I just had to go check out your site with that comment. Nicely done! Was waiting for the physics simulation to either fully settle or explode, but really neither of those happened.
I've done this. Ask Claude to permutate a prompt and run Claude Code or a subagent to observe effects until success. Claude will iterate on the prompt faster than I could and likely catch more edge cases, too. It's one of those tasks where the end goal is very clear and the agent just needs to iterate on permutations. Arguably this is a perfect match, too, because prompt engineering is really just text generation, so why not have a text generation machine do it.
Similar thoughts here. My gut tells me "if I wanted AI responses I would just ask AI", but in truth LLMs are not like Google Search. LLMs are more like a slot machine where you put your prompt in and out comes a combination of insights that are personalized to the inputs.
"I don't know" isn't in the training data. Nobody writes engineering books, science papers and blog posts that end with "well, I don't really know, the end".
There is a point where this kind of a pragmatic/contrarian take doesn't stick anymore. Is the net benefit of AI worth the cost of losing natural resources for billions? That's the question you're supposed to ponder, like, it's actually up in the air and not clear cut?
So it's better than 3.6 Flash, at half the price. I've been pretty excited about Gemini models recently, they just feel so fast after spending most of the day at work waiting for Opus 5.
I'm super curious how you make the development process interesting for him. I get that a 5yo would be good at generating ideas but how do you then go "okay, now I'll need a week in front of a screen to build all these ideas" and not have him lose interest. Do you accelerate using LLMs?
> web app I had launched using Claude was strikingly similar to his open source project, even reproducing a bug he had later fixed,
How? How do I prompt to get a literal copy of another app? When Claude/Gemini builds anything for me I have a thousand ideas for functionality improvements and UI/UX changes. How do you just stop at v1 of what Claude gives you?
This is written from a point of view of somebody who never went through the progression of CSS to CSS preprocessors to component-based frontend libraries. Tailwind is an answer to all the problems that become apparent in CSS in that progression.
I tried to embrace web components, but when I reached "declarative shadow DOM" I really just stopped seeing the point of all that complexity just to do what I can already do with a library. The target audience of that API seems to be library developers.
This post is full of things opposite of what I've found. An agent (claude code at least) will certainly refactor feature flags better than I ever could and is less likely to miss special cases. What the agent struggles with is business knowledge - context I hold in my head as a human communicating with other humans and working on a codebase over a long period of time.
Kind of excited about this. 3.5 Flash on Antigravity has surprised me recently on a hobby project. When given opportunity to plan, it can deliver on tasks that would take me a while on my own and generates responses at blazing speeds - compared to what I'm used to at work with Opus 4.8 (granted I don't use Opus 4.8 on my hobby projects so just anecdotal). While with Gemini CLI I would just watch it run in circles and run out of 5h allowance before anything useful is produced (or even approached).
Yes, it's the Youtube phenomenon - often when looking for domain-specific videos on Youtube I tend to find videos by professional Youtubers not videos by professional domain specialists (overlap is very rare but occasionally you can find a domain specialist Youtuber who applies their domain knowledge to making Youtube videos).
How? What's your secret? It's actively fighting me from discovering anything new. "Hey we'll show this artist you stopped listening to two years ago on your start screen forever. Wanna jump back in?" That's Spotify to me.
Spotify is fundamentally broken in a certain, unfixable way, IMO.
I use it and have a subscription, but I dread opening their app and looking at the starting screen that shows the same artists I listened to twenty years ago in pointless blurbs like "presave this (you can't listen to it)", "jump back in (you literally already listened to it)", "your favorite artists (not according to you but according to us)". There is no joy of discovery of new music that you haven't heard. There is no connection to other humans through music. Audioscrobbler/Last.fm is miles ahead of this. Youtube is miles ahead of this.
Here's how I discover music these days: I swipe Youtube shorts until its algorithm decides to show me an artist, then I look that artist up on Spotify. Thats how bad Spotify is - it's an audio server with search and a hundred layers of irrelevant features bolted on top.
Claude Code CLI has an agents view that lets you manage multiple sessions in parallel. I literally only ever start claude with `claude agents` these days, and use left arrow to go back to that view from any single session.
What do you disagree with? Your third sentence disagrees with your second sentence, and I can't tell which part of my comment your first sentence disagrees with.
This made me realize that forums existed only because there was somebody willing to pay for domain, hosting and maintenance with money and time. As a result most had a bus factor of 1. All the forums I know died with their maintainer moving on - and even forums "resurrected" by a community member had that exact problem. There is space for a fully distributed forum that can't die with its maintainer.
What a blast from the past. Seeing Knoppix on my room mate's PC in 2004 is what led to a 20+ year ongoing adventure with Linux, Debian, gaming on Linux, compiling games with a friggin compiler and automake, programming, it all started with that distro.
My point isn't about that, though. My point is if you start a blog post with "The C string library", my confidence in your ability to discuss the topic is shot before I even finish reading the sentence.
Starting with "The C string library" made me instantly tune out. C has a standard library, which has some string-related functions. There is no "C string library".