The AI Value System

A narrated explainer on the AI Value System, the framework behind The Value Office, a free decision tool for CXOs that runs on your own device. Captions run in the bar under the picture, and the transcript below follows along: select any line to jump to it.

What it covers

• Machine capability is not enterprise value. What AI can do and what an enterprise gains from it are different things, and better models alone do not close the gap.
• Three capabilities. Digital AI understands, Agentic AI acts, and Physical AI operates. Together they expand what machines can do.
• Business AI asks why, and who. It determines what is worth doing, rather than describing AI used in business or a vendor product.
• Two value systems. The three capabilities sit inside two value systems, Business AI and Human AI.
• AI value potential, T × B × H. Presented as a management heuristic, not a validated law: a conceptual index rather than a predicted ROI, and a hypothesis awaiting portfolio data. H is drawn as the weak link.
• Should this work exist at all? Instead of asking where AI can be used, the video places work on a range: eliminate, simplify, automate deterministically, augment, agentify, reimagine.

Explore further

• The Value Office: learn the AI Value System framework, test one initiative, and decide what is worth doing with AI. Free, and private by default.
• The Value Cube: the book on measuring what enterprise AI is actually worth.

Sources named in the video: Stanford HAI, The 2026 AI Index Report; McKinsey / QuantumBlack, The state of AI in 2026: On the road to ROI; NIST, AI RMF Playbook.