What is Behavioural Science AI: Thinking Unlocked

By Let’s Think

In Part 2 of our Behavioural Science AI explainer series, we explore how this novel technology unlocks the hidden knowledge that makes experts exceptional, on an unprecedented scale.

Image by Becky Stevens

Behavioural Science AI is a new application of artificial intelligence (AI) that combines behavioural science methodologies with generative AI (GenAI) and analytical technologies.

At first glance, this might not seem like an obvious pairing. Behavioural science is about understanding what people do and why. AI is about building computer systems that act “intelligently” by learning from existing information. The exciting thing is that when these two areas come together, the overlap gives rise to innovation and scale that neither field could achieve alone.

“We are at the intersection between generative AI and the science of understanding human expertise.”

Wendy Jephson, CEO and Co-founder, Let’s Think

To understand how this works, it's helpful to look at the underlying technology. Behavioural Science AI is underpinned by a foundational model which has been trained by Let’s Think’s team using a wealth of behavioural science frameworks and methodologies. This means it has all the creative and analytical power of domain-general GenAI, but with a crucial difference: it's designed from the ground up to work in harmony with the ways people actually think and behave. This is what allows it to elicit, decode, scaffold, and map the complex thought processes that underpin decision making.

Applying behavioural science to AI unlocks a whole new set of possibilities for supporting complex knowledge work that were never before possible. Think beyond the automation of routine tasks: what if AI didn’t just give you the answers? What if instead of replacing your thinking, it helped you understand it more clearly, and share it more easily?

Because the challenge is, a lot of the most valuable thinking that happens in our organisations isn’t written down anywhere.

Unlocking Tacit Knowledge

And this is exactly where every knowledge-intensive organisation is stuck: how do you leverage expertise when the most valuable insights are locked in experts' minds, and junior staff need access to that knowledge to develop their skills?

As we explored in Part 1, much of expert decision-making relies on tacit knowledge: the deep, intuitive understanding and pattern-recognition processes that are difficult to articulate but critical for high-stakes decisions. Yet, to understand how complex thinking can be supported and elevated by AI, we need to know what exactly is going on in professionals’ heads: no easy task.

It used to be that the best way to do this was to rely on skilled behavioural scientists who had mastered the art of drawing out expert knowledge. In fact, by combining different evidence-based behavioural science approaches, behavioural scientists have been able to elicit and analyse knowledge in various professional contexts.

 "How experts make decisions is usually not obvious to the expert themselves… techniques developed in the Behavioural Sciences can help extract this ’tacit’ knowledge from experts and discover exactly what it is they do and know that makes them so good at what they do."

Anna Leslie, Chief Scientific Officer and Co-founder, Let’s Think

However, this approach required time-consuming one-to-one interviews followed by hours of manual processing and analysis, making it resource-intensive, and difficult to scale. What’s more, there are simply not enough behavioural scientists to meet the growing demand for this expertise.

Behavioural Science at Scale

Thanks to cutting-edge technological advances, Behavioural Science AI can now surface this expertise at scale. A model trained to function like an expert behavioural scientist can adaptively navigate unique contexts and communicate with users in natural language, turning the deeply human, expert-driven process of knowledge elicitation and sharing into an intuitive and effective digital experience.

The technology works by applying the same evidence-based methodologies that trained behavioural scientists use, but at unprecedented speed and scale. Instead of lengthy interview processes, the AI can engage with experts and juniors through natural dialogue to systematically pull out, organise, and share that hard-to-explain knowledge in real-time.

This means organisations can tap into behavioural science expertise and opens the door to rethink how AI fits into professional work. We can move beyond simple task automation to understanding and enhancing human expertise and critical thinking.

Enhance Not Replace Thinking

An important thing about Behavioural Science AI is that it doesn't seek to replace people and automate away jobs, but instead to enhance the value that professionals offer, by understanding how they really think, working seamlessly with their natural decision-making to amplify their value.

This opens up some exciting possibilities for promoting reflective thinking, professional development, performance improvement and team dynamics. For example, we can integrate Behavioural Science AI into systems that help juniors learn from experts, preserve valuable knowledge before it walks out the door, and enhance critical thinking across teams.

The Future of Thinking™ Starts Here

Behavioural Science AI is changing how we can reflect, refine, and share knowledge and insights that would have previously remained hidden - transforming how we develop and apply expertise across industries.

The key here is that it accomplishes two things: it enhances how people think, whilst championing, not replacing, the professionals it is designed to support. In this new framework, AI becomes a thinking partner and knowledge amplifier.

With behavioural science leading the way, we’re not just making work more efficient, we're enhancing human cognition, fundamentally reshaping The Future of Thinking™.

Stay in the loop with Behavioural Science AI breakthroughs and insights by subscribing to our newsletter.

How Behavioural Science Reveals How We Really Think, Decide and Act

By Let’s Think

Behavioural Science AI combines behavioural science methods with GenAI and advanced analytics to surface, evolve, and amplify expert knowledge. It’s a prized asset rarely captured by traditional systems. In Part 1 of our explainer series, we unpack what behavioural science really is—and what it’s not.

Image by Becky Stevens

As AI rapidly transforms how we work and live, there's growing recognition of something essential: that technology is only as useful as our understanding of the humans using it. Behavioural science offers a powerful lens for this understanding, but it’s often misunderstood.

For many, it’s associated with nudging—those subtle design tweaks that steer people toward safer, healthier (or more profitable) decisions. Think opt-out organ donation, default pension schemes, or perfectly-timed app notifications. 

The popularity of “nudges” grew from the work of Daniel Kahneman and Amos Tversky, who famously revealed how we’re prone to cognitive biases—mental shortcuts that can sometimes lead us astray. Their research helped launch behavioural economics, and it has become shorthand for a diverse field deeply rooted in scientific rigour.

But behavioural science is much broader than that. 

So, what is behavioural science?

Behavioural science is the study of human behaviour: how people think, decide, communicate, and act, especially in social and systemic contexts. It’s a multidisciplinary field dedicated to  understanding not just what we do, but what, where, when, why and how we do things. It blends insights from both natural and social sciences, such as:

• Psychology: how we perceive, learn, remember, think and feel

• Neuroscience: what’s happening in the brain

• Sociology, anthropology & political science: how we interact within groups, institutions and systems

Behavioural scientists use diverse methods: from self-report questionnaires and experiments, to brain imaging and ethnographic observation, not just to describe behaviour, but to understand why it happens, and how it can be supported and improved. 

A common misconception, rooted in the “nudge theory”,  is that behavioural science focuses on how humans are flawed, irrational, or easy to trick.

In reality, many of these so-called flaws are adaptations that work incredibly well in real life, especially when decisions are fast, high-stakes, and complex. 

How experts actually think

Other fields of cognitive science  look not at average consumers in controlled experiments, but at real experts making tough decisions in real environments—like emergency responders, pilots and military leaders. 

Initially developed by cognitive psychologist Gary Klein, this research shows that experts often don’t compare options rationally at all. Instead, they:

• Rapidly assess the situation by matching it to past experience

• Use intuition to recognise patterns and simulate likely outcomes

• Make fast, high-quality decisions under pressure

This explains why experience matters—not because it makes people smarter, but because it shapes their internal understanding of how things work.

Enter tacit knowledge: The hidden engine of expertise

So how do experts actually know what to do, if they’re not consciously reasoning through every step? The answer lies in different types of knowledge:

• Explicit knowledge: formal, written information (books, policies, training)

• Implicit knowledge: what you know from experience, but can explain if asked

• Tacit knowledge: deep, intuitive understanding that’s hard to articulate. It just feels right

Tacit knowledge is elusive—difficult to elicit and measure, and often invisible in traditional systems. Yet most people instinctively know it exists and that it’s critical to expert decision making. For example, a trauma surgeon may instantly recognise signs of internal bleeding and act without waiting for a scan, or a fire commander may detect subtle cues of a structural collapse and change strategy on the spot. These are not guesses—they’re rapid, experience-based decisions powered by pattern recognition and mental simulation. This is tacit knowledge in action.

“I help people and organisations navigate human complexity, not with guesswork, but with insight. Technology can be a delivery mechanism for that process, and behavioural science principles can be used to design the technology” – Anna Leslie, Chief Scientific Officer and Co-founder, Let’s Think

Why this matters in knowledge work

While behavioural science insights can be used to influence others, its greater power lies in helping organisations understand and support how people actually work, especially in complex fields with frequent high stakes decisions such as law, financial services, law enforcement, and medicine. 

We can use Behavioural Science to:

• Understand expert decision making

• Support judgement under uncertainty

• Design tools and systems that work with, not against, human cognition

• Make tacit knowledge shareable and visible

At Let’s Think, we’re applying behavioural science principles to build tools that help people think brilliantly—because when they do, they make better decisions, solve harder problems, and create real impact. That’s the kind of thinking the world needs more of.

To stay on top of how we apply Behavioural Science to AI and follow insights and trends, sign up for our newsletter.

If you are new to decision making research, a helpful jargon-busting overview can be found in this A–Z of decision-making from CREST, which outlines key terms and frameworks used across the field.

From Idea to Innovation: Building the Let’s Think Knowledge Exchange

Today marks a moment of reflection on our journey at Let’s Think—one that began with a deceptively simple question: How do we capture expert knowledge before it walks out the door of a company? 

As behavioural scientists, our team has long understood how difficult it is to surface what people know, especially when that knowledge is tacit, embedded in experience, and rarely documented. Our earliest insights came from recognising that organisations everywhere suffer from knowledge loss—whether someone is leaving a role or simply too busy to share what they know.

That idea resonated widely. At events like London Tech Week, it connected us to people like the Head of Innovation at a leading financial institution. She got it. Her team faced exactly this problem: talented secondees rotating in and out, but no scalable way to retain their learnings.

So we started building—with interviews, handwritten notes, and a prototype concept for the Knowledge Exchange.

That early work laid the foundations for something far more ambitious. Serendipity played its role again when I met Sarah Harris, Director of Innovation and Knowledge at Kingsley Napley, at a UKRI Innovation Lab. Her experience in legal investigations helped bridge our thinking to a high-stakes, knowledge-rich sector: Law. The Lab’s ESG and data access goals helped refine our proposition. We realised we weren’t just capturing knowledge. We were generating a novel expertise dataset that doesn’t currently exist in most organisations.

And then came the wildcard: LLMs. Fresh on the scene in late 2022, they offered a powerful new tool for micro-eliciting knowledge. Could we develop a product that captured meaningful insights in just five minutes? In an industry like law, where time is measured in six-minute increments, that challenge was not only exciting but necessary.

We've spent the last year testing, validating, and iterating the Knowledge Exchange. We’ve trialled the product internally. We’re running pilots with law firms like Kingsley Napley. And we’re seeing, with growing confidence, Behavioural Science AI is a technology that can transform knowledge management. 

What’s been most striking is how much the journey itself has shaped the product. Unexpected opportunities like innovation labs and pivots into new industries became forcing functions—shaping our assumptions, surfacing new needs, and accelerating our thinking.

The ideas were always there. But the path forward emerged only as we walked it, and heeded the signals for the best direction to go to solve a pressing customer problem. 

Wendy Jephson is the CEO and Co-founder of Let’s Think. She is a serial entrepreneur and dually qualified as a Behavioural Scientist and Lawyer.