AI capability isn’t built in theory. It’s built in the working week.
8 Aug 2026 | Amanda Cunningham | Founder, Savvy Pixel®
Over the past week, I’ve spent a surprising amount of time thinking about one question:
What does it actually mean for a business to become capable with AI?
Not interested in AI. Not experimenting with AI. Not paying for another platform.
Actually capable.
And the answer keeps appearing in ordinary pieces of work.
This week, for example, I’ve been reviewing the AI Induction Hub in detail.
Not because I wanted to add more content for the sake of it. Quite the opposite.
I’ve been asking whether what is already there reflects the way AI is actually changing inside businesses.
One of the areas we added was AI context — specifically, what information should go where.
That sounds relatively straightforward until you consider what is happening in practice.
AI systems are increasingly able to remember information, personalise responses, connect to other systems and work across different parts of our digital lives.
That can make them significantly more useful.
It can also make it much easier to stop thinking carefully about what we are giving them, why we are giving it to them and where that information may subsequently be used.
That isn't really a prompting issue.
It is an organisational judgement issue.
Another conversation this week reinforced the same point from a completely different direction.
I met with a business owner who has built a successful operation around years of knowledge, relationships, trusted suppliers, working practices and personal judgement.
Like many established businesses, an enormous amount of its capability exists inside people.
We talked about systems, processes and opportunities to use AI.
But the more interesting question wasn't:
“Where could AI save time?”
It was:
“How much of what makes this business work is currently dependent on somebody knowing what to do?”
That distinction matters - If you simply automate activity without first understanding the knowledge behind it, you can make a business faster without necessarily making it stronger.
I’ve also experienced this at a much smaller level in my own business this week.
I bought a new AI-enabled transcription device.
The technology itself is impressive.
But owning it doesn't magically give me a useful meeting system.
I still need to decide:
What should be captured?
What shouldn't?
How will the information be organised?
What becomes an action?
What belongs in a client record?
What requires my judgement before it is relied upon?
And what should never simply disappear into an AI system because it happens to be convenient?
Again, the device isn't the capability.
The working practice around it is.
Even rebuilding parts of my own membership sales journey this week has reinforced this.
AI can help me examine wording, structure a proposition, question a customer journey and identify inconsistencies incredibly quickly.
But it cannot independently decide what Savvy Pixel® stands for.
It doesn't know which commercial compromises I am prepared to make.
It cannot determine which claims I am comfortable putting my name behind.
And it certainly shouldn't quietly become the keeper of those decisions without me deliberately structuring them.
That is why my view of AI adoption has changed considerably.
I don't think the most important question for businesses is becoming:
“How do we get everyone using AI?”
Increasingly, I think it is:
“How do we turn what this organisation knows into structured, accountable capability that people and AI can use properly?”
That includes the obvious things — policies, boundaries and governance.
But it also includes:
business context, customer knowledge, decision-making principles, processes, standards, responsibilities, approved ways of working, and knowing where human judgement must remain.
Because there is a significant difference between a company where lots of people use AI and an AI-capable organisation.
One has access to technology.
The other is building organisational capability around it.
And I suspect that distinction is going to become increasingly valuable.
It is also why I have been developing the AI Induction Hub around something broader than prompts or platform tutorials.
AI doesn't just need to be introduced to employees.
In many businesses, AI itself needs an induction into the organisation.
It needs context.
Boundaries.
Responsibilities.
Appropriate access.
And accountable review.
That is where much of my working week has taken me.
Not towards more AI for the sake of it.
Towards better structure around the AI we are already beginning to use.
And I think that is where the next stage of business AI is going too.
Originally published through Amanda Cunningham’s LinkedIn newsletter on 8 August 2026.
