Dominic Chiappe · People, capability & transformation

Thinking about how organisations perform in an AI-enabled world

AyEye · State of play

A clearer view. A better next step.

Understand models, evidence and the systems around them. Choose a route from a first explanation to team practice or organisational change.

Reviewed 15 October 2026 · Editorial edition 1.0 · Manually maintained

Stable concepts · A knowledge map

Follow the ideas, not just the headlines.

Select a concept to read its evidence and limits. Lines suggest learning connections, not causality.

A view with its limits attached.

Established describes a bounded principle, not certainty about every system. Emerging practice needs contextual review. Uncertain outcomes need testing. This is an editorial selection, not exhaustive coverage.

Established

Training fit and performance on unseen cases are different questions.

A past study or toy simulation is not a benchmark of today’s systems.

Sources: Harvard Business School AI Institute · Reviewed 2026-10-15

Established

Checking and decision ownership belong in the workflow.

Human review is useful only when it has authority, time and relevant capability.

Sources: NIST · Reviewed 2026-10-15

Emerging

A staged introduction can test more than a model demonstration.

Transfer across teams and versions remains a question to investigate.

Sources: HM Treasury · Evaluation Task Force · Reviewed 2026-10-15

Uncertain

Net value in your setting requires a meaningful comparison.

Usage and output volume alone do not establish benefit.

Sources: HM Treasury · Evaluation Task Force · Reviewed 2026-10-15

Connect learning to organisational questions.

Explore the workforce maturity wheel → · Find related ideas in the shared topic graph →

Learning tags connect to the nine workforce strands where relevant. They are discovery links, not a maturity assessment.

Read the dated AyEye story on portable evaluation