Also masculinity self placement versus self reported gender (basically looks bimodal with a little noise in between)

Real examples of this are going to be things that seem boringly obvious

Probably not reported in cohen’s D terms but strength of relationships between thermometer ratings of a political party and voting for it are very high

I think very long distance trains (Orient express?) would plausibly have been able to induce mild jet lag

Especially compared to the physical impact of travelling for hundreds of hours at high speed on horse with no sleep,

You have to travel less far further north, so St Petersburg latitude seems sensible (and there's enough land to actually travel if you go east). That latitude line is about 14,000 miles, so about 583 miles per time zone. So I don't think anyone would ever have noticed on horseback.

You adjust to time changes at about 1 hour per day. The pony express managed to get up to 200 miles per day with horse relays (usually you would just pass the message but in theory a rider could go the distance).

FWIW I think I would differentiate politics as negotiation beyond dyads. So that takes pure 2 party market transactions out of scope but any time there’s at least 3 sets of distinct interests you get politics (most things that look dyadic are actually multipolar)

So this definition would then include arguing over what film to watch as a low stakes form of politics?

I don’t think power differentials are required because parties may and usually do hold different forms of power. In fact I would say politics is what happens when power is not absolute (eg the president is very powerful but has to contend with the power of the electorate to remove her)

This example varies the level of conflict rather than power

I feel like that’s just a case where stakes are lower. There’s still power that could be used (one person could pay the other, there might be a favor economy, perhaps there’s romantic interest, reputational damage from making a poor recommendation, ultimately the physical struggle over the remote)

What’s an example of conflict without power?

Yeah, they're oddly reticent to use their really good inbuilt visual capabilities and keep trying to do everything in python.

I love the message of "just look with your robot eyes", the frustrations of dealing with LLMs.

visual tasks are the most fun to code with AI because verification is so much easier

A problem I've encountered a bunch of times is when geographic data (such as ward boundaries) is only presented in a pdf map. I got chatGPT to extract that into usable shapefiles. Chat log here if anyone wants to extract this into a systematic workflow chatgpt.com/share/69d654...

Reading an older baby naming book with some interesting suggestions

After a great discussion with @pwgtennant.bsky.social I’ll amend to “non-policy focused academic analysis”

bsky.app/profile/pwgt...

I agree with this but I think it fits better with the general interest/curiosity motivation than being able to nail down impacts in advance

The interesting question is what is the justification for a question in astrophysics? there’s no concrete decision that turns on our understanding of black hole mergers and it’s not necessary for a functioning society for citizens to understand them

Concrete example would be that a lot of people believe that “if only we could get everyone to turn out to vote then all my preferred policy positions would get implemented”. This is usually empirically untrue and I think it’s good for people not to hold these kind of false beliefs about the system

Absolutely! But I think the key difference is that health has a top down interventionist worldview (the goal is to figure out what things experts can do to help people) whereas the implied intervention for a lot of social science is widely diffusing knowledge about how people and institutions work

So if you force an answer to the value of the work in terms of decisions, it’s going to look nakedly partisan or zero sum

the societally valuable thing we’re doing is producing more informed citizens/media/politicians who are savvy about how the system actually works. But I don’t think that’s expressible in decision terms in the same way

I think it’s good for citizens in a democracy to know how other citizens think and how parties and politicians behave towards them

I think there’s value in this knowledge being widely available and understood. Ie I think it’s bad if the only people who know how to do this are working behind closed doors

I think if the implied decision is targeted at make party A more popular and party B less popular that’s always going to be a tough sell.

I think economics has an easier time of this because it’s often studying things that are closer to consensus values (reducing unemployment, increasing economic growth, reducing inflation etc)

As additional context, social science has come in for a lot of criticism for being seen as the propaganda/research arm of the woke agenda. Ie insofar as there’s a normative consensus within the discipline it’s not one that is universally shared outside