A list of AI coding tools
github.com
github.com
Seems it should be easy enough to have a basic model doing the same, by using Whisper and any sensible LLM
Talk -> streaming STT/Whisper -> some basic processing to put the text in a "markdown-like" format which allows to do structural edits (eg "change the order of the bullet points") -> send to a LLM which process the transcript and apply changes (add text, modify structure, etc)
However the biggest concern is that it would cost a lot of money to build this. If you have any thoughts/advice I'd be happy to chat.
> Cursor (editor, not yet released)
Well, it is up and running, https://cursor.sh/. I recommend it a lot - finally, not just a chat, but a tool in which a considerable fraction of programming happens in English.
Two weeks ago, I converted my blog from Gridsome to Nuxt 3 - the latter is a framework I haven't used yet. It was able to generate whole files and debug them line by line. Or write scripts that change folder structure - well-written, well-explained, and just working.
Still, I needed some knowledge of TypeScript and Vue, but it saved tons of my time - for writing, googling, and debugging. It felt more like working with a diligent, smart, and fast junior.
The goal is advertise their AI tooling under the guise of being a helpful list, so they're incentivized to make other tools look immature rather than be up to date.
Unless that's what they mean by dividing the list into "code completion tools" and "code completion LLMs"? In this case I'd adjust the naming to be more clear (I'm sure many of the tools in the first category are LLM based too)
This also benefits these tool developers: What responsibility do you have when receiving sensitive information from .env or other secret files? What if a secret escapes because you've trained an AI model on sensitive data and someone found a consistent way to access it?