5 ms·
Similarly, this is an astonishing (and infuriating) interview with SF board of ed president Gabriella Lopez: https://www.newyorker.com/news/q-and-a/how-san-fran
by jessedhillon 6y ago
Similarly, this is an astonishing (and infuriating) interview with SF board of ed president Gabriella Lopez: https://www.newyorker.com/news/q-and-a/how-san-francisco-renamed-its-schools https://www.newyorker.com/news/q-and-a/how-san-francisco-ren...
Even for SF, this is extreme.
Side note, for years I used to see the name "rayiner" and roll my eyes at the "reactionary libertarian" opinions expressed. In the past year though, I myself have come much closer to where you've consistently been. Part of it is the rapid leftward drift of major institutions -- media, workplace, education -- during the Trump years, and the distinct absence of any kind of critical discussion of this phenomenon. The other part is probably that I'm now at the point of having children soon. Either way, I now feel that the most sane and practical thing to advocate for is that there should be less institutional power available to be captured by extremists. In this case, it should matter a lot less that the school board is composed of lunatics, or that the NYT is trending towards naked culture war propagandism.
- ALittleLight 6y agoI think it's clearly better for our incompetent leaders to have less power than more, but better still would be to have competent leaders with adequate and no more power.
- geofft 6y ago> Either way, I now feel that the most sane and practical thing to advocate for is that there should be less institutional power available to be captured by extremists. I agree with this too! Which is why I think it's important the article calls out the institutional power of YC and the risk of it being captured by extremists! Frankly, leave the question of connection to white supremacy (under whatever definition) aside - as someone who was a regular Less Wrong reader, seeing Y Combinator and the other agents of Silicon Valley power go "All right, we will invest heavily in building a friendly AI under our definition of friendly," just by itself, is kind of a worst-case scenario. I don't think the interests of Valley investors are quite fully aligned the interests of humanity at large, but it's Valley investors that have the most ability to make AGI happen, and they've decided to start work. The only reason I don't find it terrifying is that I don't really agree with the Rationalists' arguments about it. But they do, and they're smart people.