positive) accounts. You haven’t replied to any of those.
I listed the preliminary evidence that led me to call minicamp a success, and you didn’t reply to that subthread yet.
You wanted to see the testimonials, and Anna posted them, and you didn’t say thanks or reply to that yet either.
I explained that further data measuring the effects of minicamp on its participants was still being gathered, but that this takes time and SI lacks available staff hours. Four other people have contacted me so far so they can free up my time by completing volunteer-doable tasks. At first you said you would volunteer, but then you apparently withdrew the offer.
As othershavesaid, your objections have been addressed and it’s hard to see why you’re still unsatisfied for now. Could you explain? Are you mostly just wanting to see additional quantitive evidence of minicamp’s effects on its participants’ lives? If so, I explained long ago that this data was still in the process of being collected and parsed.
I think a lot of the hubbub in this thread is due to different interpretations of SIAI related folks saying that the minicamp was ‘successful’. I think many people here have interpreted ‘success’ as meaning something like “definitely improved the rationality of the attendants lastingly” and I think SIAI folks intended to say something like “was competently executed and gives us something promising to experiment with in the future”.
I appreciate your efforts to decode this ‘Olympus mentality’ nonsense, and in general to make sure you’re not making communication errors. But at this point I believe you’re just wasting your time. You’ve documented your research methods better than I’ve ever seen someone do, and they certainly don’t need defending here.
On behalf of those who believe your work can positively impact the future of humanity and your time can be better spent elsewhere, I humbly request that you please file what you’ve been responding to under ‘trolling’ and move on.
I agree. This makes a perfect end point to the discussion, and unless anything actually relevant comes up, not reiteration of the same points, ignoring their previous refutations, you should stop it here.
What about the part where you ignored the things you were asking for, and kept pressing on slightly-modified issues?
Different topic. I’m being called a troll because Luke made implausible excuses, I’m calling him on it, he’s digging himself deeper, and yet people are voting him up instead of me. (And this wasn’t his first implausible one: remember the “we don’t have time”/”we can’t give that out” flip-flop?)
In any case, I’m not “pressing on slightly-modified issues”, nor ignoring anything (unless someone else replied as I would have). I listed my criteria way back when; that’s not answered, and there’s every reason to believe they have that information.
Silas, frankly, this could have been executed much more diplomatically.
Sure, if the voting pattern is to be believed, I should just make up implausible stories, and then call people trolls if they ever call me on it. What exactly should I have done differently? Write out what my comments should have looked like.
I don’t think what you’re doing is trolling, but I have a fairly tight definition for trolling—I think of it as posting driven by abstract malice, a desire to cause pain which is divorced from the topic at hand. That clearly isn’t you.
On the other hand, I think you’re engaged in a bad emotional habit—attributing a negative motivation to someone else on very little evidence, and getting stuck on the idea of that motivation.
After all, it was an announcement post, not a ‘let’s discuss topic X’ post or a post that asked questions and expected replies.
By the way, I’d like you to start labeling your topics in the future to avoid such a misunderstanding. Specifically, on all those which you deem to be “an announcement”, which we’re not supposed to discuss, or argue about, or criticize, or question, or whine about—like the mini-camp results topic—please indicate as much. Thanks.
When did you contact me via other means? Was it by email?
Yes, around the time I posted it. I don’t care if you find this implausible; I didn’t find your claims of never having seen my requests plausible either. I don’t have time to substantiate this claim either; I’m too busy.
After all, it was an announcement post, not a ‘let’s discuss topic X’ post or a post that asked questions and expected replies.
People usually expect comments on topics they start; I don’t know why you would expect otherwise just because it’s “an announcement post”. It’s the Olympus mentality that says, “I talk, you listen, replies not wanted.” It’s the mentality that replies defensively to any cross-examination of evidence you’ve presented.
Anna took the time to write up some quantitative results from our exit survey. I haven’t seen you either thank her for answering your request or give a different reply yet.
She also included testimonials as to my own effectiveness during minicamp, and other minicampers have given their own (positive) accounts. You haven’t replied to any of those.
I listed the preliminary evidence that led me to call minicamp a success, and you didn’t reply to that subthread yet. …
You wanted to see the testimonials, and Anna posted them, and you didn’t say thanks or reply to that yet either.
I didn’t assign those high priority for a reply because others already said what I would have said in reply, and it was becoming clear that I was not alone in being suspicious of this evidence.
I also didn’t ask to see testimonials because I didn’t consider those strong evidence for what you’re claiming—like others noted, they don’t really distinguish you from any other self-help camp.
I explained that further data measuring the effects of minicamp on its participants was still being gathered, but that this takes time and SI lacks available staff hours.
And I explained that you don’t get to play “evidence takes time!” while saying, “yep, I already have enough evidence to call this a success and ignore any questioning of this claim”.
In a reply to you, jsalvatier linked to additional earlier positive testimonials, to which you also did not reply.
What was there to reply to? “Oh, well, since you personally feel you had a good time, I guess I shouldn’t be suspicious that it improved anyone’s rationality”?
At first you said you would volunteer, but then you apparently withdrew the offer.
I said I would pay the ransom you’re demanding for your evidence. That was the only reason I was willing to help you. I never withdrew any offer; rather, you tried to change the topic to helping with your research, which I don’t want to do, never did, and never offered to.
As others have said, your objections have been addressed and it’s hard to see why you’re still unsatisfied for now. Could you explain?
Those are regarding a different thread and different claim; there’s nothing to explain here, and I’d appreciate if you didn’t misreport evidence.
Also keep in mind jsalvatier’s comment:
Sure thing—and you can keep it in mind too, when using increasingly strong superlatives to describe our success.
Despite multiplerequests to drop this discussion, I’d like to put a little more effort toward mutual understanding. Perhaps I’m irrationally optimistic for reconciliation and convergence.
Others have addressed the unproductive ‘attack-mode’ nature of your comments; I won’t address that here. Suffice it to say that I have plenty to learn myself about communicating diplomatically.
I also won’t say much more on the issue of my not having seen your earlier calls for evidence of minicamp’s success. I can only repeat: If you want to be sure I’ll read a particular comment, make sure you contact me directly or reply to one of my comments so that your comment shows up in the LW inbox. I do not have time to keep revisiting old posts and reading all comments made on them, and I kinda doubt anyone thinks that is the best use of my limited time when I could instead be doing research and academic outreach related to rationality and FAI theory. You may insist on attributing this to my ‘Olympus Mentality’, though I’ll try to dissuade you of this interpretation below.
As for your definite accusation that I lied when I said I hadn’t seen your earlier comments on the topic, it remains the case that I never replied to them, and they seem like comments I would have replied to given my well-documented defensiveness on LW. Just notice how tenaciously I’ve defended myself in this discussion, despite a continuous slew of character attacks.
As for your accusation that I strawmanned you, I tried to explain that unless I had a policy of checking tons of old posts for new comments it’s not clear I would have seen your original comment, but you seem to simply disagree, so I don’t think there’s much more to say about that.
Finally, you seem to have suggested that I said I made announcement posts “we’re not supposed to discuss, or argue about, or criticize, or question, or whine about—like the mini-camp results topic”, but that’s just not true. You’re welcome to discuss, argue, criticize, question, or whine about anything I post on Less Wrong. All I said was that I don’t go back and check every post for new comments, and that if you want to make sure I read something you should contact me directly or be sure to reply directly to one of my comments so that I see it in my LW inbox.
A ‘Successful’ Minicamp
jsalvatier has repeatedly suggested that we may have different ideas of what I meant when I wrote that Rationality Minicamp was a success.
As KPier wrote in response to what seems to be your original comment on this topic, “The article pretty clearly states that the claims about the effects of the camp were based on exit surveys, and that the impact of the camp is demonstrated by the projects the camp grads are now working on. You could debate whether those are good measures, but we don’t exactly have better ones.” Later, Anna and myself gave that specific evidence in more detail.
You might be willing to concede that the evidence from exit surveys and testimonials provide about as much evidence of minicamp ‘success’ as such measures are capable of providing, though that may not be much. Is that true?
But of course, you’ve been asking for stronger evidence. You’d like to see measures of rationality improvement or life success or something like that. I addressed this exact request directly in my very first comment on the topic:
We collected lots of data before and during minicamp. We are waiting for some time to pass before collecting followup data, because it takes time for people’s lives to change, if they’re going to change...
...we are still gathering data… before-and-after results will have to wait a while...
You replied that if these stronger forms of evidence don’t yet exist, then I shouldn’t claim that minicamp was a success. But again, I must repeat what KPier originally told you: My original blog post on minicamp being a success made it clear that such ‘success’ was assessed based on exit surveys and participant testimonials:
Our exit survey shows that the camp was a smashing success...
[Participants] continue to share the minicamp’s impact on themselves, and the impact they are having on others as a result, via an online mailing list and regular Skype video chats.
You seem to have interpreted ‘success’ in a different way than it was used in that blog post, perhaps to mean something like “Rationality minicamp successfully improved the rationality and life success of its participants, as demonstrated by several quantitative measures.”
But as the original blog post shows, that’s not what was meant to be claimed by calling the rationality minicamp a ‘success’.
Now, I’ll be happy to make this clearer by editing the original blog post, and by asking Eliezer to edit his post above. We could call it a ‘highly praised’ or ‘well-reviewed’ minicamp where brevity is needed, and where we have more space we could say something like “The minicamp was well-received by participants, who rated it highly in our anonymous exit survey and have given glowing reviews and reports of their resulting self-improvement. Further evidence concerning the minicamp’s effect on participants’ rationality and life success are pending.”
As for the fact that this data is still being gathered because it takes time for people’s lives to change and it takes time to parse collected data, you appear to have called this an “implausible excuse,” though I still don’t know what’s implausible about it.
Or perhaps what you meant to call an “implausible excuse” is my point about how the raw exit survey data is anonymous and private, and that’s why we can’t publish it. But I’m not sure what’s implausible about that, either. You can ask the minicamp participants themselves: We asked them to fill out one form that would be anonymous and private (the exit survey form), and another that would be identifiable and public (the testimonials form).
You also said I flip-flopped between these two “excuses”, but that’s not true. I maintain both claims. It takes time to collect and parse data on life changes and we can’t publish the private and anonymous exit form data.
Olympus mindset
You keep finding things that you choose to interpret as demonstrating my ‘Olympus mindset’ without addressing disconfirming evidence like what I gave above:
I feel I (justly) have no authority at all because (1) I learned about the intelligence explosion less than a year ago and discuss it every day with people who have thought for much longer about the subject, (2) I have completed no degrees and published no papers (yet) on the subject, and (3) I am surrounded by math and programming geniuses who inadvertently cause me to feel insecure about my relative lack of training in those fields.
Moreover, I try to speak less “from personal authority” than everyone else, via bothering to cite the scientific papers supporting many of the claims I make—and even if all I did was track down the right papers, read the abstracts, and cite them, this would still be more work than other LWers usually do to ground their claims in the scientific literature.
I’d also be curious to hear from others who think I display an ‘Olympus mindset’, and what triggers they think give them that impression. I don’t want to be giving off an inaccurate impression of myself in that way. I still practice facial expressions in the mirror because my face sometimes doesn’t clearly communicate my mindset, and obviously I still need to practice my online communication because my typed words don’t always clearly communicate my mindset, either.
EDIT: This has become an unproductive flame war, with no small thanks to my own behavior, and I will now bow out.
Now, I’ll be happy to make this clearer by editing the original blog post, and by asking Eliezer to edit his post above. We could call it a ‘highly praised’ or ‘well-reviewed’ minicamp where brevity is needed, and where we have more space we could say something like “The minicamp was well-received by participants, who rated it highly in our anonymous exit survey and have given glowing reviews and reports of their resulting self-improvement. Further evidence concerning the minicamp’s effect on participants’ rationality and life success are pending.”
As for the fact that this data is still being gathered because it takes time for people’s lives to change
I disagree with this. My intuition, supplemented by experience in somewhat analogous religious retreats, is that change happens easily in the camp environment and the question is how much of that will be inculcated enough to survive once the return to life happens.
I’d say it takes time to be sure people’s lives have changed permanently, but not too much time for them to change.
So if minicamp related posts had instead of ‘success’ said ‘we were very pleased with the execution of the camp and it gives us a promising direction to explore’ would you have felt similarly to how you feel now?
If so, why are you still arguing about whether SIAI has prematurely judged the minicamp instead of asking them to make their judgment less ambiguous?
They don’t seem to agree that they should have done anything differently, so I don’t know why I would be at the point of asking them to change something. But yes, I would appreciate characterizations of the mini-camp that aren’t extremely misleading, please pass it on through someone who they’ll listen to.
It seems rather unlikely to me that being a mini-camp participant would have more of an effect on someone’s life than being a Visiting Fellow, new techniques or not—and if I am wrong, I would very much want to encounter these new techniques!
I wouldn’t be that surprised. Explicit rationality exercises were only starting to be developed during the last month of my stay, and at that point they mostly fell into the category of “entertaining, but probably not hugely useful”. The main rationality boost came from being around others with a strong commitment to rationality, but as situationist psychology would have it, the effect faded once I was out of that environment.
The positive endorphin rush from you and lukeprog sends signals that loook just like the enthusiastic gushing I see from any week-long “how to fix your life in five easy steps!” seminar. Smart people get caught up in biased thinking all the time. I had a good friend quit AI research to sell a self-help book, so I may be particularly sensitive to this :)
Objective data means I can upgrade this from “oh bunnies, another self-help meme” to “oooh, fascinating and awesome thing that I want to steal for myself.” As long as it signals like a self-help meme, I’m going to shoot it down just like I’d shoot down any similar meme that tried to sell itself here on LessWrong.
All right, but there’s a fine line between shooting down self-help memes and unnecessarily discouraging project-builders from getting excited about their work. It’s not fun or helpful for a pioneer to have his or her every first step be met with boundless skepticism. Your concerns sound real enough to me, but even an honest concern can be rude, and even a rationalist can validly trade off a tiny little bit of honesty for a whole lot of politeness and sympathy.
Why do I say “a tiny little bit of honesty?” Well, if the minicamp were being billed as “finished,” “polished,” “complete,” “famous,” “proven,” or “demonstrably successful,” as many self-help programs are, then it would make sense to demand data supporting those claims.
Instead, the PR blurb says that “Starting on May 28th, the Singularity Institute ran a one-week Rationality Training Camp. Our exit survey shows that the camp was a smashing success, surpassing the expectations of the organizers and the participants.”
Leaving aside the colorful language that can and should characterize most press releases, this is a pretty weak claim: the camp beat expectations. Do you really need to see data to back that up?
For me, this isn’t about making SIAI transparent; it does quite enough in that regard. It’s about stopping an information cascade genie that’s already out of the bottle.
Let me put it this way: right now the ratio of “relying on the assumption of mini-camp’s success for decision making” to “available evidence for its success” is about 20-to-1. As I warned before, it’s quickly becoming something “everyone knows” despite the lack of evidence (and major suspicions of many people that it wouldn’t succeed going in). And that believe will keep feeding on itself unless someone traces it back to its original evidence.
It doesn’t reassure me that I’m told I have to keep waiting before anything’s conclusive, yet they can declare it a success now.
I just want the reliable evidence they claim to have, rather than just dime-a-dozen self-help testimonials. They collected hard data, and I gave them a list of things they could provide that are easy to gather and don’t compromise privacy, and are much more likely to be present if the success were real than if it were not. Even after AnnaSalamon’s circling of the wagons I don’t see that.
Oh, sure. The reason is easy to communicate. We explicitly told minicampers that their feedback on the exit survey would be private and anonymous, for maximal incentive to be direct and honest. We are not going to violate that agreement. The testimonials were given via a separate form with permission granted to publish THAT data publically.
Oh, sure. The reason is easy to communicate. We explicitly told minicampers that their feedback on the exit survey would be private and anonymous, for maximal incentive to be direct and honest.
I’m unclear, then, why you are citing a lack of staff hours if the data cannot be published at all.
The testimonials were given via a separate form with permission granted to publish THAT data publically.
Has the raw testimonial data been published?
I’m assuming you have data beyond exit surveys and testimonials...?
A summary of the data can be published, for example median scores for measured values. But the data can’t be published in raw form.
Not sure if raw testimonial data has been published yet. We do have data beyond testimonials and exit surveys, but that, too, requires precious staff hours to compile and write up, and it is still in the process of being collected.
Typing this stuff from a phone, pardon the brevity...
This comes off very strongly as the typical bureaucratic protectiveness—a business doesn’t want to share raw data, because raw data is a valuable resource. If you came out and said this was the reason, I’d be more understanding, but it would still feel like a major violation of community norms to be so secretive.
If simple secrecy is indeed the case, I would urge you, please, be honest about this motive and say so explicitly! At least then we are having an honest discussion, and the rest of this comment can be disregarded.
In short, what is the reason you can’t share this RAW data, which you state you collected, and which you’ve presumably found sufficient for your own preliminary conclusions? I don’t think Silas is asking for or expecting an elegant power-point presentation or a concise statistical analysis—I know I would personally love to simply see raw data.
Is there truly not a single spreadsheet or writeup that you could drop up for us to study while you collect the rest of the data?
Good grief, people. There are conspiracies that need ferreting out, but they do not revolve around generating fake data about the effectiveness of an alpha version of a rationality training camp that was offered for free to a grateful public.
I went to the minicamp, I had a great time, I learned a lot, and I saw shedloads of anecdotal evidence that the teachers are striving to become as effective as possible. I’m sure they will publish their data if and when they have something to say.
Meanwhile, consider re-directing your laudable passion for transparency toward a publicly traded company or a medium-sized city or a research university. Fighting conspiracies is an inherently high-risk activity, both because you might be wrong about the conspiracies’ existence, and because even if the conspiracy exists, you might be defeated by its shadowy and awful powers. Try to make sure the risks you run are justified by an even bigger payoff at the end of the tunnel.
I don’t think anybody is accusing the minicamp folks of anything of the kind. But public criticism and analysis of conclusions is the only reliable way to defend against overconfidence and wishful thinking.
When I ended my term as an SIAI Visiting Fellow, I too felt like the experience would really change my life. In reality, most of the effects faded away within some months, though a number of factors combined to permanently increase my average long-term happiness level.
Back then the rationality exercises were still being worked out and Luke wasn’t around, so it’s very plausible that the minicamp is a lot more effective than the Visiting Fellow program was for me. But the prior for any given self-help program having a permanent effect is small, even if participants give glowing self-reports at first, so deep skepticism is warranted. No conspiracies are necessary, just standard wishful thinking biases.
Though I think this was the third time that Silas raised the question before finally getting a reply, despite his comment being highly upvoted each time. If some people are harboring suspicions of SIAI covering up information, well, I can’t really say I’d blame them after that.
For the record, I for one don’t recall reading any of SilasBarta’s earlier comments on this topic.
I find that unlikely. That would mean you never followed up after trumpeting your success—you just posted the topic, and never bothered to come back and see what people had to say. And that you didn’t see the top comment on the 125k fundraiser thread. Then again, this is consistent with what komponisto said about your “Mt Olympus” mentality: just say stuff ex cathedra and expect everyone to fall in line or otherwise swoon.
I don’t understand this bit about my ‘My Olympus’ mentality. Until very recently I wasn’t on SI’s full-time staff. And as far as I can tell, I’ve spent vastly more time substantiating what I say by citing the relevant scientific literature (rather than relying on whatever personal authority I’m supposed to have, which I don’t think is much at all) than anyone else on Less Wrong.
And no, I don’t expect everyone to “fall in line or otherwise swoon.” It’s just that I don’t have time to write up a 20-citation research article supporting every sentence I write. If the reasons that led me to write a certain sentence aren’t available to you, as is usually the case, then you should only be updating your beliefs as much as you should given the evidence of my testimony, which in many cases should be very little.
As for you not believing me when I say that I don’t recall reading your earlier comments calling for evidence about minicamp’s success, well… the only evidence I have for you besides my testimony is that I hadn’t replied to any of your earlier comments on the matter. If you don’t believe me, well, so be it: that’s all I’ve got.
Indeed you have, and you’ve been well rewarded, with 24,000+ karma points and a full-time position at SI (with EY himself begging for money to pay you in a promoted LW post—I’ll bet that felt good!). What you haven’t earned, however, is the right to ignore people without their being offended. (The only person who might conceivably have that level of status is EY, and I think even that is debatable.)
The impression I think you give is one of writing all this great LW material, but then being “too busy” with your high-status SI work to read people’s comments on it. Surely you can see how that comes across.
I reply to many comments but certainly not all. I can’t respond to all my critics, and it’s probably unwise to do so. I’m also sympathetic to thomblake’s comment on this discussion:
To be clear, my comment pertained to reading others’ comments, not necessarily replying to them.
I would also like to stress that, while I am not sympathetic to thomblake’s comment (I could hardly be expected to be, since he effectively labeled a comment of mine “nonsense”), I have not voiced any complaint about the underlying subject (proof of minicamp effectiveness), and should not be confused with Silas. (This isn’t to say that I don’t think he has a point also; but I emphatically do not consider myself to have participated in any “trolling”.)
I entered the discussion because of the “Olympus” issue, which I had noticed in other contexts and had considered bringing up before. (Evidently I am not the only one, because my comment on the matter has—against all expectations of mine—been voted up into the 20s.)
“The impression I think you give is one of writing all this great LW material, but then being “too busy” with your high-status SI work to read people’s comments on it. Surely you can see how that comes across.”
LOL. If one does form that impression from Luke’s posts I would suggest that you:
1.) Don’t be so sensitive and get a life beyond obssessing over blog comments. 2.) Realize that people who actually get things done usually don’t make responding to blog posts a high or even medium priority in life. Seriously, most people with productive lives don’t have time to obssess over blogs.
Appreciate the good content, criticize the points worth criticizing and try not to cry about the minor points and the fact that someone doesn’t respond.
To be fair the reference class should be bloggers, not productive people in general.
My point was that either a) you didn’t do what most people do when they start topics—look at them a second time—or b) you’re lying. Neither is likely, but a) is at least consistent with the Mt Olympus mentality—why else would you never follow up on a thread with a major announcement? What other mentality would lead to that behavior?
My request for evidence was also (for most of the time it was up) the top comment on the 125k fundraising thread, which does give someone reason to be skeptical you weren’t aware anyone had asked for any substantiation.
Sure, that should be resolved for the more improbable claims.
And in the case where your justifications come from the literature, I thought you could just google the term, read the abstract, and copy-paste the citation?
Being busy with research and writing? Seriously, I don’t spend all day re-checking the comment threads on old LW posts. If you want to reach me, please contact me directly or reply to one of my comments, so it shows up in my LW inbox.
“Being busy with research and writing? Seriously, I don’t spend all day re-checking the comment threads on old LW posts.”
Amen!
Research & Writing > responding to some disgruntled blogger.
You were aware he was replying to a strawman and that I never expected that he should “spend all day rechecking …”, right? (Guess not, and ditto for those who voted you up.)
I can see the general point you were making with your criticisms and so could the other people who voted you up on your early comments on this issue. That being said I suggest you quit while you are (or were) ahead. It is too easy to let others reclaim the moral high ground if you stay on the same topic for too long and let them put you on the defensive.
This would allow you to maintain context specific credibility for use the next time the same, or similar claims were made about success that you don’t believe are justified.
Would I be unreasonable or unrealistic if I expressed a desire to not see this any of this SIAI inside baseball on Less Wrong… ever? Insofar as these rationality minicamps are something we think people who want to be more rational should take part it, obviously data on their effectiveness is very important. But insofar as the ‘success’ of the minicamp figured into SIAI’s decision to hire lukeprog (and that seems to be the issue for the moment) I could care less. I realize of course Less Wrong and SIAI are intimately connected and I’m personally at the low end on a spectrum of interest and involvement in SIAI. And I’m fine with seeing the occasional fund-raising post or strategy discussions in Discussion… after all I’m not paying for the pleasure of posting and reading here. But an ongoing flame war about a random criteria in an SIAI hiring dominates the recent comments section and is of no interest to me (and one assumes, others in my position). In the interest of keeping SIAI and Less Wrong somewhat separate shouldn’t SIAI have some other avenue donors can use to voice concerns and criticisms so that it doesn’t interfere with the interesting content here?
(And if Silas isn’t a donor… Thomas %&$@! Bayes why does anyone care?!)
For the record, Silas is a donor—listed on our donor list as having donated $2,000.
I don’t care for flame wars either. But what I do care about is that if it is permitted to make a declaration of fact on lesswrong it is permitted to to refute it. The details of what you suggest in the parent violate this. You advocating the lesswrong equivalent of true evil!
I’m sorry! I repent!
I just don’t feel like I have the standing to demand that SIAI fund-raising letters be left off the main page entirely- so I was looking for a compromise.
I don’t, but if I don’t say that out loud, other people go on loudly caring, and if I do say it out loud, I get downvoted. (Shrug.)
Errr, do you really care that much about being downvoted?
Downvotes imply net disapproval, particularly for someone whose comments get read as much as Eliezer’s. If you think of it as simply losing points, it seems trivial, but if you take it as a sign that “people seem not to like it when I do that,” it’s a meaningful consideration.
Even Eliezer needs Karma, if you ignore the sequences Alicorn and Yvain can beat him out ;)
Very well.
This is a pretty sad day for LW, to learn that you can just lie and strawman your way out of criticism, because whoever calls you on it is just “staying on the same topic for too long” :-/
(I hope you’ll pardon the digression into crude discussion of rational strategies for influence. This is a far more interesting topic than what serves for the object level at this point!)
Speak more strategically. Don’t let the verbal signals you utter be tools you use to salve and release your own feelings. Not because it is virtuous, purely because that doesn’t get you what you want. Also note that not only am I someone who consistently voted you up and those insulting you down, my comment provided stronger support for your position than your most recent comments managed. Allow me to translate what I said into ‘fun’ rather than vaguely polite:
Now, consider the difference between the above wording and what I actually said. Notice that it positions me as somewhat of an ally, assumes the criticisms you make of Luke are valid but at the same time doesn’t try to alienate me from the tribe. See why I chose to use the wording I did and, more importantly, which conceptual territory I chose to stake out and claim. A good rule of thumb is that if you are acting less savvy, constrained and strategic than the wedrifid persona then you are doing something wrong. Because I’m rather flippant and cavalier myself.
People have previously tried to assist SilasBarta on the topic of useful exposition and tone, but it does not seem to have any lasting effect.
One of the most important phrases in a rationalist’s toolkit for those rare occasions when other-optimizing is called for.
I have found some success from aiming it at myself.
I really don’t think that’s the problem here.
You had a good point, you made it, and you pointed out the problems with the responses to it. All of these comments were upvoted, many to double digits. But then your comments turned into personal attacks on Luke (suggesting he doesn’t understand the material he posts, suggesting he is lying about not seeing an earlier comment of yours). At that point, I felt (as, I’m guessing, did others), that your comments were actively counterproductive in trying to learn more about the minicamp, as well as promoting a community norm of insulting each other and assuming bad faith.
I also tend to get annoyed by, and downvote, comments to the effect of “The fact that I was downvoted reflects badly on all of you, who obviously downvoted me for [reason]” since I usually didn’t downvote for the reason mentioned and I don’t see them as a sincere attempt to understand the source of disagreement.
(The community norm should be to assume bad faith as much as is suggested by evidence. The extent to which bad faith is assumed shouldn’t be a product of a community norm. Insulting is rarely useful, of course.)
Given the human tendency to get emotionally involved in an argument, I think a rule of “assume bad faith as much as is suggested by evidence” qucikly devolves into “assume bad faith”. If you want to argue for a community norm of “assume bad faith as much as suggested by the evidence even after updating on all the evidence that people are really bad at evaluating other people’s motives”, I wouldn’t neccesarily disagree, but in practice, I think that looks a lot like “assume good faith”.
If there are known flaws in a method of inference, taking them into account should be part of what’s done when performing an inference, or what’s meant when suggesting to perform it. There should be no distinction between suggesting to look for a fact and suggesting to take into account possible flaws in the method of looking for that fact. This is simple exercise of the human power, something to encourage, not work around.
But, for instance, we know that flaws in our way of thinking about politics are so pervasive that we’ve decided to avoid it as much as possible. I would argue that flaws in our way of assessing whether other people in an online argument are arguing in good faith are nearly as pervasive, to the extent that assuming good faith is a better heuristic than assuming bad faith as much as is suggested by evidence.
And people who use an “assume good faith” model still change their mind once the evidence starts to accumulate; it’s about what your default assumption is, not whether it’s ever appropriate to say “You are arguing in bad faith.”
For what it’s worth, I agree that you’re doing the right thing.
Thank you. For your karma’s sake, though, you might want to keep that to PM.
Accumulating karma is trivial; burning it is something else entirely.
Strawman much? Returning once to a topic you started != “spending all day re-checking the comment threads on old LW posts”.
In order to see your earlier comments I would have had to return to that topic or specifically to the $125k post that I didn’t write, and I’d need to have done that sometime after your comments appeared on one of those posts, and I would have had to have comments sorted so as to see your comment before I gave up scrolling through all the comments there, and something about the first bit of your comments would need to have grabbed my attention so that I would have read it rather than continuing to scroll down.
My LW comment-reading behavior doesn’t ‘zoom in’ on comments from SilasBarta about minicamp. Either I would need to have gotten lucky or I would have to be doing that kind of thing with a broad selection of LW posts (and the comments made to them), and that’s just not the case, because I don’t “spend all day re-checking the comment threads on old LW posts.”
Asking generally—is there a compelling technical reason we don’t have an option to “subscribe” to a thread, or be emailed notifications of direct replies to our comments? Or even that there was a reply, if not the reply itself?
I am mildly irked that I have to go check my LW inbox for this; I think it reinforces my light tendencies towards online OCD and ADD to work to a pull model rather than a push.
(Or if there is one, please enlighten me, as I had thought I had searched sufficiently to find one if it existed.)
Even now, there are only 3 threads on your mini-camp result announcement topic, and mine is at the top, 16 comments in total. 90% of the discussion is about the request for evidence. No need to “zoom in” on anything, nor re-check frequently. Please, stop trying to come up with stories to account for not having seen it; it’s obvious you just never came back.
And really, it’s not some kind of mortal sin or anything—I don’t see why you’re goint to such lengths to justify it.
If I was a Confessor I would have tazed you by now.
I am alright with your original questions on this, but now you’re stretching. You seem to be going to unnecessary extremes to find fault with anything and everything that Luke has said on this. I judge this a violation of sanity.
That’s probably why Confessors don’t exist. We’re not ready from them; we haven’t grown up enough to cope with even a single, tiny dissenting voice without resorting to threats of counterfactual violence.
Thanks for pointing that out. Typing quickly on the go does not afford much spell/grammar checking.
And yes, by all means, I only meant that from reading (most of) the comments and discussion on this topic that I in my current state would have tazed him, had I the job description of a Confessor. I didn’t mean to imply that I was exceptionally good at judging sanity violations in any way, just a reference and a pithy statement of my view.
I would have tazed you in turn. Not because you tazed Silas—I’d have done that too for his own sake. Rather, I’d have tazed you for the reasons you gave. You are observing two people bitching at each other each with their own (vastly different) kind of insanity and siding with the one with the most status and whose insane bitching is the most skilled (and socially typical). You are tazing the unsophisticated, lower status insane bitcher.
The evidence given suggests you are well suited to be a player in the social environment but not a confessor. In the future, when it matters, you can be expected to act as a social enforcer and not as a last bastion of sanity.
In the interest of full disclosure, I read the majority of this exchange in unordered chunks from the Recent Comments and mostly-backwards by going up context levels and trying to figure it out. And like I said to paper-machine I don’t mean to say I’m exceptionally good at judging sanity violations, just being pithy.
I’ll probably later on read them in some more-ordered fashion and see if I would taze luke too (even taking into account your claim you would).
Glad to know you’d be there to taze me if I started to go insane. It is appreciated. Not that I’m evaluating you as a fully superpowerful Confessor at the moment or anything. Here’s a question though… who would you have tazed first?
I can see why reading in that order/style would leave you just shooting Silas. :)
Chronologically Luke. He was insane way back when the claims were first made/not defended and Silas hadn’t gone insane yet. If I were to enter the room now after observing from outside I would shoot Silas first, pointedly shoot Luke as well and give everyone else in the room a stern look. Then I’d confiscate your tazer and turn in my confessors hood myself. Because I don’t want that kind of responsibility.
I’d keep the tazers. Because I have yet to meet anyone who I would trust to confessor at me, even though there are those whose advice I value. I would always take care to position myself with my back to the wall such that I could see the movements of any confessors and rely on my reflexes and laser tag prowess to protect me from any nosy interventionists. If necessary I’d take them all out in a massive confessor tazing spree.
Can I take back what I said about being cool with you tazing me?
I think I’m just going to go read this thing in order and ignore any responses to my comments for a bit...
That sounds like an inefficient use of your time (also note that the conversation spans several posts).
This isn’t even an interesting thread relative to other flame wars we’ve had!
Yes, but I’ve got the complicated issue of taking your interjection entirely truthfully. I don’t strongly believe you have any motivation to lie to me, but I may want to go through a few just to verify.
In any case, I’m not going to do it now, just when I have some spare time and am not browsing other comments.
I only really started posting comments in March of this year. Reading the comments at all about a month or so before then, and have been reading LW itself for a little over a year. I may still be a little green for any of the more interesting flame wars.
And yet crap, I’m already doing what I said I wouldn’t. Shucks.
“It’s not the crime, it’s the coverup.”
If Luke wants to say he just ignored comments when he trumpeted the success (like a good Olympian) -- fine.
If he wants to invent stories about how seeing my request would have required “spending all day re-checking the comment threads on old LW posts” or how it would have been difficult see my comment in the massive thicket of 3 threads with mine at the top, etc etc etc—then he’s making things up, which is not fine.
Yes, I can think of someone who warranted a Confessor zapping.
I think there’s a reasonable middle ground between fire-and-forget posting that you’re pushing, and the obsessive checking of posts that luke (very clearly) hyperbolized.
I think I remember coming back once soon after posting, but YES! This is what I’ve been trying to say! I never saw your earlier comments concerning the evidence for minicamp’s success, but you said you thought I was lying about that.
Lying or never following up because of the Olympus mentality, yes. Neither reflects well on you. In any case, I did contact you via other means and got no response that way either. (And other people have been made aware of this issue, who probably contacted you as well.)
Make sure to thank the folks who have modded you up for your strawman comment and your elaborate excuses for how you could have missed the entire discussion on a thread you started.
So are you going to settle on the “too busy” or “never saw it” position?
When did you contact me via other means? Was it by email? Who else do you think contacted me about the issue? As I’ve said, I don’t recall seeing your earlier comments on the matter. I’ve also said I don’t recall returning to the minicamp announcement discussion more than once after I originally posted it, but I don’t think this was because I have a ‘Mount Olympus’ mentality—more likely, it’s because I was busy doing other things. After all, it was an announcement post, not a ‘let’s discuss topic X’ post or a post that asked questions and expected replies.
I am curious what’s giving you the impression that I have an ‘Olympus mentality’, though, and whether others have gotten the same impression. The feeling ‘on the ground’ is quite different. I feel I (justly) have no authority at all because (1) I learned about the intelligence explosion less than a year ago and discuss it every day with people who have thought for much longer about the subject, (2) I have completed no degrees and published no papers (yet) on the subject, and (3) I am surrounded by math and programming geniuses who inadvertently cause me to feel insecure about my relative lack of training in those fields.
Moreover, I try to speak less “from personal authority” than everyone else, via bothering to cite the scientific papers supporting many of the claims I make—and even if all I did was track down the right papers, read the abstracts, and cite them, this would still be more work than other LWers usually do to ground their claims in the scientific literature. (Of course this isn’t always the case; I’m talking mostly about claims made in my articles about psychology and neuroscience.)
(Also, I don’t just start with personal claims and then do a Google scholar search for supporting evidence. I start with a question and then read textbook and review article excerpts to figure out which researchers are studying the topic, and then I read or skim their articles on the topic to figure out what we know, how we know it, and what we don’t know. And then I post my claims and cite the studies I found that guided me to make those claims.)
Back to your requests for evidence of minicamp’s success, and my impact upon it....
Anna took the time to write up some quantitative results from our exit survey. I haven’t seen you either thank her for answering your request or give a different reply yet.
She also included testimonials as to my own effectiveness during minicamp, and other minicampers have given their own (positive) accounts. You haven’t replied to any of those.
In a reply to you, jsalvatier linked to additional earlier positive testimonials, to which you also did not reply.
I listed the preliminary evidence that led me to call minicamp a success, and you didn’t reply to that subthread yet.
You wanted to see the testimonials, and Anna posted them, and you didn’t say thanks or reply to that yet either.
I explained that further data measuring the effects of minicamp on its participants was still being gathered, but that this takes time and SI lacks available staff hours. Four other people have contacted me so far so they can free up my time by completing volunteer-doable tasks. At first you said you would volunteer, but then you apparently withdrew the offer.
As others have said, your objections have been addressed and it’s hard to see why you’re still unsatisfied for now. Could you explain? Are you mostly just wanting to see additional quantitive evidence of minicamp’s effects on its participants’ lives? If so, I explained long ago that this data was still in the process of being collected and parsed.
Also keep in mind jsalvatier’s comment:
Luke,
I appreciate your efforts to decode this ‘Olympus mentality’ nonsense, and in general to make sure you’re not making communication errors. But at this point I believe you’re just wasting your time. You’ve documented your research methods better than I’ve ever seen someone do, and they certainly don’t need defending here.
On behalf of those who believe your work can positively impact the future of humanity and your time can be better spent elsewhere, I humbly request that you please file what you’ve been responding to under ‘trolling’ and move on.
I agree. This makes a perfect end point to the discussion, and unless anything actually relevant comes up, not reiteration of the same points, ignoring their previous refutations, you should stop it here.
To me, your comments here look like trolling, but I guess YMMV.
Calling someone on phony excuses is “trolling”? (That’s the only reason the recent thread has stretched out so far.)
Next time, I guess I should shut up when Luke makes an(other) implausible claim?
What about the part where you ignored the things you were asking for, and kept pressing on slightly-modified issues?
I’d call that trolling, along with the tone of some of your comments.. Silas, frankly, this could have been executed much more diplomatically.
Different topic. I’m being called a troll because Luke made implausible excuses, I’m calling him on it, he’s digging himself deeper, and yet people are voting him up instead of me. (And this wasn’t his first implausible one: remember the “we don’t have time”/”we can’t give that out” flip-flop?)
In any case, I’m not “pressing on slightly-modified issues”, nor ignoring anything (unless someone else replied as I would have). I listed my criteria way back when; that’s not answered, and there’s every reason to believe they have that information.
Sure, if the voting pattern is to be believed, I should just make up implausible stories, and then call people trolls if they ever call me on it. What exactly should I have done differently? Write out what my comments should have looked like.
I don’t think what you’re doing is trolling, but I have a fairly tight definition for trolling—I think of it as posting driven by abstract malice, a desire to cause pain which is divorced from the topic at hand. That clearly isn’t you.
On the other hand, I think you’re engaged in a bad emotional habit—attributing a negative motivation to someone else on very little evidence, and getting stuck on the idea of that motivation.
By the way, I’d like you to start labeling your topics in the future to avoid such a misunderstanding. Specifically, on all those which you deem to be “an announcement”, which we’re not supposed to discuss, or argue about, or criticize, or question, or whine about—like the mini-camp results topic—please indicate as much. Thanks.
Cheers,
Silas
Yes, around the time I posted it. I don’t care if you find this implausible; I didn’t find your claims of never having seen my requests plausible either. I don’t have time to substantiate this claim either; I’m too busy.
People usually expect comments on topics they start; I don’t know why you would expect otherwise just because it’s “an announcement post”. It’s the Olympus mentality that says, “I talk, you listen, replies not wanted.” It’s the mentality that replies defensively to any cross-examination of evidence you’ve presented.
I didn’t assign those high priority for a reply because others already said what I would have said in reply, and it was becoming clear that I was not alone in being suspicious of this evidence.
I also didn’t ask to see testimonials because I didn’t consider those strong evidence for what you’re claiming—like others noted, they don’t really distinguish you from any other self-help camp.
And I explained that you don’t get to play “evidence takes time!” while saying, “yep, I already have enough evidence to call this a success and ignore any questioning of this claim”.
What was there to reply to? “Oh, well, since you personally feel you had a good time, I guess I shouldn’t be suspicious that it improved anyone’s rationality”?
I said I would pay the ransom you’re demanding for your evidence. That was the only reason I was willing to help you. I never withdrew any offer; rather, you tried to change the topic to helping with your research, which I don’t want to do, never did, and never offered to.
Those are regarding a different thread and different claim; there’s nothing to explain here, and I’d appreciate if you didn’t misreport evidence.
Sure thing—and you can keep it in mind too, when using increasingly strong superlatives to describe our success.
Despite multiple requests to drop this discussion, I’d like to put a little more effort toward mutual understanding. Perhaps I’m irrationally optimistic for reconciliation and convergence.
Others have addressed the unproductive ‘attack-mode’ nature of your comments; I won’t address that here. Suffice it to say that I have plenty to learn myself about communicating diplomatically.
I also won’t say much more on the issue of my not having seen your earlier calls for evidence of minicamp’s success. I can only repeat: If you want to be sure I’ll read a particular comment, make sure you contact me directly or reply to one of my comments so that your comment shows up in the LW inbox. I do not have time to keep revisiting old posts and reading all comments made on them, and I kinda doubt anyone thinks that is the best use of my limited time when I could instead be doing research and academic outreach related to rationality and FAI theory. You may insist on attributing this to my ‘Olympus Mentality’, though I’ll try to dissuade you of this interpretation below.
As for your definite accusation that I lied when I said I hadn’t seen your earlier comments on the topic, it remains the case that I never replied to them, and they seem like comments I would have replied to given my well-documented defensiveness on LW. Just notice how tenaciously I’ve defended myself in this discussion, despite a continuous slew of character attacks.
As for your accusation that I strawmanned you, I tried to explain that unless I had a policy of checking tons of old posts for new comments it’s not clear I would have seen your original comment, but you seem to simply disagree, so I don’t think there’s much more to say about that.
Finally, you seem to have suggested that I said I made announcement posts “we’re not supposed to discuss, or argue about, or criticize, or question, or whine about—like the mini-camp results topic”, but that’s just not true. You’re welcome to discuss, argue, criticize, question, or whine about anything I post on Less Wrong. All I said was that I don’t go back and check every post for new comments, and that if you want to make sure I read something you should contact me directly or be sure to reply directly to one of my comments so that I see it in my LW inbox.
A ‘Successful’ Minicamp
jsalvatier has repeatedly suggested that we may have different ideas of what I meant when I wrote that Rationality Minicamp was a success.
As KPier wrote in response to what seems to be your original comment on this topic, “The article pretty clearly states that the claims about the effects of the camp were based on exit surveys, and that the impact of the camp is demonstrated by the projects the camp grads are now working on. You could debate whether those are good measures, but we don’t exactly have better ones.” Later, Anna and myself gave that specific evidence in more detail.
You might be willing to concede that the evidence from exit surveys and testimonials provide about as much evidence of minicamp ‘success’ as such measures are capable of providing, though that may not be much. Is that true?
But of course, you’ve been asking for stronger evidence. You’d like to see measures of rationality improvement or life success or something like that. I addressed this exact request directly in my very first comment on the topic:
You replied that if these stronger forms of evidence don’t yet exist, then I shouldn’t claim that minicamp was a success. But again, I must repeat what KPier originally told you: My original blog post on minicamp being a success made it clear that such ‘success’ was assessed based on exit surveys and participant testimonials:
You seem to have interpreted ‘success’ in a different way than it was used in that blog post, perhaps to mean something like “Rationality minicamp successfully improved the rationality and life success of its participants, as demonstrated by several quantitative measures.”
But as the original blog post shows, that’s not what was meant to be claimed by calling the rationality minicamp a ‘success’.
Now, I’ll be happy to make this clearer by editing the original blog post, and by asking Eliezer to edit his post above. We could call it a ‘highly praised’ or ‘well-reviewed’ minicamp where brevity is needed, and where we have more space we could say something like “The minicamp was well-received by participants, who rated it highly in our anonymous exit survey and have given glowing reviews and reports of their resulting self-improvement. Further evidence concerning the minicamp’s effect on participants’ rationality and life success are pending.”
As for the fact that this data is still being gathered because it takes time for people’s lives to change and it takes time to parse collected data, you appear to have called this an “implausible excuse,” though I still don’t know what’s implausible about it.
Or perhaps what you meant to call an “implausible excuse” is my point about how the raw exit survey data is anonymous and private, and that’s why we can’t publish it. But I’m not sure what’s implausible about that, either. You can ask the minicamp participants themselves: We asked them to fill out one form that would be anonymous and private (the exit survey form), and another that would be identifiable and public (the testimonials form).
You also said I flip-flopped between these two “excuses”, but that’s not true. I maintain both claims. It takes time to collect and parse data on life changes and we can’t publish the private and anonymous exit form data.
Olympus mindset
You keep finding things that you choose to interpret as demonstrating my ‘Olympus mindset’ without addressing disconfirming evidence like what I gave above:
I’d also be curious to hear from others who think I display an ‘Olympus mindset’, and what triggers they think give them that impression. I don’t want to be giving off an inaccurate impression of myself in that way. I still practice facial expressions in the mirror because my face sometimes doesn’t clearly communicate my mindset, and obviously I still need to practice my online communication because my typed words don’t always clearly communicate my mindset, either.
EDIT: This has become an unproductive flame war, with no small thanks to my own behavior, and I will now bow out.
Support!
Okay, I’ve done both these things.
Thanks, I appreciate it.
I disagree with this. My intuition, supplemented by experience in somewhat analogous religious retreats, is that change happens easily in the camp environment and the question is how much of that will be inculcated enough to survive once the return to life happens.
I’d say it takes time to be sure people’s lives have changed permanently, but not too much time for them to change.
Luke wins the flame war! Huzzah!
Yes he did! The smarter and tougher man won.
So if minicamp related posts had instead of ‘success’ said ‘we were very pleased with the execution of the camp and it gives us a promising direction to explore’ would you have felt similarly to how you feel now?
If so, why are you still arguing about whether SIAI has prematurely judged the minicamp instead of asking them to make their judgment less ambiguous?
They don’t seem to agree that they should have done anything differently, so I don’t know why I would be at the point of asking them to change something. But yes, I would appreciate characterizations of the mini-camp that aren’t extremely misleading, please pass it on through someone who they’ll listen to.
It seems rather unlikely to me that being a mini-camp participant would have more of an effect on someone’s life than being a Visiting Fellow, new techniques or not—and if I am wrong, I would very much want to encounter these new techniques!
I wouldn’t be that surprised. Explicit rationality exercises were only starting to be developed during the last month of my stay, and at that point they mostly fell into the category of “entertaining, but probably not hugely useful”. The main rationality boost came from being around others with a strong commitment to rationality, but as situationist psychology would have it, the effect faded once I was out of that environment.
The positive endorphin rush from you and lukeprog sends signals that loook just like the enthusiastic gushing I see from any week-long “how to fix your life in five easy steps!” seminar. Smart people get caught up in biased thinking all the time. I had a good friend quit AI research to sell a self-help book, so I may be particularly sensitive to this :)
Objective data means I can upgrade this from “oh bunnies, another self-help meme” to “oooh, fascinating and awesome thing that I want to steal for myself.” As long as it signals like a self-help meme, I’m going to shoot it down just like I’d shoot down any similar meme that tried to sell itself here on LessWrong.
All right, but there’s a fine line between shooting down self-help memes and unnecessarily discouraging project-builders from getting excited about their work. It’s not fun or helpful for a pioneer to have his or her every first step be met with boundless skepticism. Your concerns sound real enough to me, but even an honest concern can be rude, and even a rationalist can validly trade off a tiny little bit of honesty for a whole lot of politeness and sympathy.
Why do I say “a tiny little bit of honesty?” Well, if the minicamp were being billed as “finished,” “polished,” “complete,” “famous,” “proven,” or “demonstrably successful,” as many self-help programs are, then it would make sense to demand data supporting those claims.
Instead, the PR blurb says that “Starting on May 28th, the Singularity Institute ran a one-week Rationality Training Camp. Our exit survey shows that the camp was a smashing success, surpassing the expectations of the organizers and the participants.”
Leaving aside the colorful language that can and should characterize most press releases, this is a pretty weak claim: the camp beat expectations. Do you really need to see data to back that up?
“Please give us money” and “Co-organized and taught sessions for a highly successful one-week Rationality Minicamp” are stronger claims.
For me, this isn’t about making SIAI transparent; it does quite enough in that regard. It’s about stopping an information cascade genie that’s already out of the bottle.
Let me put it this way: right now the ratio of “relying on the assumption of mini-camp’s success for decision making” to “available evidence for its success” is about 20-to-1. As I warned before, it’s quickly becoming something “everyone knows” despite the lack of evidence (and major suspicions of many people that it wouldn’t succeed going in). And that believe will keep feeding on itself unless someone traces it back to its original evidence.
It doesn’t reassure me that I’m told I have to keep waiting before anything’s conclusive, yet they can declare it a success now.
I just want the reliable evidence they claim to have, rather than just dime-a-dozen self-help testimonials. They collected hard data, and I gave them a list of things they could provide that are easy to gather and don’t compromise privacy, and are much more likely to be present if the success were real than if it were not. Even after AnnaSalamon’s circling of the wagons I don’t see that.
I think this is largely a case of people reading different things into ‘success’.
Oh, sure. The reason is easy to communicate. We explicitly told minicampers that their feedback on the exit survey would be private and anonymous, for maximal incentive to be direct and honest. We are not going to violate that agreement. The testimonials were given via a separate form with permission granted to publish THAT data publically.
I’m unclear, then, why you are citing a lack of staff hours if the data cannot be published at all.
Has the raw testimonial data been published?
I’m assuming you have data beyond exit surveys and testimonials...?
A summary of the data can be published, for example median scores for measured values. But the data can’t be published in raw form.
Not sure if raw testimonial data has been published yet. We do have data beyond testimonials and exit surveys, but that, too, requires precious staff hours to compile and write up, and it is still in the process of being collected.
Typing this stuff from a phone, pardon the brevity...