• 0 Posts
  • 81 Comments
Joined 1 year ago
cake
Cake day: July 1st, 2023

help-circle









  • But why bother with creating a new language, and duplicating all the features your language already has, in a weird way?

    If I want a list of UI items based on an array of some data, I can just do items.map(item => 〈Item key={item.id} item={item} /〉), using the normal map function that’s already part of the language.

    Or I can use a function, e.g. items.map(item => renderItem(item, otherData)) etc.

    JSX itself is a very thin layer that translates to normal function calls.







  • You need to look at this from a practical standpoint.

    The vast majority of phone apps are not local-only. They are merely the frontend to services provided by some company - e.g. a Reddit app is really about Reddit the service, a food delivery app is about the service, not the locally running code, etc.

    Apple controls what users can and cannot install on devices made by them, but the web and things like PWA are an alternative that would be viable for some portion of these.

    You can make a web app that can be added as an icon on the homescreen, can access the camera, location, notifications, storage, authentication (e.g. require fingerprint), etc. It still can't do everything native apps can do, but it would be good enough for a good portion of popular apps.

    But in China, that is not really possible without the government's approval either, because China requires the same kind of registration and an ICP license for websites, otherwise things will get blocked. Which, even if you could install anything you want on a device, would effectively limit you to purely local-only apps anyway.





  • Now, there was a paper that instantiated a couple dozen LLMs and had them run a virtual software dev company together which got pretty good results

    Dude, you need to take a closer look at that paper you linked, if you consider that "pretty good results". They have a github repo with screenshots of some of the "products", which should give you some idea https://github.com/OpenBMB/ChatDev/tree/main/misc .

    Not to mention the terrible decision making of the fake company (desktop app you have to download? no web/mobile version? for a virtual board game?)

    (Also the paper never even tried to prove its main hypothesis, that all this multi agent song and dance would somehow reduce hallucinations and improve performance. There is a lot of good AI stuff coming out daily, but that particular paper - and the articles reporting on it - was pure garbage.)