Data point: seeing these predictions take over PB.com is killing what little motivation I had left for using the site.
Yes, we desperately need a tagging feature. Does anyone have any ideas on how to implement one?
Does anyone have any ideas on how to implement one?
Yes, many people do. Many others just need to google it.
For the record: I've been trying, on and off, to set up a development environment for PB for a while now, but always kept getting error'd because of version inconsistencies. But that was before I knew of Bundler. I'll try again in the near future.
Then again, I'm just starting out with rails and I have other things to do with my time, so don't expect any miracles.
EDIT: Retraction was accidental.
Yeah, I'm more busy with my start-up than I had anticipated (retarded, huh?) and had to put my study of Ruby on hold. Let me know if you need any testing help or other grunt-work.
'Take over'? Are they eating your predictions? Do you wake up and find your predictions have been infected with comments about MoR? Did you have a quota I didn't know about and these MoR predictions are filling up your hard drive? Hardly anyone of them have a due date in the next year, so you can't be complaining about their judgment or controversies, so what are you complaining about?
This isn't a complaint, it's feedback.
Specifically, for instance, what's happening when I view the "Happenstance" or "Recent predictions" lists is that I have to scroll, sometimes over a screen's worth, to get past a big chunk of predictions which I can tell are of no interest to me. This is demotivating (even though I have read HPMoR), a trivial inconvenience.
To me (and I can perfectly understand not everyone feels that way) the HPMoR predictions are noise, as are the private predictions such as "I will complete 100 push-ups by next month". I have no interest in refining my map of a fictional universe, or my map of a total stranger's motivational structure or muscular strength.
The appeal to me of PredictionBook isn't as a repository of predictions which are private to one person or that have a very narrow appeal. I can totally understand if someone else chooses to use it that way - though it's worth noting that its UI is not consistent with that usage.
What I'm expecting from PredictionBook is a diverse stream of somewhat challenging predictions which motivate me to take the risk of being wrong. What other PredictionBook users get in return is my occasionally contributing a prediction that makes the site more interesting, if by some epsilon value, for everyone else.
I'm not paying anything, so I don't have a right to complain about the service - I'm also free to quietly stop using it. I just thought it would be of some value to other users to give, as it were, an exit interview.
I'm not paying anything, so I don't have a right to complain about the service
Nonsense. You have a right to complain about anything. And some optimal prices are negative; there's nothing magical about a service you pay $0 to use (not counting the time you spend using it!) as compared to $0.01 or -$0.01.
No, he's right to complain. We really do need a tagging system so we can filter out the kinds of prediction statements we don't want to see. For instance, I resolved not to read anymore HPMoR (I stopped at about chapter 65) until I complete certain others goals (of which, I use PredictionBook to help with motivation), but it is very difficult for me to avoid spoilers with the site's current setup. I'm sure others have similar problems.
http://predictionbook.com/predictions/3237 which is currently
HP MoR: Voldemort never intended Harry’s own death on the night of his parents’ death.
says in a comment
ArisKatsaris changed their prediction from “In “Harry Potter and the Methods of Rationality” it will be revealed that Voldemort never intended Harry’s own death on the night of his parents’ death.” 2 days ago
Thing is I don't remember editing the text of that prediction myself. Now this is the sort of minor edit that might be thought to be utterly forgettable, but two days ago my Internet usage was extremely limited, and I'd not spend it editing predictionbook predictions.
It's highly unlikely that I was hacked with the sole purpose of changing "In “Harry Potter and the Methods of Rationality” it will be revealed that " to "HP MoR: ".
So is this some sort of glitch in predictionbook which falsely records me editing the prediction, when in reality it was edited by some moderator? I'd prefer this to be fixed, as I don't like actions wrongly assigned to me -- though the fact that this particular instance is harmless enough, makes this a non-priority for me in this particular instance.
Either way I'd like a clarification, if possible, so that I'm sure my memory isn't so glitchy as to forget this.
So is this some sort of glitch in predictionbook which falsely records me editing the prediction, when in reality it was edited by some moderator?
Yes, that was gwern. The site doesn't keep track of who changed the prediction, so it just assumes the author of the prediction ( = you) did it.
(The sheer amount of "they just didn't care" inherent in the site design boggles the mind.)
Though rarely updated, PredictionBook is open source. So at least you don't have to wait for the creators to fix it, if you know how to code. I wrote up an Issue for this bug, describing some starting points for fixing it.
Correct. I edited all the predictions to read 'HP MoR: ' for uniformity and searchability (once Google re-indexes them).
Suggestion: the person who memory charmed Rita Skeeter was Gilderoy Lockhart, who was the most famous master of Memory Charms ever mentioned in canon.
Digging through my posting history...
Voldemort is the product of Dementation, and is brain damaged. Dumbledore is in the early stages of dementia. Grindelwald will return as an antagonist to Dumbledore. The Source of Magic is a mildly unfriendly artificial intelligence. Names are needed for magic because the Source of Magic uses the names as pointers to the information in other people's heads. The first prophecy was faked by Dumbledore. Dumbledore caused and intended Sirius to betray the Potters. A person who's created a horcrux survives his death as the brain emulation running on the horcrux, not as a disembodied spirit. Quirrell is being possessed by a horcrux of Voldemort hidden on his person. The button Quirrell threw to Hermione was a horcrux. Voldemort will create a flesh and blood copy of himself from one of his horcruxes. Voldemort will defect against a magically produced identical copy of himself.
They're independent predictions.
That's, um, an interesting selection. Are you going to record them or are there a few which are particularly important I could record?
after doing something else involving Harry
Why does this sound more pervy than uploading part of his consciousness into Harry?
Suggested prediction: Gilderoy Lockhart will be mentioned on the basis of his reputation, but not make a personal appearance.
I'm sorry I might have missed it, but where did the notion of humans being turned into house elves come from?
"ELF!" [said the Sorting Hat]
Huh? Harry remembered Draco mentioning a 'House Elf', but what was that exactly?
Judging by the appalled looks dawning on the faces around him, it wasn't anything good -
IIRC, Anubhav thought it might be the penalty - to serve House Malfoy forever (bwa ha ha ha). And might explain why House Elves are so rare etc despite their apparent power.
Aw, I thought I was the first to suggest it -- http://lesswrong.com/lw/ams/harry_potter_and_the_methods_of_rationality/61e9
Maybe he got it from you? I know I sure didn't consider it likely enough to be worth recording.
In retrospect, I was putting far too much weight on Dumbledore's prognostication that Azkaban was out of the question. My model of Dumbledore didn't make mistakes like that.
Also, if the Wizengamut went around house-elfing people, that's pretty much guaranteed to make Harry declare war on them, which seemed like an obvious goal of the story arc even back then.
But I doubt it will come up in the story again, so unless EY weighs in, I'm not sure if we'll know one way or the other.
Possibility that this fiction concludes year one, with Harry and Voldemort still alive, and a second book picks up with year two?
What would a 'second book' be? A full-blown official sequel? (Without a resolution of the plot, I'd wonder if it really makes sense to dub any division the end of 'book 1' and the start of 'book 2'.) Might be better to predict about sequels in general.
I was thinking a full blown sequel, where part of voldie's plan is revealed, but not near all of it- we resolve the issues surrounding the stone, and perhaps gain some clearer insight into what dumbledore's motivation is, but the war is still on.
Hmm. Gwern, looking over a few of these, I feel like your percentages given are consistently too high. Then again, I guess we'll see how it works out in the end.
You're probably right. I have good handles on real-world outcomes, but for fiction there's so many freedoms that I think I anchor on initial suggestions too much.
The recent spate of updates has reminded me that while each chapter is enjoyable, the approaching end of MoR, as awesome as it no doubt will be, also means the end of our ability to learn from predicting the truth of the MoR-verse and its future.
With that in mind, I have compiled a page of predictions on sundry topics, much like my other page on predictions for Neon Genesis Evangelion; I encourage people to suggest plausible predictions that I've omitted, register their probabilities on PredictionBook.com, and come up with their own predictions. Then we can all look back when MoR finishes and reflect on what we (or Eliezer) did poorly or well.
The page is currently up to >182 predictions.