The machines do the work.
People answer for it.

Ed Fassio's body of work on human-AI collaboration: an operating system of governance, capability, and value. Its design principle is that the machines do the work; people answer for it.

An AI strategist, author, and founder of ByteBrain. He wrote The Agentic Frontier Series (The Constitution of the Frontier, Bots in the Boardroom, The Last Human in the Deal) and the peer-citable paper Decision-Scoped Delegation, and he advises enterprises on governing and enabling AI agents.

A constitution for human-AI collaboration: fifteen principles in five articles, plus two meta-rules — every principle must cost something, and no principle may rest on what machines can't do. Each principle names what it costs and who holds it. Read the principles.

Delegation is a contract for one consequential decision, not a score for how mature an agent is. Every grant states what may be decided, within what bounds, with what evidence — and a named human, the Sponsor of Record, can inspect, halt, or refuse. The framework · the technical paper is peer-citable at Zenodo DOI 10.5281/zenodo.23126767.

Five rungs of delegated authority, from Execute at the bottom to Set the Bounds at the top. The top rung — setting the bounds themselves — stays with humans. It gives an organization one shared language for what may be handed to machines, rung by rung. Climb the ladder.

Three questions that organize the whole body of work: who decides (the decision layer), who can work (the capability layer), and who benefits (the value layer). Every AI collaboration program needs an answer to each, in writing, with evidence. The three questions.

A method for measuring AI returns on two ledgers at once: Enterprise ROI, where released capacity only counts as value once it is converted, with evidence; and Human ROI, which tracks where returned hours actually went. Its core rule: released capacity is not realized value. The two ledgers.

Take the AI Governance Readiness Scorecard: fifteen questions across five dimensions, self-scored in about five minutes. It routes you to a band — from foundations missing to fully operating — and shows which dimension to work on first. Free, no signup. Take the scorecard.

The three Agentic Frontier Series books are on Amazon — Bots in the Boardroom is available now. The Decision-Scoped Delegation paper is peer-citable at Zenodo DOI 10.5281/zenodo.23126767, with a companion repository at DOI 10.5281/zenodo.23126782. This site also offers a free article, an audio overview, and a downloadable slide deck on the resources page.

Three offerings, one per layer of the Collaboration Trinity: the AI Governance and Enablement Foundation (the decision layer, a six-week engagement), the Delegation Ladder course (the capability layer, taught to whole workforces), and the AI Value Realization Review (the value layer, measuring returns on both ledgers). See the offerings or email info@bytebrain.org.