Why AI Needs a Human Layer

Much of the conversation around AI has focused on helping people understand the technology.

• How do we write better prompts?
• How do we increase adoption?
• Which model should we choose?
• How do we encourage employees to use AI responsibly?

They’re all sensible questions. As AI becomes part of everyday work, organisations naturally want to help people get the most from it.

Chatting with Liz Villani, founder of #BeYourselfAtWork – Sounds Like Me® From iAM, however, I realised there might be another question we’re overlooking. Rather than asking how people should adapt to AI, she asked what would happen if AI adapted to people. At first, it sounds like a subtle distinction, but the more I thought about it, the more significant it became.

Today’s AI models know an extraordinary amount about the world; they can draft reports, summarise meetings, analyse documents and answer complex questions in seconds. Yet for all that knowledge, they begin every interaction knowing almost nothing about the person using them.

They don’t know your experience, they don’t understand the judgement you’ve built over years of work and they can’t see the values that shape your decisions or the context that influences why you choose one course of action over another. Unless you deliberately provide that information, AI produces the most statistically likely response rather than the one that genuinely reflects you.

Perhaps that’s the gap the Human Layer is trying to address.

What is the Human Layer?

Liz describes the Human Layer as the missing connection between a person and the AI they’re using. Rather than treating AI as a tool that produces the same type of answer for everyone, the Human Layer is designed to help AI reflect the individual behind the prompt.

That means more than writing in someone’s preferred tone of voice. It means giving AI an understanding of the judgement, values, experience and context that shape how someone works. Instead of producing the most statistically likely response, the aim is to produce one that is more representative of the person asking the question. In other words, rather than asking people to adapt to AI, the Human Layer asks whether AI can adapt to people.

A different way of thinking about AI

It’s tempting to think of the Human Layer as simply another AI product or feature. I think the more interesting way to view it is as a shift in perspective.

For much of the past two years, organisations have understandably focused on the technology itself. Which platform should we use? How quickly can we roll it out? How do we encourage adoption? Those questions matter, but they only tell us whether AI is being used. They tell us very little about whether it is helping people apply their own judgement more effectively.

That matters because access to AI is becoming increasingly equal. Most organisations can now use the same models, ask similar questions and produce similar outputs. Competitive advantage is unlikely to come from the technology alone. It will come from the people using it and the different experiences, perspectives and decisions they bring to the table.

Perhaps that’s why the idea of a Human Layer feels so timely. It shifts the conversation away from making AI more capable and back towards making people more effective. Rather than treating human judgement as something AI eventually replaces, it treats it as the very thing AI should be built around.