> there is pressure to ship crappy code

this is true with or without AI

Yes but now there is so much more volume of code created in the same amount of time.

I would hope that actually testing the changes is done regardless of whether or not AI is used

LLMs are far more flexible in what you can create, opening up many niche use cases for non programmers (or those with very limited programming experience).

For example, I use LLMs for one specific thing, making plugins for an app I use (which need to be written in javascript/typescript). No code tools wouldn't be of any use to me here.

No code tools put you in a box that limits what you can create, whereas LLMs allow you to code pretty much anything (though of course how far you can get does depend on having at least some technical ability/knowledge).

in antigravity gemini sometimes inserts its CoT directly into code comments lol

its so easy to accidentally hit enter though lol, I usually type larger prompts in my notes and copy paste then finished

Blanket statements; what are they good for? Nothing

it's better than at launch, but I still get random model response errors in anti-gravity. it has potential, but google really needs to work on the reliability.

It's also bizarre how they force everyone onto the "free" rate limits, even those paying for google ai subscriptions.

yeah I've been getting better results with codex (gpt5) vs claude

if only gemini cli could edit files without getting stuck in an infinite loop

A few weeks ago I asked gemini cli to do something pretty simple and it ran for like 12 minutes and then failed with an exception. Haven't tried it again since.

I asked it go modify a maven pom.xml to support grpc. forgot about it and blew through my usage.

While joplin's notes are stored in a database, they are still markdown files