Before getting to the Mailbag, I want to say something about the increasingly salient A.I. debates happening online. Trying to think about policy involves forming views about a lot of contested questions, many of which don’t have immediately obvious answers. That means you necessarily find yourself deferring to an extent to the judgment of other people. Which means that something very important to do in life is decide how to assess which people are reliable sources of information.
The way that I think a lot of people go about doing this is kind of basic friend/enemy politics. They emotionally or intellectually anchor around some particular point and then just try to align with the people who are aligned with them on that.
And I think that’s a mistake. The fact is that for basically any widely held political view, a huge share of the people who hold the view are going to be ignorant, intellectually sloppy, or otherwise badly flawed in their reasoning. That’s just what people are like.
What you ought to do instead is try to see which people casually toss off false statements about things that are easily verifiable. One salient example this week is that I’ve seen a lot of conservative-minded people who decided to align with the White House’s regulatory skepticism on A.I. safety saying that the people who are pushing concerns about the risks of A.I. are the exact same people as “climate doomers” and anti-nuclear activists.
Anyone who is saying this either doesn’t know what they are talking about or (worse) does know what they are talking about and is choosing to lie to you.
That’s not a good person to listen to on this subject. For better or worse, the relationship between the climate movement and the effective altruist community has been very antagonistic over the years precisely because the latter group did not accept the totalizing and apocalyptic narratives the climate people were invested in.
To be maximally generous to Erickson, he was just popping off without looking into it. That’s not ideal epistemic behavior, but I think we’ve all done it (certainly I have) and it can be useful as a learning experience. But if you pop off, get told you’re wrong by better-informed people, and then plow ahead unashamedly, that’s bad. Don’t do that. Don’t listen to people who do that.
The Digital Entomologist: Will Stancil and Parker Molloy got into it on Bluesky. Molloy’s theory seems to be that Talarico showed he’s transphobic, and it acceptable to work to reduce his support which could result in Paxton winning — possibly leaving a 50/50 Senate which remains in the thrall to the Trump personality cult. Her argument seems to be that the election is the best time to try to influence a politician. It is better to defeat the one who is closer to you, than to keep the pressure on for the next 6 years. How does post election influence work? How can one convince trans activists that they are better off talking to Talarico for the next 6 years than being ignored by Paxton?
I certainly agree that vote blue no matter who reduces the influence of those whose votes can be taken for granted during the election, but I also think that those voters still have a lot of influence during the term if they in part of the coalition. Or is that incorrect?
I think the big thing that has gone weirdly missing from a lot of progressive thinking about the political system is the concept of “persuade people that you are correct.”