You're probably doing something the LLM has less fluency with.

Definitely would concede this, but I haven't had any issues so far. It's been able to make things look exactly as I've dictated them, even with relatively complex visualization of git branching. I get the sense that, with how capable Opus 5.5 is, "less capable" no longer means "not capable enough".

▲

0x457 5 hours ago | parent | prev [-]

LLMs can for sure write GTK4 (or Relm4), Swift UI, whatever windows uses, Kotlin for Android.

▲

catigula 5 hours ago | parent [-]

They can write brainfuck. I said less fluency.

	▲	0x457 4 hours ago \| parent [-]
		What I meant: they can write it pretty well. Probably because share of good quality code of native apps is much higher bad quality, due to all js slop we had even before LLM slop.