I have increasingly considered myself a philosopher in the naïve sense that I like encountering, engaging with, accumulating and imparting wisdom, for all of which I expected the rationalist community to be a nurturing and motivating one, and indeed, I would not say I was disappointed -- so I have been collecting momentum and (hopefully) skills for thinking in an insightful, interesting and nuanced way... Then, whenever I found ideas that I believed to be original and worth sharing, I would go on to discuss it with people in this crowd to pick it apart, find underdeveloped bits, flaws etc.
However, since this Summer, this has gradually become more fruitless in a specific way: I have started to encounter more and more frequently the same piece of advice in response. That is, that I am poking at something difficult and I better ask Claude.
There is both an honestly valuable and frightening side to this. On the one hand, humility and situational awareness about current Claude capabilities being increasingly likely superior to any response the addressed person may give in a reasonable amount of time is objectively indispensable to rational thinking and subjectively dear to me. Of course, what I shall focus on is the other hand.
Permit me just a tiny anecdote in the spirit of "how things used to be better in older times." An attitude about how intellectual cooperation can work that was very highly emphasized during my early years of (engineering) university was that any of us shall be able to receive anyone else looking for help with some problem, and they shall either provide useful help, or redirect to someone else who is more likely to do so. Eventually who is seeking assistance will either stumble upon the right person for them or figure out a novel solution. It is not the only value for the community, as in potentially each member, to have a lot of object-level knowledge, but it is also to know people with diverse expertise, different from yours and each other's.
What happens if there is an expert in our network who you perceive most likely to know better how to aid in the majority of cases when you get asked? It might be obvious that you will recommend them in the majority of the cases, i. e. disfavor recommending alternative experts in your network. That is just what the merit of said expert earns by having been helpful so many times. They are a marvelous person to know, worth introducing them to those who ask you. Those other experts, maybe not so much.
Does this line of reasoning simply translate to the case when the outstanding expert is unhuman? In particular, how often can we be reassured that it is worth discouraging a new social bond for someone in pursuit of the subject's shortest path to sorting out what they see as a predicament?
I for one am slowly coming to the realization that the treasure I am looking for are friends I can make along my way, and I would not mind if that is a bit more of a long and winding one, for exactly that reason. But this goes beyond me. I am sure that natural social networks' global properties have relied on a micro-scale structuring principle, e. g. as in the Granovetter hypothesis, that is being disrupted.
Now there is a new structuring principle on the scene and I would not be surprised to learn that actual social science experiments can expose how individuals' neighborhoods in social networks are transforming, and the theory that predicts what kind of phase transition that drives society towards be already published.
A caveat to mention: my case and motivation may be extreme for 2 reasons. As I take my philosophy seriously, I try to foster those acquaintances and friendships where I can most openly bring my thoughts, to the point that whether or not I can may end up deciding the fate of my connection. The other is that because of talking to rationalists, I mostly talk to people who are optimistic about the trends in Claude's capabilities. Both of these coalesce to make me unusually susceptible to the process described here.
Well, or I may not be so unusual after all, you be the judge.
I have increasingly considered myself a philosopher in the naïve sense that I like encountering, engaging with, accumulating and imparting wisdom, for all of which I expected the rationalist community to be a nurturing and motivating one, and indeed, I would not say I was disappointed -- so I have been collecting momentum and (hopefully) skills for thinking in an insightful, interesting and nuanced way... Then, whenever I found ideas that I believed to be original and worth sharing, I would go on to discuss it with people in this crowd to pick it apart, find underdeveloped bits, flaws etc.
However, since this Summer, this has gradually become more fruitless in a specific way: I have started to encounter more and more frequently the same piece of advice in response. That is, that I am poking at something difficult and I better ask Claude.
There is both an honestly valuable and frightening side to this. On the one hand, humility and situational awareness about current Claude capabilities being increasingly likely superior to any response the addressed person may give in a reasonable amount of time is objectively indispensable to rational thinking and subjectively dear to me. Of course, what I shall focus on is the other hand.
Permit me just a tiny anecdote in the spirit of "how things used to be better in older times." An attitude about how intellectual cooperation can work that was very highly emphasized during my early years of (engineering) university was that any of us shall be able to receive anyone else looking for help with some problem, and they shall either provide useful help, or redirect to someone else who is more likely to do so. Eventually who is seeking assistance will either stumble upon the right person for them or figure out a novel solution. It is not the only value for the community, as in potentially each member, to have a lot of object-level knowledge, but it is also to know people with diverse expertise, different from yours and each other's.
What happens if there is an expert in our network who you perceive most likely to know better how to aid in the majority of cases when you get asked? It might be obvious that you will recommend them in the majority of the cases, i. e. disfavor recommending alternative experts in your network. That is just what the merit of said expert earns by having been helpful so many times. They are a marvelous person to know, worth introducing them to those who ask you. Those other experts, maybe not so much.
Does this line of reasoning simply translate to the case when the outstanding expert is unhuman? In particular, how often can we be reassured that it is worth discouraging a new social bond for someone in pursuit of the subject's shortest path to sorting out what they see as a predicament?
I for one am slowly coming to the realization that the treasure I am looking for are friends I can make along my way, and I would not mind if that is a bit more of a long and winding one, for exactly that reason. But this goes beyond me. I am sure that natural social networks' global properties have relied on a micro-scale structuring principle, e. g. as in the Granovetter hypothesis, that is being disrupted.
Now there is a new structuring principle on the scene and I would not be surprised to learn that actual social science experiments can expose how individuals' neighborhoods in social networks are transforming, and the theory that predicts what kind of phase transition that drives society towards be already published.
A caveat to mention: my case and motivation may be extreme for 2 reasons. As I take my philosophy seriously, I try to foster those acquaintances and friendships where I can most openly bring my thoughts, to the point that whether or not I can may end up deciding the fate of my connection. The other is that because of talking to rationalists, I mostly talk to people who are optimistic about the trends in Claude's capabilities. Both of these coalesce to make me unusually susceptible to the process described here.
Well, or I may not be so unusual after all, you be the judge.