Skip to content
Join 7,000+ leaders following Alastair's work on LinkedIn.

Framework Library

The named models, scorecards and rules behind how we help teams adopt AI - one shareable card each.

Browse by topic

Browse by layer

Frameworks and principles

Mindset Butler vs Advisor Mindset The Human Advantage Prompting The Core Four Prompting The Deep Research Brief Prompting The GOAL Framework Risk Where Does It Run? Risk Risk Zones Risk The Verification Dial Risk What Counts as Private AI? Risk The Vendor Evaluation Card Risk The Permission Ladder Risk The Four Seats Choosing Four Measurement Lanes Choosing RATES Choosing The Three Implementation Tiers Choosing The Four Pillars of Opportunity Choosing The Pilot Charter Choosing The Pilot Playbook Choosing Scale / Pivot / Kill Adoption The Three Ways of Working Adoption Make / Sponsor / Escalate Adoption The Three-Audience Model Adoption The Three Circles Adoption HumanSpark AI Maturity Model Automation The Seven Moves Mindset The Boring AI Manifesto Mindset The Capacity Paradox Mindset AI is a Skill, Not Software Mindset The Amplification Problem Mindset Cognitive Offloading Prompting Ask Me Questions Risk The Expert's Advantage Risk AI Drafts, You Sign Risk Defensive Programming Risk Shadow AI Risk Produce vs Act Choosing The Golden Test Set Choosing Rules Before Guesses Adoption Operator to Orchestrator Adoption The Leadership Literacy Floor Adoption The Two Adoption Gaps Adoption The Reduction Trap Adoption The Quit Question Adoption Tasks, Not Roles Adoption Recruit the Sceptic Adoption The Pipeline Problem Automation AI Literacy Is Not Automation Literacy Automation The Automation Owner

Frameworks25

Mindset2

// mindset

Butler vs Advisor

AI helps two ways - Butler executes a task, Advisor thinks with you.
→ open // mindset

The Human Advantage

Four things AI cannot commoditise - your moat as an expert.
→ open

Prompting3

// prompting

The Core Four

Four portable files - Who, What, How, Why - that give any AI your context.
→ open // prompting

The Deep Research Brief

An eight-part brief for deep-research prompts - GOAL with more rigour.
→ open // prompting

The GOAL Framework

A four-part structure for a prompt that gets a usable answer.
→ open

Risk & safety7

// risk & safety

Where Does It Run?

Every AI system has a venue, and the venue decides where your data goes, what it costs and what it can do.
→ open // risk & safety

Risk Zones

A traffic-light filter for where it is safe to start with AI.
→ open // risk & safety

The Verification Dial

Check AI's work by its stakes and track record, and earn your way down from 100%.
→ open // risk & safety

What Counts as Private AI?

Consumer, business and local AI handle your data differently.
→ open // risk & safety

The Vendor Evaluation Card

Four checks before approving an AI tool for business data.
→ open // risk & safety

The Permission Ladder

How much autonomy you give an AI permission to have.
→ open // risk & safety

The Four Seats

A four-role split that keeps AI advice separate from AI action.
→ open

Choosing & ROI7

// choosing & roi

Four Measurement Lanes

Measure usage, the work, the business result and the people, separately.
→ open // choosing & roi

RATES

A five-factor score that finds the highest-ROI tasks to automate first.
→ open // choosing & roi

The Three Implementation Tiers

Three ways to add AI to a process - very different cost, speed and risk.
→ open // choosing & roi

The Four Pillars of Opportunity

The four places AI creates value - to look past 'just save time'.
→ open // choosing & roi

The Pilot Charter

A one-page agreement that defines a pilot before any work begins.
→ open // choosing & roi

The Pilot Playbook

One workflow, taken from selection to a Scale, Pivot or Kill verdict on a date set in advance.
→ open // choosing & roi

Scale / Pivot / Kill

The three-way verdict that ends a pilot before it can drift.
→ open

Adoption & strategy5

// adoption & strategy

The Three Ways of Working

Use a product, direct a tool, run a service: who does the work, and is anyone watching?
→ open // adoption & strategy

Make / Sponsor / Escalate

Is it yours to do, yours to back, or someone else's to decide?
→ open // adoption & strategy

The Three-Audience Model

Translate your AI message three ways - for the Board, IT, and the team.
→ open // adoption & strategy

The Three Circles

Leadership, people, technology: most firms only fix technology.
→ open // adoption & strategy

HumanSpark AI Maturity Model

Where you sit on the path to AI adoption, in five levels.
→ open

Automation1

// automation

The Seven Moves

Seven moves that turn work you understand into a system you own: Find, Map, Split, Design, Build, Prove, Own.
→ open

Principles23

Mindset5

// mindset

The Boring AI Manifesto

The most valuable AI work is boring - not the headline projects.
→ open // mindset

The Capacity Paradox

The team most buried in the work has the least time to help fix it.
→ open // mindset

AI is a Skill, Not Software

How well the person uses the tool decides the results.
→ open // mindset

The Amplification Problem

AI takes bad inputs and scales them, with confidence.
→ open // established concept

Cognitive Offloading

Hand a skill to the machine often enough and you stop practising it - the individual cost of heavy AI use.
→ open

Prompting1

// common technique

Ask Me Questions

Tell the AI to interview you first, instead of writing the perfect brief.
→ open

Risk & safety5

// risk & safety

The Expert's Advantage

People who know the work well check AI's output faster and spot mistakes sooner.
→ open // risk & safety

AI Drafts, You Sign

AI produces the draft; a named human checks and owns what goes out.
→ open // standard practice

Defensive Programming

Put predictable rules around unpredictable AI: rules first to handle the bulk, rules after to check the output.
→ open // industry term

Shadow AI

Staff are already pasting company data into free AI tools.
→ open // risk & safety

Produce vs Act

Does an unattended AI act, or only produce and hand to a person?
→ open

Choosing & ROI2

// standard practice

The Golden Test Set

Test the AI on 10 to 20 examples with known right answers before launch.
→ open // choosing & roi

Rules Before Guesses

If the answer is already known as a stable rule or calculation, encode it rather than asking AI to rediscover it every time.
→ open

Adoption & strategy8

// adoption & strategy

Operator to Orchestrator

The shift from working one AI conversation to directing several.
→ open // adoption & strategy

The Leadership Literacy Floor

Before you back AI projects, watch someone use today's tools on real work.
→ open // adoption & strategy

The Two Adoption Gaps

Two gaps: people not using the tools you bought, and people using them while the work stays the same.
→ open // adoption & strategy

The Reduction Trap

Time AI saves gets filled with more work unless a leader decides what it's for.
→ open // adoption & strategy

The Quit Question

One question that finds real pilot candidates by asking about pain.
→ open // adoption & strategy

Tasks, Not Roles

AI takes the tasks people dread out of a job. The job stays.
→ open // adoption & strategy

Recruit the Sceptic

Put one sceptic on the pilot team to find problems before rollout.
→ open // adoption & strategy

The Pipeline Problem

Junior work is how people become senior. Automate all of it and nobody gets there.
→ open

Automation2

// automation

AI Literacy Is Not Automation Literacy

AI literacy is working well with an AI. Automation literacy is turning work you understand into a repeatable system.
→ open // automation

The Automation Owner

You can own an automation without writing its code. The owner knows what it is for, what it may change, where judgement comes in, how it is checked and who is called when it fails.
→ open

The Complete AI Toolkit