This article made me feel better about my relatively unorganized digital life and knowledge graph. The constant struggle of organizing information, versus actually using it to accomplish goals. https://borretti.me/article/unbundling-tools-for-thought
And another: whether text-predictors might take agent-like action to make future text easier to predict. You might think no, but consider the closely-related recommender systems, like the YouTube or Amazon algorithms...
Comment discussions: whether the structure of the transformer architecture is able to, in principle, carry out complex enough computations to simulate conscious minds.
Good discussion about the limits (or not) of simulators like ChatGPT. https://www.lesswrong.com/posts/MmmPyJicaaJRk4Eg2/the-limit-of-language-models
(Warning: The post is structured strangely. It starts by arguing for unlimited simulation power, and then counterargues. Don't give up early. The comments also have good discussion.)
That is: George gave Kathy his heart. The very next day Kathy gave George's heart away to Andrew. How does that … work?
(Figurative) hearts are not traditionally a transferrable asset. Kathy could give *Kathy's* heart away, but not George's.
I think about this a lot. Perhaps too much. https://web.archive.org/web/20171216072414/http://squid314.livejournal.com/332946.html
@fugueish@infosec.exchange Interesting. My intuitions about what non-technical people would be more comfortable with differ. I suspect nice OS/browser integrated software like 1Password is easier for them to use than copying and pasting from a cloud doc. (Mobile copy/paste and app switching is nontrivial!)
But, I guess both are harder than using the same password everywhere. So probably they'll continue to do that.
For a White Elephant or Secret Santa gift exchange with a limit of (say) ¥4,000, is 4000 × 1¥ coins the best gift or the worst gift? 🤔
Before you answer, remember https://www.pbs.org/newshour/economy/the-economics-of-wasteful-spending-the-dead-weight-loss-of-christmas
@stilkov This is a technically hard problem because it's equivalent to cross-site tracking. I.e., separate servers all knowing the user's same identity. And browsers are working hard to make that impossible.
@kevinswiber Can't rename old public APIs. But on the web, we no longer use any term but URL. https://url.spec.whatwg.org/#goals https://url.spec.whatwg.org/#url-apis-elsewhere