Let’s say you’re interested in transformers. In a big city, you could go to a meetup, talk to different people working on this stuff at a production level (maybe there’s even some guy who works on it at Google), talk about tips or pitfalls that no one ever publishes and potentially have the conversation veer off to DuckDB or some obscure topic. When you have a gathering of like minds, the conversation can go in unpredictable directions and you can potentially land in very interesting places.
In a less urban area, this is far less likely to happen because there are just fewer people with the same interests (unless you’re lucky). I grew up with friends who were absolutely brilliant (high fluid intelligence) who come from farming families. But they just weren’t interested in what I was interested in. The core of intellectual pursuit isn’t just smart people, but the confluence of people who have the same interests and who are also well positioned.
You might ask, can’t you just learn this stuff online and talk to people on Reddit? But the reality is that positive effects of randomness require real life undirected interactions with the right people.
(This is for instance why people are willing to relocate to a cold city like Montreal (-30°C/-22°F in the winter) to work in Bengio’s lab for a couple of years, just so they can overhear lunchtime conversations about how to train certain models. A lot of this knowledge is caught and not taught.)
I’ve gotten so many ideas from just random conversations with well positioned people who happened to be doing something important and interesting. It’s not just about being smart.
> In a less urban area, this is far less like. The core of intellectual pursuit isn’t just smart people, but the confluence of people who have the same interests and who are also well positioned.
The idea that spontaneous interactions lead to innovation is wonderfully captured in Kevin Simler’s essay Going Critical
Did you not experience the golden age of ICQ, irc, msn chat? It is absolutely possible to form strong friendships entirely online - it just takes the same amount of time and effort to nourish them. I still have fond memories of friendships with the people I used to chat with a couple hours almost daily, and we still know what's going on in each other's lives 20 years later!
Consider the collections of letters from snail mail friendships some classic authors used to maintain. Also purely done in writing.
It sounds to me like you’re just describing the limitations of textual communication, which I agree is limited in all the ways you have written in this post except the last.
But, it’s entirely possible to have thoughtful, deep, honest discussions with individuals over a textual medium and to develop meaningful relationships in that way. I have done so. Often these relationships start in a more open public setting and become more meaningful in a private space, similarly to real life.
Let’s say you’re interested in transformers. In a big city, you could go to a meetup, talk to different people working on this stuff at a production level (maybe there’s even some guy who works on it at Google), talk about tips or pitfalls that no one ever publishes and potentially have the conversation veer off to DuckDB or some obscure topic. When you have a gathering of like minds, the conversation can go in unpredictable directions and you can potentially land in very interesting places.
In a less urban area, this is far less likely to happen because there are just fewer people with the same interests (unless you’re lucky). I grew up with friends who were absolutely brilliant (high fluid intelligence) who come from farming families. But they just weren’t interested in what I was interested in. The core of intellectual pursuit isn’t just smart people, but the confluence of people who have the same interests and who are also well positioned.
You might ask, can’t you just learn this stuff online and talk to people on Reddit? But the reality is that positive effects of randomness require real life undirected interactions with the right people.
(This is for instance why people are willing to relocate to a cold city like Montreal (-30°C/-22°F in the winter) to work in Bengio’s lab for a couple of years, just so they can overhear lunchtime conversations about how to train certain models. A lot of this knowledge is caught and not taught.)
I’ve gotten so many ideas from just random conversations with well positioned people who happened to be doing something important and interesting. It’s not just about being smart.