This post has nothing to do with rationality and is likely to scare away readers who haven’t the slightest idea what “Peano arithmetic” refers to. And calling your conclusion “obvious” sends a rather “obvious” message to a large subset of potential readers.
I have written many technical posts, and the next one I’ve got planned will be more difficult than this one. Several of those posts got promoted and I also got some comments like this one saying my half-baked ideas were a big part of what attracted them to LW. That’s probably a bit too much flattery, but still :-) To be fair, I also got some comments similar to yours. Eliezer has also written many technical posts like the cartoon guide to Löb’s theorem which is more complex than most of my posts, including this one.
I’d really like to continue writing about AI-related technical stuff on LW. Would removing the word “obvious” satisfy your complaint?
I think it would also help if you spelled out “Peano arithmetic” in the title instead of “PA.” After reading the title, I had no idea what this post would be about.
When reading a technical post, it would be nice to see something like an “assumed reader’s background” note before all the gory details. And maybe a non-technical summary for the rest of us.
For example, in your case you could state that the target audience is (say) a 3rd year math undergrad or higher, familiar with the concepts of <...> and <...>
Perhaps you could put something in the title or first line of the post to signal that it will be (a) only of interest to people who enjoy mathematical philosophy and (b) not indicative of the kind of writing normally seen here. The fact that similar posts of yours have attracted readers to LW is, I acknowledge, evidence against my previous comment.
Also please do omit the term obvious from future articles. It’s infuriating (for me at least) to read technical articles and not understand something the author has labeled “obvious”.
If possible you might want to motivate why whatever you have written provides insight into rationality. EY, I think, at least eventually always does this with his writings.
Perhaps you could put something in the title or first line of the post to signal that it will be (a) only of interest to people who enjoy mathematical philosophy and (b) not indicative of the kind of writing normally seen here.
Done, sort of.
Also please do omit the term obvious from future articles.
Okay,
It’s infuriating (for me at least) to read technical articles and not understand something the author has labeled “obvious”.
whuh? Your past comments indicate that you’re a college professor and have written a book on game theory! I’m surprised… Okay, point taken.
If possible you might want to motivate why whatever you have written provides insight into rationality.
Most of my technical posts are about the mathematics of decision theory and AI, not human rationality. That is also a traditional LW topic that predates me. In particular, I’m very interested in AIs that try to prove theorems, and this post is the sort of theoretical result that could be relevant to those. Also it’s relevant to my next post which will be about decision theory, if I don’t refute that result first :-)
If I see a bunch of stuff that I’m not familiar with and don’t immediately want to make the effort to become familiar with, I just click out of the thread. I think that if things like this composed a larger fraction of the posts here, then there might be a problem, but that there really isn’t one at the moment.
This post has nothing to do with rationality and is likely to scare away readers who haven’t the slightest idea what “Peano arithmetic” refers to. And calling your conclusion “obvious” sends a rather “obvious” message to a large subset of potential readers.
(I didn’t downvote your comment.)
I have written many technical posts, and the next one I’ve got planned will be more difficult than this one. Several of those posts got promoted and I also got some comments like this one saying my half-baked ideas were a big part of what attracted them to LW. That’s probably a bit too much flattery, but still :-) To be fair, I also got some comments similar to yours. Eliezer has also written many technical posts like the cartoon guide to Löb’s theorem which is more complex than most of my posts, including this one.
I’d really like to continue writing about AI-related technical stuff on LW. Would removing the word “obvious” satisfy your complaint?
Please do.
You could go all mathematician and use ‘trivial’. ;)
I think it would also help if you spelled out “Peano arithmetic” in the title instead of “PA.” After reading the title, I had no idea what this post would be about.
When reading a technical post, it would be nice to see something like an “assumed reader’s background” note before all the gory details. And maybe a non-technical summary for the rest of us.
For example, in your case you could state that the target audience is (say) a 3rd year math undergrad or higher, familiar with the concepts of <...> and <...>
Made some edits. Though writing a nontechnical summary is a little beyond my skills right now :-)
Perhaps you could put something in the title or first line of the post to signal that it will be (a) only of interest to people who enjoy mathematical philosophy and (b) not indicative of the kind of writing normally seen here. The fact that similar posts of yours have attracted readers to LW is, I acknowledge, evidence against my previous comment.
Also please do omit the term obvious from future articles. It’s infuriating (for me at least) to read technical articles and not understand something the author has labeled “obvious”.
If possible you might want to motivate why whatever you have written provides insight into rationality. EY, I think, at least eventually always does this with his writings.
Done, sort of.
Okay,
whuh? Your past comments indicate that you’re a college professor and have written a book on game theory! I’m surprised… Okay, point taken.
Most of my technical posts are about the mathematics of decision theory and AI, not human rationality. That is also a traditional LW topic that predates me. In particular, I’m very interested in AIs that try to prove theorems, and this post is the sort of theoretical result that could be relevant to those. Also it’s relevant to my next post which will be about decision theory, if I don’t refute that result first :-)
If I see a bunch of stuff that I’m not familiar with and don’t immediately want to make the effort to become familiar with, I just click out of the thread. I think that if things like this composed a larger fraction of the posts here, then there might be a problem, but that there really isn’t one at the moment.
(comment retracted because it sounded dismissive)