Update: The service has been improving a lot. Infra is more robust, higher query limits, more sources, a further optimized Claude Prompt with much improved guidance for meaningful embedding vector manipulation. Someone could write some interesting quicktakes from the research you’re easily able to do with this.
Update: The service has been improving a lot. Infra is more robust, higher query limits, more sources, a further optimized Claude Prompt with much improved guidance for meaningful embedding vector manipulation. Someone could write some interesting quicktakes from the research you’re easily able to do with this.