The missing middle in business AI is becoming clearer

Our starting observation

Through The Remix, we keep meeting people who have finished an introductory AI course, tried the tools and then asked the same question: what do I do with this inside my business?

That is different from needing another beginner course. It is the gap between using AI occasionally and making it genuinely useful in the work.

What the research is showing

JPMorganChase Institute found that many small businesses using AI are still in testing and exploration rather than integrating it across operations.

A 2026 survey of Hong Kong SMEs found the same practical barriers: limited skills, unclear returns, a lack of suitable use cases and too little time. Only 14% of adopters were using AI for process automation or workflow improvement.

At the other end of the market, the direction is clearer. OpenAI reports that leading organisations are moving from AI assistance to execution. Salesforce and Anthropic's new Claudeforce partnership connects Claude to company data, workflows, business rules, actions and governance.

These sources describe different markets and business sizes. Together, they reveal the same gap from opposite sides.

The signal

The next divide may not be between businesses that use AI and businesses that do not.

It may be between those experimenting with disconnected tools and those building enough structure for AI to work with the business: clear information, repeatable workflows, permissions, boundaries and human judgement.

That is our interpretation of the evidence, not a claim made by any one source.

What we are hearing at The Remix

People do not need to be told to “learn AI” again. Many have started.

They are asking how to move beyond generic training without becoming technical specialists. They want to apply AI to a real business problem, know what information it needs, understand what must remain human and see whether the change is actually worthwhile.

An approved participant comment captured it simply: “A course for each level would be valuable.”

This is a real signal from our work, but it is not evidence about every business in the Northern Rivers.

How this connects to our work

The Remix is the learning bridge: practical ways for Appliers and Builders to move from introduction to useful application around real work.

The Signal Studio goes further into implementation: organising business knowledge, redesigning a workflow, setting boundaries and building the systems that let AI work safely inside an organisation.

We are developing both because the missing middle is where many smaller businesses appear to be getting stuck.

What this means for you

If you have tried AI but it still feels separate from the way your business runs, you may not need another tool.

Start with one piece of work that matters. Ask what information it relies on, where judgement sits and what a better result would look like. That is the beginning of integration.

What we are watching

We are now testing whether this pattern holds more widely across the Northern Rivers: where people get stuck after entry-level learning, which workflows are ready for change and what kind of support actually helps.