Hacker News new | ask | show | jobs
by SkyPuncher 230 days ago
This is it? A novel model and an interface I don't care to use.

I don't understand the push for multi-agent right now. Getting a single agent to work reliably is a problem. Getting multiple running, with possibly overlapping diffs just seems to be impossible.

2 comments

They all seem to have hit the upper ceiling with state of current LLM's and chat interfaces so trying to entice you with minor updates.
I am still waiting to do voice controlled chats where I don't have to speak in a foreign language and keep correcting the input.

Now there is something of value to actually improve.

Let’s you burn cash faster.