This matches my impression in a certain sense. Specifically, the density of gears in the post (elements that would reliably hold arguments together, confer local validity, or pin them to reality) is low. It’s a work of philosophy, not investigative journalism. So there is a lot of slack in shifting the narrative in any direction, which is dangerous for forming beliefs (as opposed to setting up new hypotheses), especially if done in a voice that is not your own. The narrative of the post is coherent and compelling, it’s a good jumping-off point for developing it into beliefs and contingency plans, but the post itself can’t be directly coerced into those things, and this epistemic status is not clearly associated with it.
How do you think Zoe’s post, or mainstream journalism about the rationalist community (e.g. Cade Metz’s article, perhaps there are other better ones I don’t know about) compare on this metric? Are there any examples of particularly good writeups about the community and its history you know about?
I’m not saying that the post isn’t good (I did say it’s coherent and compelling), and I’m not at this moment aware of something better on its topic (though my ability to remain aware of such things is low, so that doesn’t mean much). I’m saying specifically that gear density is low, so it’s less suitable for belief formation than hypothesis setup. This is relevant as a more technical formulation of what I’m guessing LoganStrohl is gesturing at.
I think investigative journalism is often terrible, as is philosophy, but the concepts are meaningful in characterizing types of content with respect to gear density, including high quality content.
I am intending this more as contribution of relevant information and initial models than firm conclusions; conclusions are easier to reach the more different relevant information and models are shared by different people, so I suppose I don’t have a strong disagreement here.
Sure, and this is clear to me as a practitioner of the yoga of taking in everything only as a hypothesis/narrative, mining it for gears, and separately checking what beliefs happen to crystallize out of this, if any. But for someone who doesn’t always make this distinction, not having a clear indication of the status of the source material needlessly increases epistemic hygiene risks, so it’s a good norm to make epistemic status of content more legible. My guess is that LoganStrohl’s impression is partly of violation of this norm (which I’m not even sure clearly happened), shared by a surprising number of upvoters.
Do you predict Logan’s comment would have been much different if I had written “[epistemic status: contents of memory banks, arranged in a parseable semicoherent narrative sequence, which contains initial models that seem to compress the experiences in a Solomonoff sense better than alternative explanations, but which aren’t intended to be final conclusions, given that only a small subset of the data has been revealed and better models are likely to be discovered in the future]”? I think this is to some degree implied by the title which starts with “My experience...” so I don’t think this would have made a large difference, although I can’t be sure about Logan’s counterfactual comment.
I’m not sure, but the hypothesis I’m chasing in this thread, intended as a plausible steelman of Logan’s comment, thinks so. One alternative that is also plausible to me is motivated cognition that would decry undesirable source material for low gear density, and that one predicts little change in response to more legible epistemic status.
If you are genuinely asking, I think cutting that down into something slightly less clinical sounding (because it sounds sarcastic when formalized) would probably take a little steam out of that type of opposition, yes.
This matches my impression in a certain sense. Specifically, the density of gears in the post (elements that would reliably hold arguments together, confer local validity, or pin them to reality) is low. It’s a work of philosophy, not investigative journalism. So there is a lot of slack in shifting the narrative in any direction, which is dangerous for forming beliefs (as opposed to setting up new hypotheses), especially if done in a voice that is not your own. The narrative of the post is coherent and compelling, it’s a good jumping-off point for developing it into beliefs and contingency plans, but the post itself can’t be directly coerced into those things, and this epistemic status is not clearly associated with it.
How do you think Zoe’s post, or mainstream journalism about the rationalist community (e.g. Cade Metz’s article, perhaps there are other better ones I don’t know about) compare on this metric? Are there any examples of particularly good writeups about the community and its history you know about?
I’m not saying that the post isn’t good (I did say it’s coherent and compelling), and I’m not at this moment aware of something better on its topic (though my ability to remain aware of such things is low, so that doesn’t mean much). I’m saying specifically that gear density is low, so it’s less suitable for belief formation than hypothesis setup. This is relevant as a more technical formulation of what I’m guessing LoganStrohl is gesturing at.
I think investigative journalism is often terrible, as is philosophy, but the concepts are meaningful in characterizing types of content with respect to gear density, including high quality content.
I am intending this more as contribution of relevant information and initial models than firm conclusions; conclusions are easier to reach the more different relevant information and models are shared by different people, so I suppose I don’t have a strong disagreement here.
Sure, and this is clear to me as a practitioner of the yoga of taking in everything only as a hypothesis/narrative, mining it for gears, and separately checking what beliefs happen to crystallize out of this, if any. But for someone who doesn’t always make this distinction, not having a clear indication of the status of the source material needlessly increases epistemic hygiene risks, so it’s a good norm to make epistemic status of content more legible. My guess is that LoganStrohl’s impression is partly of violation of this norm (which I’m not even sure clearly happened), shared by a surprising number of upvoters.
Do you predict Logan’s comment would have been much different if I had written “[epistemic status: contents of memory banks, arranged in a parseable semicoherent narrative sequence, which contains initial models that seem to compress the experiences in a Solomonoff sense better than alternative explanations, but which aren’t intended to be final conclusions, given that only a small subset of the data has been revealed and better models are likely to be discovered in the future]”? I think this is to some degree implied by the title which starts with “My experience...” so I don’t think this would have made a large difference, although I can’t be sure about Logan’s counterfactual comment.
I’m not sure, but the hypothesis I’m chasing in this thread, intended as a plausible steelman of Logan’s comment, thinks so. One alternative that is also plausible to me is motivated cognition that would decry undesirable source material for low gear density, and that one predicts little change in response to more legible epistemic status.
I expect the alternative hypothesis to be true given the difference between the responses to this post and Zoe’s post.
If you are genuinely asking, I think cutting that down into something slightly less clinical sounding (because it sounds sarcastic when formalized) would probably take a little steam out of that type of opposition, yes.