I have one project that is very complex and for this I can't and don't use LLMs for.

I've also found it's better if you can get it code generate everything in the one session, if you try other LLMs or sessions it will quickly degrade. That's when you will see duplicate functions and dead end code.

Many of these are annoying LLM training/scraping bots (in my case anyway). So while it might not crash them if you spit out a 800KB zipbomb, at least it will waste computing resources on their end.

But outside of this if someone is determined they can issue fake documents at this level of provenance.

Drivers licenses for example you can buy the printing machine and blanks (illegally) so you actually need to check the registrar in that location.

Much of this feels like when they did studies on people who take mushrooms for example feel like they are more productive, but when you actually measure it they aren't. It's just their perception.

To me the biggest issue is that search has been gutted out and so for many questions the best results come from asking an LLM. But this is far different from using it to generate entire codebases.

It doesn't operate on the same level as (human) intelligence it's a very path dependent process. Every step you add down this path increases entropy as well and while further improvements and bigger context windows help - eventually you reach a dead end where it degrades.

You'd almost need every step of the process to mutate the model to update global state from that point.

From what I've seen the major providers kind of use tricks to accomplish this, but it's not the same thing.

Mind you in the early days pages used to have hundreds if not thousands of text links all over the place, the only sites that do this now are the hardcore conspiracy sites where the author just adds several new links a day.

So in this dimension at least web UIs have changed for the better.

I did forget you could do this, though I'm fairly sure in the case of a browser (at least it used to) this will block the entire UI.

Suppose as well there are a thousand people lined up to make the wooden part but hardly any for the metal, then surely the ones who work on the metal part will (try to) command a higher wage too.

In the first example: there is no such thing as a blocking sleep in javascript. What people use as sleep is just a promise wrapper around a setTimeout call. setTimeout has always created microtasks, so calling a sleep inline would do nothing to halt execution.

I do agree that dangling Promises are annoying and Promise.race is especially bad as it doesn't do what you expect: finish the fastest promise and cancel the other. It will actually eventually resolve both but you will only get one result.

Realistically in JS you write your long running async functions to take an AbortController wrapper that also provides a sleep function, then in your outer loop you check the signal isn't aborted and the wrapper class also handles calling clearTimeout on wrapped sleep functions to stop sleeping/pending setTimeouts and exit your loop/function.

As someone using Kafka, I'd like to know what the (good) alternatives are if you have suggestions.