I realize (having been back and forth on this) that the answer is a very strong yes.

LLMs are based on language. The concepts/"nodes" LLMs tie together are useful because humans have created these nodes over the entire history of language, culture, and science. Humans naturally over millennia have been filtering out useful concepts from useless ones, refining language to efficiently carry meaning.

If we broke down all abstractions to fundamentals, there would be billions of things that would need to be tied together to make any useful product, almost impossible. Abstractions are required for solving problems efficiently. A high level abstraction that fits well in many solutions is insanely powerful.

LLMs are not immune to this effect, they need abstractions so that they don't have to prove every single thing they do from first principles, they can instead lean on language to shortcut a ton of thinking.

So when a framework gives you useful high level structure, patterns, and abstractions, you can do a lot less reasoning, and work a lot more efficiently, to solve a lot more problems. You could have one word (e.g. controller) mean a whole ton of useful things that you otherwise would've had to describe with multiple words. You could have references to existing concepts (e.g. `String#camelize`) that you no longer need to expand and define each time.

Bottom line, abstractions let LLMs solve problems faster, more correctly, with less reasoning.

P.S. Language is all abstractions, and we can solve problems well because we created these effective abstractions. The question I have about LLMs going forward is whether they can create new useful abstractions (new words, concepts, categories) that are actually time-tested and proven to contribute to good problem solving, without human input/verification.

But almost every comment here is incredulously pointing out how inept police is to not be able to solve such an easy/obvious problem. There is nothing easy and obvious about it.

We need one of those "what every programmer should know about time zones" articles about common police and military misconceptions.

Serious crimes are generally a low percentage stat in society overall, which is part of the reason why it's so hard to judge "oh, this isn't serious". Most police encounters aren't, except the ones that are.

But this seems like a good crime recipe:

1. Swat somebody for months.

2. Go into their house once police stops showing up.

We can all say “well, if you get many calls about one location, maybe you should stop showing up” but wtf are you supposed to do if a real crime is being committed? The same couple getting swatted one day could get criminals breaking in another, possibly for the same reason even. How can you make the call not to show up if your entire existence and role in society is to show up?

What real solutions are there?

Edit: a bunch of identical responses to this comment with "just don't show up". I know it's easy to sit and judge from the comfort of your keyboard, but do a little bit of thinking. What should the rule be? When is it okay for police to receive a distress call and not show up in actual reality with actual lives at stake? Who's responsible for the consequences of not showing up?

Edit2: now people switched to "show up, but softer, knock on the door". Now we can agree, because this comment is about showing up, not about specifics. Yeah, more considerate approach is warranted. To be clear, I agree that police should be punished for violating laws and conduct policies. Yet there is the other side of it, there's still an alert and it could still be real. Police still have to be mind-f'ed into figuring out every time if there's real danger or not.

Edit3: People who say "police can catch the callers" and suggest other triumphant movie-esque feats make me feel like we need an article "Falsehoods programmers believe about police".

No affiliation, just arrived at Paseo after a bunch of research.

    Those who claim LLM-generated software is good enough:

    Haven’t written code in ages
    Cannot spot if their code figuratively had 6 fingers!
    Have a low bar for what good looks like
    Don’t care about quality or NFR
    Have difficulty understanding an S-curve

> one cannot buy the technical competence

The most charitable interpretation here is "just by paying for AI tokens, you don't get smarter", with which I would agree. And yet your full comment seems to suggest that there's no way to buy technical competence in general.

But Omarchy is literally buying it[1]. And so does every business that pays tech employees.

[1]: https://omarchy.org/teams/

If there was a piece of software that needed 3 colors set, and there was a choice to set the 3 colors yourself, or pick a color preset that some person spent time combining just right, a lot of people would appreciate the latter.

Sitting and making subjective taste-based considerations is a huge energy and time sink, and takes a certain kind of individual to actually enjoy. Even if you did take time on one thing, you still hope that you don't have to do that on others. Having somebody sweat these details for you, so your baseline is strong, is a lot of value in practice, because we're human and we like nice things, and knowing that even if we don't customize it, somebody already did with great care.

Why even listen to premade music. You can just throw together samples/notes yourself. Why listen to lyrics/poems, you can just put together some words yourself. Why hire an interior designer. You can just make it the way you like yourself. Etc.

[1]: https://github.com/achiya-automation/safari-mcp#vs-apples-of...