Board Advisor — AI Execution & Advisory

Remote role with quarterly in-person meetings if convenient. Advisory commitment of approximately 4–6 hours per quarter.

About the role

Keep our execution methodology honest—identify what truly moves a program from pilot to practice, and where our approach breaks.

Why this seat exists

The embedded execution side of the practice relies on a methodology: find the invisible human judgment inside an AI-mediated process, classify it, name the seats, choose the loop posture deliberately, instrument the human contribution, and run a cadence. This methodology is a claim about how organizations change, and such claims are often partly wrong in ways their authors cannot see. This advisor is the person who tells us which parts.

Responsibilities

• Pressure-test the execution methodology against real-world successes and failures at scale
• Challenge human-in-the-loop design, identifying where our three-posture framework is too simplistic and what genuine oversight looks like
• Advise on governance and regulatory trajectory, including model risk, the EU AI Act, and sector supervision, and what enterprises will soon need to evidence
• Assess the economics of embedded engagements, including how buyers evaluate them and what makes them believe claims about return on investment
• Review engagement design on live work, particularly the critical first thirty days where programs are won or lost
• Identify when our thesis is overreaching, as strong arguments can harden into dogma

Qualifications

The role is suited for several profiles:

• An operator who has carried an enterprise AI program from pilot into production and can describe the organizational work, not just the technical aspects
• A COO or transformation leader who has implemented an execution layer and understands the political costs involved
• A model risk, audit, or regulatory executive who has examined AI governance and can specify what evidence satisfies supervisors
• A human factors or decision science practitioner who can challenge our assumptions about automation bias and vigilance
• A consulting partner who has run these programs and is candid about what didn’t work

Requirements

• Commitment of roughly four to six hours per quarter, plus time to review research before publication
• Willingness to engage in substantive disagreement—this role is ineffective if you agree with us
• Comfort with our position that we do not sell autonomy, tools, or retain control—if you believe this is commercially naive, disclose it early
• Disclosure of any conflicts with a current employer or client

Benefits

We do not publish the terms, but the participation is structural and permanent rather than a token retainer. The package is designed to make this role worth taking seriously and is presented to you in full, in writing, early in the process.