Category: Blog

Alberini hypothesises that early unrecalled memories may function as schemas upon which adult memories are built. Like the foundations of a home, they remain concealed but crucial. — Read on www.theguardian.com/science/2025/feb/16/why-cant-we-remember-our-lives-as-babies-or-toddlers

Every American stands at the end of a continuing, decades-long effort to build and maintain the systems that support our lives — Read on marginalrevolution.com/marginalrevolution/2025/02/how-the-system-works.html

some people would think that reading a baby a book about farm animals is teaching them about farm animals, but really it’s teaching them about the concept of a book and how there’s new information on each page of a single object, but really, beyond that, it’s teaching them how language works, and beyond that…

all of Musk’s preening on social media obscures what is actually happening. That is what content is really good at doing. It feels transparent to see an influencer bake bread in her grandma chic retro kitchen or to see a billionaire storm a corporate headquarters to vanquish his enemies to unemployment. But content does not…

This from a comment: LLMs are not: intelligent. But most humans aren’t either. We keep comparing LLMs to the top 1% or 0.1% in a given field and laughing because, well, they can’t do the tasks the top in a given field can do. But LLM are handily crushing your average person at cognitive tasks.…

Michael Polanyi’s Republic of Science principle is a valuable guide that describes the way the scientific community ideally organizes itself so knowledge is generated and shared freely, leading to beneficial innovations. When scientists are free to work on problems that fit their abilities and interests and are well-informed about the work of others, they learn,…

How will the best current models be knit together in stacked, decentralized networks of self-improvement, broadly akin to “the republic of science” for human beings? — Read on marginalrevolution.com/marginalrevolution/2025/02/the-three-levels-of-ai-understanding.html

so many of the patterns we were using against Mechanical Turk a decade+ ago can provide hints about potential ways to apply LLMs. Ie. the way you would explain the task to a human assistant… — Read on simonwillison.net/2025/Feb/11/llm-sort/

I think once you get into the, “I’m not using this as a utility or for education, it’s not serving me in any way, it’s not a tool, it’s not building community, it’s fraying community, and I’m just doom scrolling,” then you’re outside of the CUE. You need to log off. — Read on kottke.org/25/02/is-social-media-good-for-you-apply-the-cue-test

See, “search” has never just been “search.” Search has always been an epistemic journey. You’re building knowledge, you’re not in and out with one query. You google first with a vague intention, you learn some of the vocabulary. You google more, back and forth, and you pick up the terminology and trade-offs – whether you…