💡 New AI Workflow tip: Use voice to move faster

Since I discovered Aqua Voice (YC 24), I speak ~80% of my prompts and pseudocode - and my iteration speed has tripled.

LLM (Cursor, Claude, etc.) open → Hold Fn → Speak → Release → Pastes. Done.

The breakthrough isn’t just better dictation software, it’s that LLMs effortlessly untangle broken grammar and spelling

This is nice, was looking for something similar.

I think if Cursor had audio output could replace the ideation process. Grok 3 is really cool to discuss ideas, but if you wanna do it as part of existing product, than it lacks the context. Otherwise, was the best so far.