liveBook · Manning

Cost

As explained in Agent Design Patterns

Cost is the expenditure associated with operating an AI agent, summarized here primarily by resource usage or monetary charges.

Cost is the resource use or monetary expenditure associated with operating an agent. It is a design concern alongside predictive performance, interpretability, and alignment. Computational efficiency belongs on the cost side of that distinction: an agent can do its job well while using more resources than necessary. section 1.2.2

One way to reduce cost is to give the agent a process that quickly reaches appropriate resources for a task, rather than encouraging an extensive, expensive search for a plan or solution. Lower cost is a possible benefit of agent design, not a guaranteed one. section 1.1.2

Evaluate cost against what a design change achieves:

  • Check whether the change improves the agent’s ability to handle tasks; predictive performance and cost are separate concerns. section 1.2.2
  • Account for repeated work. Decompose-Process-Recompose can improve capability, but piece-wise processing usually adds a constant overhead per input fragment. section 8.2.1
  • Consider latency separately: processing fragments in parallel can improve latency even when the decomposition adds overhead. section 8.2.1

A large-input agent

An agent splits a large input into fragments, processes each fragment, then recomposes the results. This may let it handle the input more accurately, but every fragment introduces processing overhead. Parallel work may shorten the wait without removing that overhead. section 8.2.1

How Agent Design Patterns covers it

How deep it goes
Distinguishes computational efficiency from predictive performance and identifies fragment overhead, but gives no cost formula or accounting procedure.
Examples
Contrasts expensive reasoning searches with quick resource selection and notes per-fragment overhead when processing large inputs.
What you can do after
Evaluate/understand cost trade-offs when choosing agent processes or decomposing large inputs.
How it connects
Connects cost to predictive performance, interpretability, alignment, Decompose-Process-Recompose, and latency.

Also explained in

Compare how 17 books explain cost →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Agent Design Patterns. The section links point to the parts of the book it draws on.