Insights

What we're learning, as we build.

Field notes, client stories and considered perspectives on putting AI to work, from the people building CogniScale and helping teams use it.

In brief

CogniScale’s Insights page collects field notes, client stories and considered perspectives on putting AI to work. It covers the team’s experience of building AI race teams and helping organisations use AI with clear ownership and evidence.

Client stories
The systems stay. What changes is who carries the work.
Reality check18 Nov 2026 · 4 min read
Before you upgrade to the system with AI stamped on itSoftware suppliers are adding AI to every system and asking you to upgrade. Before you do, ask whether an AI colleague could already work with the software you have.
We’ll all become orchestrators of our own AI agents.
Future signals15 Oct 2026 · 3 min read
We won’t need new ‘agent orchestrator’ rolesLast week I spoke at the Chartered Institute of Marketing (CIM) and Henley Business School’s annual South East conference in Cookham, a beautiful village beside the Thames.
Change who does the steps, not every step.
Reality check2 Aug 2026 · 6 min read
AI value doesn't start with a redesign programmeThe assumption is that value from AI requires redesigning the organisation, re-engineering the workflows and retraining everyone first. It doesn’t. Start with the people and processes that work today, and test value before committing to wider change.
Six things to get right.
Reality check24 Mar 2026 · 6 min read
The end of business as usual: six things enterprises must get right with AILast week I spoke at Stockholm Fintech Week. The talk covered what we've learnt at CogniScale helping enterprise teams build AI agents. Here are the six things most organisations are getting wrong.
Find your sensible place with AI.
Future signals1 Feb 2026 · 6 min read
Find your sensible place with AIIt's not often I hear an interview that genuinely reframes a familiar subject. Marc Andreessen's recent conversation with Lenny Rachitsky did that for me.
Success creates its own problems.
Reality check7 Jan 2026 · 11 min read
The problem AI success createsOnce teams begin using AI successfully, another challenge appears: keeping the work visible, maintained and accountable, with clear governance, discoverability and evidence of value.
Sequence, not supremacy.
Industry insight6 Jun 2025 · 8 min read
AI-first: let's clear up the confusion once and for allDuolingo's AI-first backlash. Klarna's chatbot disaster. The public thinks AI-first means mass redundancies. Here's what it means - and why it matters for your team.
What Microsoft’s AI report tells leaders
Reality check10 Jan 2025 · 5 min read
What Microsoft’s AI report tells leadersMicrosoft's research on AI at work finds a gap between adoption and productivity gains. What separates the two, and what it means for your team.
The team principal character
Everyone gets an AI race team.Start with one team and one pressure point. We'll show you what it does on a normal Monday.