Karma as Money
How do you gather a theory of Counterfactuals, Karma, and Economics, into a revised algorithm for thinking about Lesswrong?
Thinking of Karma as money.
There are a lot of things that one may consider worth saying on Lesswrong. Things that go against the agenda, things that may make people unconfortable, things that are different from what the high-ranking officials would prefer to read here. But we don’t do it, because we don’t want to “loose” precious Karma points. Each Karma point loss is felt as an insecurity, as a tiny arrow penetrating the chest. But should it be that way?
Here is the alternative: Think of Karma as money. You work hard for getting a few karma points by writing interesting stuff on superintelligence and whatnot, society rewards you by paying some karma points. Then you go there and write something you think people need to hear, but will downvote for sure, at least initially. Some people by now will be very rich, which affords them the opportunity of saying a lot of things that they are not sure will get themselves upvoted, but are sure should be posted.
Citizen: Wait, you said counterfactuals...
Yes, just like your State doesn’t really care or like you going out in your hovercraft through the river and using equipment to climb a mountain, so the people here may not care about putting attention into that idea which you think they should hear. Thus, they dowvote it. They make you pay for their attention. If you mentalize it as “they are drawing my soul and life is worthless if karma is negative”, then you are much less likely to end up posting something controversial that may be counterfactually relevant.
Just like efficient charity donation works because the vast majority of people are not paying to effectively cause others into being happier, using karma as money works because the vast majority of people are afraid their soul is being sucked every time a downvote comes. But it isn’t, this is just the price people charge for their attention, if you think the way I’m tentatively suggesting. It is just a test worth trying, not necessarily something that I fully endorse. I like the idea, and have been using it since forever. Every post linked here, or an earlier subpart of it, has been negative at some point, and from before posting, I knew it would be a “costly one”. Try it, if you are rich, you may have nothing much to loose, and more controversial but useful stuff will show up with time.
Let’s see how much this costs.
It’s not that hard to build up enough karma that it would be unfeasible to “spend” it all on heavily downvoted posts. If Karma is money, it’s money that can only be used to buy something that most people who have a lot of it don’t want anyway.
If you’re sure that a topic should be posted about, it’s a good idea to write about it in such a way as to minimize the chances of it being downvoted, otherwise your post will probably be less useful than it could have been.
I agree with your advice, but I suspect that many folks have interesting, relevant, reasonable quality insights that they don’t write posts or comments about because they are worried about downvotes.
If the OP causes some of those posts to be made that wouldn’t have been made, then it probably provides net benefit to LessWrong. That said, I completely endorse your second paragraph. There’s no reason to court downvotes that could have been avoided.
Ditto
I don’t agree. I don’t agree with many downvotes, and consider them bad behavior. It’s a bad idea to reward bad behavior by giving in to it.
I think once I passed a thousand karma I basically stopped caring about it. To date, I’ve made two exceedingly self-indulgent comments, fully expecting them to get heavily downvoted, and it’s hard to care about either of them.
I can report that I have a great many things I think might greatly contribute, but never post anything because I am absolutly unable to handle even a low risk of geting downvoted due to mental helth issues. I utterly despise the karma system for this reason and it has basicaly robed LW of my entire public contribution, wich could have been great. I routinely use only PMs to avoid these issues when I think I can help somone specific, but often I dont know who to contact.
What if you had a script that hid downvotes? Or one that used an unknown algorithm to hide all downvotes and also some upvotes so a lack of upvotes could still mean that you really got upvoted? Or maybe one that added fake downvotes so that you would never have certain knowledge of having been really downvoted? Etc. Personally I like your commenting and would enjoy more of it.
because a) The very rationality habits I’ve learnt here finds it extremely anaesthetic to hide information from myself like that. b) My brain would default to assuming everything that might have been downvoted has so, and thus only have problems more often. c) The culture that having the voting system in general causes would still be there, and what really hurts is the status loss from people wanting to downvote me, not losing the arbitrary point total.
Why not post on your own blog, and occasionally link worthwhile content from open threads?
Having your own blog is pretty much LW without the karma.
Because then his links could be downvoted.
I have a G+ account and a twitter, and find myself never using them. Turns out I don’t have anything to say to the audience “people who follow me personally”, and if I use it only for what you propose I might as well use Pastebin, with is really silly.
If karma strongly matters for your actions, I would recommend to do rejection therapy in real life. In the large scheme of things karma just doesn’t matter.
Would you recommend the same therapy for anything that people care about but doesn’t ultimately matter in “real life”?
I don’t think that people participate in lesswrong to gather karma. They generally participate primarily to have conversations about rationality or other topics that get discussed on this website.
What do you mean with “real life”? I recommend “real life” rejection thearpy, because getting rejected online is easy and thus no good training.
I don’t think that online discussions don’t matter. Participating on lesswrong can give you useful knowledge that matters a lot more than your lesswrong karma score. I also do believe that ideas matter and that it matters to convince others of ideas.
If anyone has anything they want to say, but are afraid to because of karma, feel free to message me and I might post it for you.
Obviously if it is actually stupid I won’t post it, but if you’re just nervous I can take the weight off your shoulders (in return for any possible karma gain.)
I think I missed a part of the meta-discussion: Are there any consequences for having a high or low karma, other than position on the leaderboards?
Some readers may take you a little more seriously if you have very high karma, or a little less seriously if you have very low karma.
In practice, though, if you have very high karma then the chances are that you’re active enough that those readers will recognize you and form whatever opinions they do more on the basis of experience than of your karma score, and the state of having very low karma doesn’t tend to persist very long: someone who posts things other LW participants like will soon not have very low karma, and someone who doesn’t will usually notice and give up. (Or improve.)
I have to click through to your user page to look at your karma score; do a statistically significant fraction of readers check karma before evaluating a post?
I doubt anyone does it often. Probably many people never do it at all. (I think I’ve been curious enough to bother once or twice, ever.)
Hardly; I think you need a minimum of 10 karma to create new posts, and if you have negative karma you can’t comment more than once per five minutes or something like that (the time might depend of your karma, and it may be 30-day karma and not total karma). Apart from that, nothing.
You also need to pay 5 karma to comment on highly down voted posts (I think this means posts with −5 karma or more).
And you can’t downvote more than 4 times your total number of upvotes (or net upvotes, I can’t recall).
Negative net karma can also reduce the rate at which you can post.
It’s not like money because it’s not scarce. I could generate karma easily by saying the right things on new posts, if I ever felt the urge to, which I never will because it’s accreted anyway with no specific effort toward that goal. I’ve abandoned two other accounts to change my username and the reset to 0 karma didn’t affect me one bit; I’ve never tried giving away all of my savings, but it would probably change my day-to-day life a lot.
Karma is a numeric representation of how people reacted to your post, including the votes by people who don’t feel compelled to respond verbally—a substitute for the voting by attentive gazes or intentional ignoring that would serve the same purpose in meatspace. It’s body language for the internet.
False dichotomy. Karma correlates a lot more with the quality of writing than with the topic.
The biggest thing karma correlates with is number of posts/comments made, unless they suck.
I meant karma per post, naturally.
Karma per comment is correlated mostly with how many people read the comment, provided that it’s witty or insightful or funny or correct enough to be considered mostly upvoteworthy.
Getting such a comment in quickly (before a post is promoted) on a popular post is worth much more than that same comment made a few days after promotion.
But the amount, and the quality, of the attention you get from someone downvoting what you’ve written is typically very low, and it’s not any higher on account of the downvotes. I think the analogy you’re trying to make simply isn’t there.
Wow. There’s a set of assumptions in here that are quite foreign to the way I think of LessWrong.
Karma is not money, as it cannot be exchanged for anything. I’d be interested to know how many people care about their karma level at all, beyond the basic thresholds of being able to post and comment when they like. I seriously doubt it’s a “vast majority”.
More importantly, having lots of karma can’t be spent to get attention. You can make lots of posts, but there’s no cost to anyone to ignore them (and downvote them). It may be possible to hurt the community a little by posting a lot of crap before you run out of karma (which could be a long time if you exploit a karma pump on some threads), but I suspect it won’t cause much harm, and you’ll mostly be ignored.
Still more importantly, you’re not going to get very much value out of Less Wrong if you focus on things you’d like others to hear, as opposed to things you want to learn. It’s not a group you can preach to, it’s a group that will help you find holes in your beliefs and patch them.
Karma is a way of keeping score in a game that one may or may not be playing.
At 71% positive, I’ve yet to receive a downvote or upvote that was anything other than N++ regarding my opinion of the Less Wrong community.
N++?
My read on that was “increment N” in c-like syntax, i.e. “me too”-ism, where N is the number of people professing an opinion.
Increment N is right, but reference statistics “number of data points”.
e.g. “Less Wrong voters value posts which use community passwords and applause lights, P~.85, N=20”
Right, got it.
Really? You’ve never been downvoted when you thought it was unjustified?
I don’t use karma to keep score, only as one of many sources of information about other people’s opinion. When one of the comments I’ve posted gets a much different response than I expected, it is my expectation that was at fault.
Let me be more concrete.
I don’t cross-post my blog posts in the Open Thread because I don’t think they are on topic.
By contrast, I don’t post on certain topics related to rationality, mindkilling, and social interaction because I expect downvotes, regardless of quality or on-topic-ness.
If I understand correctly, you only post when it is both on-topic and you expect positive karma? Are there two different terminal values in play here?
Generally, yes.
I also post points in favor of feminism and post-modernism when I think they are on topic and would enhance the rationality of the average LW user. But I expect those topics to get downvoted, so I don’t always make such posts, even when I think the expected value of the post would be positive for readers.
Really? You let imagined karma downvotes keep you from posting quality material? That’s no way to live.
He lets expected karma downvotes keep him from posting quality material, not imagined downvotes.
If the purpose of karma is not to influence posting behavior in a positive way, then what IS its purpose?
Expected downvotes are still just imagined downvotes.
And Karma’s purpose is Karma’s problem; not Tim’s, and not mine.
Tim thinks he has quality material to post that he expects some people to vote down. On balance, I don’t think Tim’s purposes are served by this. He’s letting himself be controlled by some handful of people he disagrees with, for no significant gain.
Writing a quality post takes effort that is orders of magnitude more costly than 3 thumbs down. If it’s worth spending that effort in the first place, a handful of downvotes should not change that that—they’re insignificant by comparison.
Tim actions are not consistent with his conscious values, as estimated by me.
You might be right. I think my best defense is that your analysis of expected gain fails to account for opportunity cost. I could go troll a religious forum. And I could do so in a way that would enhance the rationality of the average reader (if believed / accepted by that reader). But that caveat is pretty huge when evaluating opportunity cost.
If I’m understanding you correctly, the expected/imagined/whatever thing is a red herring; your comment applies just as well to actual downvotes. Your claim is Tim ought not take downvotes into consideration when evaluating the value of a post, whether they be expected downvotes on expected posts, imaginary downvotes on imaginary posts, or actual downvotes on actual posts.
Yes?
Karma’s purpose is not unrelated to this question. If net karma approximates an expression of the community’s judgment of the value of the post to the community (which it’s supposed to, though it’s not clear it does) then net downvotes indicate the community judges that a post is of negative value to the community. Tim might update away from his belief that the post is valuable based on that judgment. (And, relatedly, update away from it based on his expectation of that judgment for an unsubmitted post.)
Of course, you might say in turn that the judgment of the community (whether actual or expected) should itself not cause Tim to change his belief.
It’s not a red herring, it’s another reason.
No. Downvotes are information. Don’t ignore them. But don’t let them keep you from what you consider the right thing. The context was of him not posting what he considered quality material, anticipating downvotes.
I don’t suppose that at all. Do you?
I generally behave as though net karma does mean that, under the principle that this is one way I can encourage it to more closely approximate meaning that over time, and I would prefer it did so.
Note that one reason to be wrong about how people will respond to a given stimulus is to overestimate how much of a desirable trait that they have. So while “your expectation is at fault” to the extent that correctly predicting response is a goal, the update in such circumstance is still in the direction of less respect for the people you are attempting to predict. That can be disheartening at times.
(For what it’s worth I’ve found that by doing lots of commenting experiments on LessWrong it’s actually possible to develop a fairly detailed model of LW karma attribution habits, even to the point of, say, noticing which geographic regions have which voting habits. This is sorta costly, maybe, but makes “karma is like X” approximations unnecessary.)
After banking some karma, negative karma points don’t inhibit me, whether for posting what people need to hear, or as a result of responding to downvoted posts.
In fact, negative karma when I think I’m right are extra sparkly karma points to me.
I still like to get positive rubs for a good idea or a well turned phrase, but negative karma has little effect. If I didn’t mean it, I wouldn’t have said it in the first place. And I’m self confident enough that disapproval of others is unlikely to convince me I’m wrong. Karma is a blunt instrument to express disapproval. Was it a stupid idea, obviously wrong? Did you dislike the tone? Disagree ideologically?