Links for November

Networks, OpenAI, Institutional Rot, Wonderdrugs, Wild West

Thanks for reading! And special thanks to my new paid subscribers!

As always, if you’d like to support Bullfish Hole you can leave a tip at this Stripe link (you can edit the amount) or with Paypal. If you’re not already one, you can become a free or paid subscriber at the link below.

Now, on to some items of interest from around the interwebs.

Trunk or Treat

I didn’t read this in time for the October post, but at Systematic Hatreds, Paul Musgrave talks about the politics of trunk or treat.

If you’re not familiar, it’s a form of trick-or-treating where people gather their vehicles in a parking lot and kids go from car to car for candy. We took our kids to one this year at the local community center, and it was like trick-or-treat mixed with tailgating: People decorated their cars in various themes and some even set up a grill to make hotdogs and such.

It’s certainly a lot more convenient than going to door to door if you live in a rural area, as we do. But Musgrave argues the practice doesn’t have rural origins, cropping up first as a church-led movement in mid-sized cities and suburbs. He gives several reasons it might have caught on, including safety concerns and the physical geography of modern communities.

Social Networks

Rob Henderson writes that on surveys, people tend to report that others go to more parties, have more friends, and have bigger social networks than themselves. He connects this to something called the friendship paradox: On average, your friends have more friends than you do. This is due to the structure of social networks, where some people are much more highly connected than others. The social butterflies, the lynchpins of community — since they have so many ties, they tend to show up more often in other people’s networks. We are actually more likely to know people who are more highly connected than ourselves.

It reminded me of this TED talk by sociologist Nicholas Christakis. He explains how we can take advantage of this fact to sample people who are closer to the center of social networks. Take a sample of people, ask them to nominate a friend, and then take a sample of the nominated friends: The second sample will, on average, be more central than the first. Since the central people are first to get whatever is spreading through the network — including diseases — having a sample of them gives you early warning that a big wave of illness (or whatever else) is coming.

Institutional Rot

This piece from Lorenzo Warby at Not On Your Team But Always Fair examines how a society’s institutions can degrade over time.

For example, while meritocratic institutions have a lot of advantages, they too get janky with age. One way they decay, writes Warby, is by selecting for capacity over character: Over time, manipulative social climbers learn how to game the system to its detriment.