widdershins

I've seen recent AIs make detailed and technically impressive 3D models. You might argue "they're not doing spatial reasoning, they're making measurements with code and doing math to configure relative positions". Fine, but at a certain point that becomes functionally indistinguishable from spatial reasoning.

Yeah, I'm not an archery nerd and even I could see that the archery looked very poor compared to other movies like Lord of the Rings. There didn't seem to be any weight to the draw in most casts.

Agreed. Work through the Emacs tutorial to find out how vanilla editing works. Then look at buffer management, window management (Emacs 'windows' are actually more like frames in other apps) and simple tooling like search. Then start to tweak some settings in your config, find the bare minimum of packages you need to scratch your worst itches (for me: Vertico, Corfu, Avy and Dumb Jump). Finally, figure out tree-sitter modes, project/projectile mode and a couple of other foundations.

It will be a struggle. It was around 2 months before I felt remotely comfortable in Emacs. And nearly a year before I really felt at home. It's a long road, but gradually you mold the editor to yourself so tightly that you'll never be able to go back. The remarkable thing is that the progression never stops. The tool just keeps getting sharper and sharper.

Britain and Portugal have one of the world's oldest alliances, established in 1386.

At various times, especially starting in the 1700s, British wars with France meant that French wine and brandy became unavailable, or was even banned. In reponse, British merchants discovered that fortified wine from the Duoro valley could sate the British market's taste for strong wines, and could survive the long journey across the Bay of Biscay and the Channel. They began setting up port wine trading houses which became known as 'shippers'.

Fun fact: Britain's longest-serving Prime Minister, William Pitt the Younger, started drinking port at 14 as a cure for gout. He apparently drank a bottle a day for most of his life. Port exacurbates gout. He died at age 46.

It's more minimalistic, that's true. But there's nothing stopping you writing or downloading an array library so you can do this:

  enemies = array.filter(enemies, function(e) return e.alive end)

Or even setting a metatable so you can do:

  enemies = enemies.filter(function(e) return e.alive end)

I signed up and I'm finding browsing a repo quite snappy. Good job.

Neither the title nor the article implied so.

This is fantastic. I don't know if it tells us anything about the human brain, but it's so cool to be able to do emperical science on large neural networks like this.

Yeah, this speed is excellent! I'm using GPT-5 mini for my "AI tour guide" (simply summarizes Wikipedia articles for me on the fly, which are presented on my app based on geolocation), and it's always been a ~15 second wait for me before streaming of a large article summarization will begin. With GPT-5.4 it's around 2-3 seconds, and the quality seems at least as good. This is a huge UX improvement, it really starts to feel more 'real time'.

It's working reasonably well for me. But this is inside a well-established codebase with lots of tests and examples of how we structure code. I also haven't used it much for building brand new features yet, but for making changes to existing areas.

> I imagine an LLM would be very good at writing Elisp to leverage EMacs’ strong core functionality

Yes, they are pretty good. I have set up GPTel (an excellent Emacs package for interacting with LLMs) with some tools allowing it to run Elisp, inspect files (Elisp functions know what file they were defined in, so it's easy to find stuff) and read Emacs documentation. LLMs use this to good effect, and iterate on my config very nicely.

I'm not familiar with the Web Audio API, but I am interested in audio APIs in general. Could you explain where the designers went wrong? The templates listed in this (very nice!) app seem to look relatively sane as these things go.

Ask it to cite the document and page, and verify it yourself.

It lasted because React genuinely did introduce a useful new paradigm - of UI being a (mostly) pure function of state. This has actual benefits. Unfortunately the way this was implemented was (very) suboptimal, and now they're locked in.

I agree that this stubborn clutching to that original idea, despite better options that acheive the same paradigm having arisen in the meantime.