Using AI to learn objective things is acceptable. However, as long as it's combined with your own experience, because AI can't possibly understand your entire world, any subjective answers will be disgusting, disastrous, obsequious, and boring.
I bought a Claude membership a few days ago. I asked him to fix a React issue—a very simple UI modification with almost no logic. He still failed to understand it. And after three attempts, the 5-hour limit was reached. This was a disaster. I had to immediately buy a CodeX membership and also tried Image2. I won't give Claude another chance.
Well written. I also wrote in my personal notes yesterday that if you don't organically and continuously develop code, it's hard to say you truly "own" it.
Like self-driving cars, at least you remember the scenery along the way, but now it just teleports you to another place and then shows you the recording.
This kind of review is ineffective. Such ghosted code might be acceptable for small tools, but for databases or similar systems, it's really worrying.
I've now basically stopped granting the agent any write permissions and returned to how I wrote codes 2 years ago with manual QA. The thing is, it's actually more efficient in turns of tokens and the results.
I really like Obsidian, but its features are still too much for me. Are there any aesthetically pleasing, faster alternatives that simply render Markdown and LaTeX? It would be even better if it also supported mixed inputs like Obsidian.
Note: Nick Collins (the author of this book) and Alex McLean created Algorave. The time I spent learning from the Algorave community was crucial to my later work on Glicol (https://glicol.org/).
Btw, I have a feeling that if you want to learn about computer music, you can send the PDF to LLM and ask what the chapter is about and how to represent it using csound or supercollider.
My experience is that with computer music, you have to keep experimenting and listening in order to truly understand and innovate.
Interesting. But if you claim "prompt in Godot game out", how do you deal with assets? I think assets pipeline is one of the most challenging parts in game dev. Is there anything similar but for Bevy?
"The stronger boundary protects the machine while the agent is coding, testing and improvising. It does not protect the rest of the world from the permissions you have already granted. A better-isolated runtime will not stop the bot from spraying outbound messages, sending a stupid email, or otherwise turning your authority into a minor public nuisance."
The core issue, to me, is that permissions are inherently binary — can it send an email or not — while LLMs are inherently probabilistic. Those two things are fundamentally in tension.