How thinkers and writers must compete with LLM's

Longreads + Open Thread newsletter: 2025-May-24 by Byrne Hobart

The Patel interview is too long for my patience, but I love Byrne's comment! I added the emphasis.

Nilay Patel interviews Microsoft CTO Kevin Scott. This interview mostly focuses on the economics of AI, specifically the question of how anyone will make money online if AI agents are circumventing ad-supported models, and are trained on tokens they haven't paid for. He seems confident that we'll work something out, which will still have implications for the online content business: it means that being able to write at the same skill as an LLM is basically worthless, but being able to produce things that they can't, and that will influence their behavior, is very, very worthwhile.

How thinkers and writers must compete with LLM's

Longreads + Open Thread newsletter: 2025-May-24 by Byrne Hobart

The Patel interview is too long for my patience, but I love Byrne's comment! I added the emphasis.

Nilay Patel interviews Microsoft CTO Kevin Scott. This interview mostly focuses on the economics of AI, specifically the question of how anyone will make money online if AI agents are circumventing ad-supported models, and are trained on tokens they haven't paid for. He seems confident that we'll work something out, which will still have implications for the online content business: it means that being able to write at the same skill as an LLM is basically worthless, but being able to produce things that they can't, and that will influence their behavior, is very, very worthwhile.

Make a plan to think better

Cars (and before them, horses) reduced some people's ability to walk long distances. Unless they had a plan to avoid the degradation of that capability. I think we all need a plan to think even when we don't have to do so. 

Svblte: Thoughts on thinking, 2025-May-16 by Dustin Curtis

The irony is that I now know more than I ever would have before AI. But I feel slightly dumber. A bit more dull. LLMs give me finished thoughts, polished and convincing, but none of the intellectual growth that comes from developing them myself. The output from AI answers questions. It teaches me facts. But it doesn’t really help me know anything new.

While using AI feels like a superhuman brain augmentation, when I look back on the past couple of years and think about how I explore new thoughts and ideas today, it looks a lot like sedation instead.

And I’m still stuck. But at least I’m here, writing this, and conveying my raw thoughts directly into your brain. And that means something, I think, even though an AI could probably have written this post far more quickly, eloquently, and concisely. It’s horrifying.

Voicing ideas

This article has me thinking about making more voice memos and notes. 

Ness Labs: Thinking Out Loud: How to Use Your Voice in Knowledge Work, 2025-May-15 by Anne-Laure Le Cunff

Reduce cognitive load. Psychologist Alan Baddeley’s model of working memory suggests a “phonological loop” — the part of the brain that holds spoken and written material. Speaking activates this loop, reducing cognitive load and freeing bandwidth for deeper reasoning.

Improve clarity. Studies show that explaining ideas aloud (even to yourself) improves understanding — a phenomenon known as the self-explanation effect — which can boost problem-solving and retention.

Strengthen memory. Reading aloud or repeating things vocally has been found to improve encoding in long-term memory. This is linked to the production effect, where memory is better for words that are spoken than those silently read.

A bunch of reasons NOT to use AI

Like Marie, I've been given a bunch or reasons to use AI. Instead of replying, I usually just keep quiet. I don't like to argue, and I'm pretty sure anything I say is going to be contested--maybe even ridiculed. So I'm glad Marie applied her energy to explaining why it's not necessarily a great idea. Here's just one of her 11 different reasons. 

Young Vulgarian: 11 things I hate about AI, 2025-May-9 by Marie Le Conte

Some proponents of generative AI will argue that going back and forth with the computer helps creativity, as opposed to stifling it: as someone who has to be creative a lot of the time, I just don't think it's entirely true. Sometimes you do need to force yourself to do the hard thing a lot of the time, so that you're then able to do the easy thing quicker.

It's also a good habit to pick up, as starting from nothing is, I think, good for the brain. I see it a lot in oil painting, which I started doing last year. There's nothing as horrifying as a blank canvas, but the first steps you take when faced with it are often the most important ones. Which colour should I use to make my wash? What am I mixing to begin with, for that first layer? Am I drawing some outlines in charcoal first, or just vibing it?

By taking these first few decisions, I start building the foundations of whatever project I'm embarking on. It's a different exercise from, say, paint-by-numbers, and there's a reason why one of those is a more respected field than the other.

Today's Young Couples Have Little Room in their Lives for Babies

I worry that people are going to have fewer intergenerational connections than we had in the past. How can that NOT hurt the cultural divide??

Tangle: Would a baby bonus boost the birth rate?, 2025-April-29 by Isaac Saul:

I think there are plenty of good reasons for the government to pursue pro-brith policies, but I don’t think they’ll be effective in encouraging more births — the decreasing fertility rate seems so much deeper than just a financial question. 

This is, in my view, primarily a social and cultural issue: Young Americans live different lives than they did 50 years ago, with different goals, different dreams, and different pursuits. Both parents are often focused on their careers, which diasporizes their family and friend village. On top of that, young Americans have growing numbers of peers who don’t want kids so they can be free to travel, socialize, and pursue their careers without limitations. The upshot is that people on the fence about having kids today feel more societal pressure not to than in years past. 

We’re going to need to turn those cultural pressures around if we want to meaningfully impact fertility rates. 

 

And ways to have fun with AI

Found via Adam Tinworth's newsletter, One Man & His Blog... 

Om Malik used ChatGPT to (successfully) brainstorm a colour palate for some art. What's notable is the fun he had with the experience:

What surprised me most was not the machine’s intelligence (which is easy enough to marvel at), but the quiet way it handed back a piece of the world. Its palette didn’t describe April in grand terms. Instead, it was poetic and almost literary, taking cues from the light, the soil, and the ordinary shapes of change. In doing so, it reminded me that the season is not something we wait for but something to observe and experience.

Which raises some interesting questions:

You could never have this kind of interactive fun with “search” as we have known it. And that is why, the new way is going to be a problem for Google.

Of course, what makes this work is that Om is looking for ideas not answers. That's the right way to use a guessing machine.