⌘K

AI in Practice

AI in Practice

Practical thinking about what it takes to use AI inside a real business. Beyond access to the tools: strategy, people, processes, judgement, security, costs and the messy work of implementation.

All AI in Practice Digital Leadership Digital Decisions Search & Digital Front Door Fractional Leadership Wineries Beverage

How to let AI make stupid mistakes without allowing it to make expensive ones

How to let AI make stupid mistakes without allowing it to make expensive ones

AI needs somewhere safe to be wrong. Start with a pilot, set the boundaries and stay close enough to keep stupid mistakes from becoming expensive ones.

AI has already entered the building — AI is already entering the organisation through existing tools, remote work, external partners and deliberate pilots.

AI has already entered the building

AI is already inside the software, workflows and supplier relationships many businesses use every day. The question is no longer whether to adopt it, but whether what is already happening is being managed deliberately.

Leaders don’t need to learn ten AI tools by lunchtime

Digital Leadership

Leaders don’t need to learn ten AI tools by lunchtime

AI has changed the environment leaders work in, not the job of leadership. The real question is not which tools leaders know, but where AI can materially help the business.

AI can make customer service easier. Unless...

AI can make customer service easier. Unless...

AI can make customer service faster and easier. But good automation needs clear boundaries, escalation and handoff when the customer needs something more.

Why do I have to speak wine to buy it?

Why do I have to speak wine to buy it?

Wine buyers often know why they're buying without knowing which bottle fits. Better product information and new ways to discover wine create an opportunity to meet them where they are.

As a competent professional, AI can still catch you out

As a competent professional, AI can still catch you out

AI can produce polished, convincing work at extraordinary speed. But professional judgement, context and responsibility still belong to you.

Why AI implementation is lumpy

Why AI implementation is lumpy

Easy access to AI is not the same as implementation. Inside a business, people, policies, security, costs and workflows make progress lumpy.

AI Didn’t Publish That

AI Didn’t Publish That

AI can make a draft look finished before the thinking is finished. Someone still needs to check it, question it and decide whether it is ready to publish.

"We should be using AI" is not a strategy

"We should be using AI" is not a strategy

“We should be using AI” may start the conversation. A useful pilot starts with a business need, clear boundaries and a decision worth trusting.