I… don’t think I understand what this has to do with my comment? I agree that it’s not overwhelmingly obvious, but what does that have to do with my comment (or Christian’s for that matter?).
I guess maybe this whole thread just feels kinda confused, since I don’t understand what the goal of Christian’s comment is.
Christian responds to your comment with “you want to divert resources away from the thing you usually work on” (with possibly an implication that Christian cares a lot about the thing he thinks you usually work on, and doesn’t want fewer resources allocated to it.)
You respond “huh that’s a weird way of phrasing it. why would it be ‘divert away?’ we’ve always worked on other projects”
That seemed, to me, to be a weird way of replying, because, like, theory of mind says that ChristianKI doesn’t know about all those other projects. If you assume the LessWrong team mostly builds LessWrong, it’s quite reasonable to respond to a query about “what stuff should we build other than LW?” with “that sounds like you’re diverting resources away from LessWrong”. And a more sensible response would have been “ah, yeah I see why you’d think that if you think we only build LessWrong, but actually we do other projects.”
(moreover, I think it actually is plausibly bad that we spread our focus as thinly as we do. I don’t think it’s an obvious call because the other projects we work on are also important and it’s a reasonable high-level-call for us to be “the rationality infrastructure team” rather than “the LessWrong team”. But, a priori I do feel a lot more doomy about small teams that spread themselves thin)
I think it makes sense to have multiple installations of the same software, so that the EA-Forum and AI Alignment Forum as they reuse the code base. Code has a feature of often providing exponential returns and thus it makes sense to double down on good projects instead of speading efforts.
I… don’t think I understand what this has to do with my comment? I agree that it’s not overwhelmingly obvious, but what does that have to do with my comment (or Christian’s for that matter?).
I guess maybe this whole thread just feels kinda confused, since I don’t understand what the goal of Christian’s comment is.
My read was:
Christian responds to your comment with “you want to divert resources away from the thing you usually work on” (with possibly an implication that Christian cares a lot about the thing he thinks you usually work on, and doesn’t want fewer resources allocated to it.)
You respond “huh that’s a weird way of phrasing it. why would it be ‘divert away?’ we’ve always worked on other projects”
That seemed, to me, to be a weird way of replying, because, like, theory of mind says that ChristianKI doesn’t know about all those other projects. If you assume the LessWrong team mostly builds LessWrong, it’s quite reasonable to respond to a query about “what stuff should we build other than LW?” with “that sounds like you’re diverting resources away from LessWrong”. And a more sensible response would have been “ah, yeah I see why you’d think that if you think we only build LessWrong, but actually we do other projects.”
(moreover, I think it actually is plausibly bad that we spread our focus as thinly as we do. I don’t think it’s an obvious call because the other projects we work on are also important and it’s a reasonable high-level-call for us to be “the rationality infrastructure team” rather than “the LessWrong team”. But, a priori I do feel a lot more doomy about small teams that spread themselves thin)
I feel like Raemon got where I was coming from.
I think it makes sense to have multiple installations of the same software, so that the EA-Forum and AI Alignment Forum as they reuse the code base. Code has a feature of often providing exponential returns and thus it makes sense to double down on good projects instead of speading efforts.