The most interesting things to read to start the new year

Dear Friends,

I want to start with a wonderfully strange essay from Ben Lerner, “The Hofmann Wobble,” on Wikipedia and the nature of the Internet. You realize you’re in an odd place in the first paragraph when the author writes, “I remember, wrongly, that I was listening to a book on tape,” and then describes the book. Then he steps into an impossible mass of crickets. Soon, he’s haunting Berkeley, California, creating fictional personas and sock puppets and rewriting Wikipedia 14 hours a day to advance his beliefs, correct historical injustices, change our understanding of the Iraq war, manipulate his students, and perhaps do the dirty work of a tech billionaire. The details of his editing, and the nature of Wikipedia, are too specific to believe that the story is all fiction. It’s also, though, not clearly factual. And it does make you think differently about the world’s most important encyclopedia, which is also one of the key underlying sources for our modern large language models.

I was taken in, too, by Tom Scocca’s disquieting essay about the medical mystery that afflicts him just as his career in the journalism industry he’s thrived in takes a turn as well. (Scocca is the author of many, many classic Internet pieces, including a famous work on Smarm that inspired this rejoinder by Malcolm Gladwell, back when I was the editor of NewYorker.com.) Another disquieting piece I couldn’t put down is this one from the WSJ, about a man who falls apart while trying to succeed as an Instagram wellness influencer. And this story, by Molly Minta, about life in a Mississippi county with more than a thousand residents but just 42 adults with college degrees.

For interviews, I recommend reading this delightful short conversation with Jill Lepore and this chat with Yann LeCun, one of the most interesting people in AI. Then there’s this wonderful podcast in which the philosopher John Gray defends and defines pessimism to the optimist Tyler Cowen. Here’s an excerpt from one of Gray’s answers:?

“For me, there’s no providence of any kind in history. There’s no logic in history, although particular situations may have a logic of their own. But the logic, of course, may not be benign. It may be, to use your word, absurd. That’s to say, we may find human beings recurrently trapped in situations in which what they do is bound to produce results different from, or even opposite from, the ones they want. I think that is a recurring human situation. There’s no logic like Hegel thought or Marx thought or even Mill thought … There’s no logic in which history develops through a series of successive stages to some higher and higher levels … I would say the highest point of recent science, recent physics, might be a recognition that the world is finally unintelligible or absurd.”

I was also captivated by this story about a possible murder in the Argentinian mountains, this bonkers crypto scam, Katherine Wu on the American chestnut tree, and this lovely essay about rejected book covers. And here’s a very smart AVC essay on predictions for 2024, along with an argument for a convergence of AI and blockchain. Plus: “Innovation never waits for rules and regulations. But it eventually gets them.”? Over Christmas, I picked up an extraordinary book off my mother's shelf, which I had read many decades ago: The Education of Little Tree. Then, as a friend pointed out after I read it, the author is not at all who he seemed to be.

One more note on tech, due to a recommendation from my friend Chris Anderson, I’ve started listening to The Cognitive Revolution podcast, a superb series with tech builders. I recommend this one with Shane Legg and this one about the drama at Open AI and some stories you haven’t heard. Oh, and a reminder, my friends at the New America Foundation are now accepting applications.

This section is sponsored by Elastic, whose search analytics platform securely harnesses search-powered AI to unlock insights from massive amounts of data instantly. Elastic's platform helps any business drive tomorrow’s outcomes today.?

Back to AI: usually, I interact with chatbots the way I interact with a search engine. I ask the models to help me understand complicated issues or to help me write stories for my kids. But there are lots of people in the AI field who believe this type of use isn’t the future. In their view, we won’t primarily be interfacing with big, public LLMs. We’ll be talking to private LLMs trained on our data and designed to serve our needs.

Maybe, for example, we’ll have personal virtual psychologists, as?this Scientific American piece explains. Maybe they’ll be confidants, as?Kyle Chayka writes in an excellent article in the New Yorker. Or maybe they’ll be lovers.

I endorse some of this future. I’m fully in favor of chatbot tutors: private systems trained to help kids learn math or to help me improve my Spanish. So perhaps there will be times when a well-trained system can help people in need, as I discussed in this conversation with Karlie Kloss and Mustafa Suleyman. But there are also use cases that seem like nothing but trouble. Developing an emotional relationship with a bot seems like a recipe for manipulation. Just read this story about a bot that allegedly played a role in convincing a man in Belgium to commit suicide to prevent climate change.?

I love AI, and I use it every day. I’m obsessed with understanding the benefits and the risks, and I think the former outweigh the latter. I worry, though, about manipulation and anomie. We’re just now starting to untangle the ways that social media, and phones more generally, have harmed our mental health. If you have a kid with a phone, I strongly suspect there was at least one moment over the holiday break when you were tempted to put it in the microwave.

Best wishes, and happy reading, for 2024.?

cheers * N

I hope you enjoy this newsletter, which, thanks to your support, now has more than 400,000 readers. Please continue to forward it to anyone else who might enjoy it. They can?sign up here.