Your agent's clock is not the model's clock

Two parallel tools can do forty seconds of work in twenty seconds. Before buying faster inference, separate model-call latency, tool waits and the cost of a validated change.

Commenting on Bubbles has been removed. Why? If you want to talk about this post, write to the author, use the comments on their blog if it has them, or reply with a "RE:" post on your own blog. Suggestions? feedback@bubbles.town