But the model doesn't need to read the node_modules to write a React app, it just needs to write the React code (which it is heavily post-trained to be able to use). So the fair counter example is like:
function Hello() {
return <button>Hello</buttton>
}
Fair challenge to the idea. But what i am saying is that every line of boilerplate, every import statement, every configuration file consumes precious tokens.
The more code, the more surface area the LLM needs to cover before understanding or implementing correctly.
Right now the solution to expensive token limits is the most token-efficient technology. let's reframe it better. Was react made to help humans organize code better or machines?
Is the High Code-to-Functionality Ratio 3 lines that do real work > 50 lines of setup really necessary?
At current prices you can pretty much get away with murder even for the most expensive models out there. You know, $14/million output tokens. 10k output tokens is 14 cents. Which is ~40k words, or whatever.
The way to use LLM's for development is to use the API.
I'm not so worried about the money but more about context rot. I used spec driven development for a week and I had constant compacting with Claude code. I burned 200€ in one week and now I'm trying something different: only show diffs and try to always talk to me in interfaces.
I do think that at some point there will be frameworks or languages optimised for LLMs.
Tangent but the screen scratches mentioned from the keyboard are annoying. Every generation of MacBook has suffered from them and everyone’s seems to have it after a while, I guess it’s just physics when it gets stuffed in a bag etc, but any good tips to avoid it? I’ve heard mixed things about keeping a cloth in there
I'm on my third mac which has a glossy screen and tight tolerances. I won't sweat it.
The hardness of the coating is getting better with every generation. I'm on my M1 MacBook, which I'm using for ~4 years every day. The scratches are visible, but becomes almost invisible when I clean the screen. Moreover they're invisible when the screen is on. My 10 year old Unibody also has them, and while they're worse, they're not visible either.
The best defense is having a padded case and using it even carrying in a bag. Carrying a cloth can bend and damage the screen in worse ways.
I thought this was an absolutely fascinating look into reverse engineering hardware. If you’re a musician, check out the plugins at https://dsp56300.wordpress.com/ - you can bit perfect emulate famous 90s synths including Access Virus, Waldorf MicroQ and XT, Nord Lead and now Roland JP8000
- Imagine what you could do with 16-bit games of the era with one or more of these models embedded. Swap the model depending on the use case within the game. Great for adventures, RPGs, strategy, puzzle, and trading games (think Elite). With 512K or 1MB of RAM, plus 2 - 4 floppies (which became increasingly common as the era wore on), you could probably do a lot, especially if the outcomes of conversations can result in different game outcomes
- Back in the day nobody was really trying to do anything serious with AI on 8 or even most 16-bit machines, because nobody thought they were powerful enough to do anything useful with. Now the thinking has changed to how much somewhat useful intelligence can I cram into the least powerful device, even if that’s only for fun?
- Imagine showing this running on a CP/M machine, like the C128, to a serious AI researcher working back in the 1980s. Minds blown, right?
- Now spool forward 10 years into the 1990s and think what PC hardware of that era would have been capable of with these limited language models. I wonder what that era might have looked like with something that seems like somewhat useful conversational AI? A sort of electro-steampunk-ish vibe maybe? People having really odd conversations with semi-capable home automation running via their PCs.
Yeah I’ve almost never got in an Uber that was notably unclean or damaged in some way in London. Most of the times I’ve got one in SF, it’s been an unpleasant experience and so I now Waymo when I can there.
It is, but I don’t think this a bad offering - up until recently all iPhone cameras were 12mp, so you still get “good enough to print” quality. I guess it’s a bit of marketing speak but I don’t mind - to me it seems they made good choices on the lenses this year. 5x always seemed a bit too much without something in between but hopefully 4x is a decent compromise while still enabling “8x” (which I suspect is important for marketing and honestly will be quite fun)
Buzz was so great. I came across it when I learned James Holden produced his early stuff on it and was hooked. It had a good community with hundreds of synths and effects you could download from Buzzmachines. It was such a shame they lost the source code, I’d have loved to see how it developed.
This made me google Buzz and it turns out it was recoded by the author and there was an update in 2022. Not sure if it’s still developed. I’m not on a Windows machine so won’t be able to try it easily unfortunately, wish it was open source so it could be ported to Mac
If you work for a large org you’ll have some kind of enterprise agreement in place guaranteeing this. I can’t imagine they’d risk violating it regardless, the outcry could ruin them
If you work for a large org which has an official AI policy and agreements, sure. In those cases there is no problem; you're sending your employer's code to these companies in compliance with your employer's policy.
function Hello() { return <button>Hello</buttton> }
reply