Research, Architecture and Training
Not commentary — a position
Generic AI commentary describes what already happened. What follows is the firm's position on ten structural shifts that are underway now: what each one is, why it matters before it is obvious, and what it asks of an organisation that intends to be ahead of it rather than behind it.
“The next decade of AI strategy will be won by whoever sees the shift first.”
The ten shifts
Each carries a position and the advisory engagement that acts on it. Ten shifts, ten engagements, no duplication between them.
The Agentic Enterprise
“The most disruptive shift is not that AI can generate content — it is that AI can now take bounded actions across systems, make sequences of decisions and complete multi-step work without human presence at every step. Organisations that treat agents as faster automation miss the point: agents are the first form of digital labour that reasons about its task, not just executes a scripted procedure.”
The enterprises that will lead the next decade are those that design trustworthy agentic infrastructure — registries, permissions, approval gates and kill switches — before the agents arrive at scale. Most have not. The gap between organisations that have designed for agents and those that have merely deployed them will become visible in governance reviews, regulatory audits and operational incidents before it becomes visible in competitive outcomes.
What we publish, and what we do not
Published
The ten positions above, the Cognitive Coverage construct that runs through the Control line of DecisionOS, and the course material in the Academy. All of it is readable in full, without an email address.
Not yet published
Periodic research — an outlook, a technology radar, an index of coverage across organisations — is in development and is deliberately not named here. Comparative research needs data behind it before it can honestly claim to compare anything, and a title with nothing behind it is a catalogue entry, not research.