The primary point of this post was to illustrate Double Crux, not to actually highlight the outcome of the discussion.
Which is fair enough, but—the outcome of a Double Crux is a pretty important part of it! (Right??) Any illustration has got to include it. After all, you say this at the start of your post:
SilentCal put together the Productive Disagreements Practice Thread, where people could post controversial opinions and then resolve them via Double Crux.
If the “and then resolve them” part is missing, then one may reasonable question the whole point of the exercise, yes? After all, we hardly need any special rationality technique just to have some possibly-productive-possibly-not exploratory banter about something…
Anyhow, your updated summary is very helpful to me in forming an opinion about what I should take away from this demonstration, so, thank you. :)
>If the “and then resolve them” part is missing, then one may reasonable question the whole point of the exercise, yes?
That… was contained in the original post though. (If you hadn’t read it except for the summary, I actually have a mild sense that the post would be better without the summary, to force people to actually read it. The summary isn’t an intrinsic part of the Double Crux, that’s just the part that makes it easily digestible to outsiders. The point of the post was to demonstrate what the back-and-forth-and-eventual-resolution-looks-like)
I read it. I just didn’t see a resolution in it. You say:
The point of the post was to demonstrate what the back-and-forth-and-eventual-resolution-looks-like
But actually, where is the resolution? Even with the summary (which, again, is helpful indeed), I see none!
… which is not a criticism of your post, not exactly. I mean, I certainly didn’t expect to see a resolution; I was just pointing out that indeed there is not really one (contra your suggestion that indeed there is).
… is it possible that we have different ideas about what constitutes a “resolution”?
Huh, I thought that last 2 back-and-forths between gjm and I were basically nothing but resolution.
There wasn’t a summary of all of the resolution from throughout the post, but I had a pretty clear sense of where we had ended up agreeing, what single crux we had isolated (which was specified in one of my final posts), and what the next steps were to operationalize that crux into an experiment if we wanted.
I’m assuming/guessing that by resolution you meant “explicit summary of the outcomes of the conversation”. Did you mean something else?
Well… let me put it this way. If the outcome of the conversation / exercise is substantial enough to need a summary, and that summary of the outcome itself runs into multiple paragraphs, then, yeah, we have some different expectations.
Or… how about like this:
Imagine a table (aside: I’d really love it if the commenting software here supported tables…), with two rows and two columns. The rows: Raemon and djm. The columns: Before and After. In other words: “Should HPMOR be on the front page? Before the Double Crux exercise, Raemon said yes and gjm said no. After the Double Crux exercise, Raemon said __ and gjm said __.” What goes in the blanks?. (In other words, I am looking for a mere 2 bits of information here!)
(Of course, there are some possible outcomes that don’t quite fit into that very narrow framework, such as “we decided the question was malformed” or “one or both of us is now agnostic on the matter” or something along those lines. Still, I should not expect even this class of outcomes to be so complex that they can’t at least be indicated with a single sentence!)
A large chunk of the point of Double Crux, IMO, is that questions substantial enough to have this sort of disagreement are almost always malformed.
The point of Double Crux is to reduce one question “should HPMOR be on the front page?” to another question that we both agree would answer the first question (“Did more than X% of people who read HPMOR radically have their life changed for the better?”)
Thanks!
You say (in your updated bit):
Which is fair enough, but—the outcome of a Double Crux is a pretty important part of it! (Right??) Any illustration has got to include it. After all, you say this at the start of your post:
If the “and then resolve them” part is missing, then one may reasonable question the whole point of the exercise, yes? After all, we hardly need any special rationality technique just to have some possibly-productive-possibly-not exploratory banter about something…
Anyhow, your updated summary is very helpful to me in forming an opinion about what I should take away from this demonstration, so, thank you. :)
>If the “and then resolve them” part is missing, then one may reasonable question the whole point of the exercise, yes?
That… was contained in the original post though. (If you hadn’t read it except for the summary, I actually have a mild sense that the post would be better without the summary, to force people to actually read it. The summary isn’t an intrinsic part of the Double Crux, that’s just the part that makes it easily digestible to outsiders. The point of the post was to demonstrate what the back-and-forth-and-eventual-resolution-looks-like)
I read it. I just didn’t see a resolution in it. You say:
But actually, where is the resolution? Even with the summary (which, again, is helpful indeed), I see none!
… which is not a criticism of your post, not exactly. I mean, I certainly didn’t expect to see a resolution; I was just pointing out that indeed there is not really one (contra your suggestion that indeed there is).
… is it possible that we have different ideas about what constitutes a “resolution”?
Huh, I thought that last 2 back-and-forths between gjm and I were basically nothing but resolution.
There wasn’t a summary of all of the resolution from throughout the post, but I had a pretty clear sense of where we had ended up agreeing, what single crux we had isolated (which was specified in one of my final posts), and what the next steps were to operationalize that crux into an experiment if we wanted.
I’m assuming/guessing that by resolution you meant “explicit summary of the outcomes of the conversation”. Did you mean something else?
Well… let me put it this way. If the outcome of the conversation / exercise is substantial enough to need a summary, and that summary of the outcome itself runs into multiple paragraphs, then, yeah, we have some different expectations.
Or… how about like this:
Imagine a table (aside: I’d really love it if the commenting software here supported tables…), with two rows and two columns. The rows: Raemon and djm. The columns: Before and After. In other words: “Should HPMOR be on the front page? Before the Double Crux exercise, Raemon said yes and gjm said no. After the Double Crux exercise, Raemon said __ and gjm said __.” What goes in the blanks?. (In other words, I am looking for a mere 2 bits of information here!)
(Of course, there are some possible outcomes that don’t quite fit into that very narrow framework, such as “we decided the question was malformed” or “one or both of us is now agnostic on the matter” or something along those lines. Still, I should not expect even this class of outcomes to be so complex that they can’t at least be indicated with a single sentence!)
> such as “we decided the question was malformed
A large chunk of the point of Double Crux, IMO, is that questions substantial enough to have this sort of disagreement are almost always malformed.
The point of Double Crux is to reduce one question “should HPMOR be on the front page?” to another question that we both agree would answer the first question (“Did more than X% of people who read HPMOR radically have their life changed for the better?”)