Podcast Spotlight: Talking AI Education and Timberlearn with Web Usability

Dive into the long-form insights from our podcast on our new product Timberlearn below, or listen to the full conversation in audio format.

Tell me a bit about how you came to be together today.

We met at Orange, working together in the online team led by Keith Reville. Keith’s now a Timberyard NED.

Being there at the dawn of e-commerce was a huge ‘right place, right time’ break for us. We delivered some of today’s foundational tools like click to chat and click to collect.

A while after leaving Orange we reconnected, worked on a few projects together, and the Timberyard idea started to germinate.

Tell us about Timberyard

We’re a digital agency focused on enterprise website builds. Some have an ecommerce element, others don’t. But all are built on best of breed composable commerce technology. Our clients are in diverse sectors, including education. One is a leading exam board who we migrated onto a composable tech stack and ingested a number of complex other extranets to become part of the same estate.

We now manage and evolve their website, content and products across all subjects to support students, teachers, exam assessors etc.

We’ve had a front row seat for the birth and evolution of ecommerce. Timberyard distills everything we’ve seen and done into a lean, commercially-focused model, seen through a sharp ‘client lens’.

We’ve done the yards, and experienced and solved the same problems clients bring to our door. Or as we like to say, Real People, Real Experience. Real Results.

What are you seeing in the education sector when it comes to AI use both from students and institutions?

Institutions have shown healthy caution towards AI. But they’re beginning to accept it’s here to stay, including it in their strategic planning at a senior level.

This shift has two elements:

The first is operational change. Institutions want the day-to-day efficiency gains of AI - streamlining admin, better resource management, and reducing manual workloads.

Then there’s the Educational Product - what’s being taught, and how. AI is reshaping both.

It’s influencing teaching and learning methods. Students use AI tools to support and personalise their learning, creating new classroom dynamics. Then there’s the crucial issue of course content - the institution’s priceless IP. AI can twist and turn it until it’s unrecognisable. As a result its integrity and meaning can be compromised.

The teaching and the technology need to co-exist. But how? This issue of balance - of quality, ownership, and trust - is what’s keeping education providers up at night.

Imagine students in different regions, and even countries, uploading standardised course content into Gemini, Claude, GPT etc. Each tool uses different rules and applies different analysis and context to educational material that is precision-crafted, and has a high degree of specificity. The AI’s interpretation could incorporate a breadth of influences from a range of examining boards, with students taking an AQA exam exposed to elements from Pearson or CCEA course content.

For us it’s a question of intelligent adaptation, which is the enduring hallmark of progress. Or finding the rainbow without putting up with the rain!

What risks are institutions most worried about right now?

They want to make sure students are supported, and are using AI in the right way - to enhance their studying, not just to give them the answers.

They’re learning how to adapt assessments to emphasise critical thinking and real world application, while also recognising the importance of AI literacy as a core skill. Operationally they’re keen to leverage AI for tasks like speeding up marking, and spotting cheating.

It’s a paradox - institutions want to restrict and control AI use while also using AI tools to detect its influence and make themselves more efficient overall.

How should education providers be thinking about protecting their course content in an AI-enabled world?

The institution’s IP is its gold, whether that’s course content, curriculum structure or teaching approach. The chat platforms are businesses - training is integral to their value. So they’re willing to enter into value exchanges with students, such as free plans which, of course, drive document uploading. But this puts that IP out in the open where it’s no longer protected.

AI has the potential to support different learning styles. Where do you see real opportunities for improving accessibility or personalisation in learning? – Jason

We had this exact conversation last week in relation to SEN (Special Educational Needs) and AI’s potential to be trained to adapt course/curriculum content to better accommodate their requirements.

UK class sizes are 26-27 for primary and 22-23 for secondary. On average, one in six pupils have SEN, but the education system operates on a structured curriculum basis.

There’s a real and transformational opportunity for AI to make the curriculum, and learning it, more accessible to students with SEN.

For education leaders who feel curious but cautious, what questions should they be asking before adopting any AI solution?

How’s content, data, and information being used once added to the model?

Do you have the right controls over it?

Who’s the responsible human?

What can they monitor to ensure the AI tool is operating as it should, today and in the future?

Different levels of accountability and control may be needed, depending on what the AI is for. But there should always be appropriate human oversight across any AI solution to manage risk, understand how it’s working, and be able to continue training and adapting it over time.

You’ve been working on a platform called Timberlearn. How did that idea come about, and what specific problem were you trying to solve?

A team member whose daughter is studying a T-Level veterinary course was struggling to access and consume the course content. Google searches and questioning of generic AI was getting her nowhere - responses were generic and not aligned to the course in question.

From our internal conversations about useful AI implementations focused on a need or actual use case, Timberlearn was born.

We wanted to support students with accurate, context-specific responses to questions or tasks using curated content relevant to their course and curriculum. In this case one created and managed by the education provider, rather than through questions entered into AI tools.

We’ve begun training it in other key subjects and, with a focus on SEN, are developing it with the aim of delivering meaningful outcomes for students and the education sector as a whole.

What has the Timberlearn experience changed about how you think about AI and its role in education more broadly?

It’s helping us evolve our use of AI as a company. It’s been a thrill seeing the team pick up this challenge and run with it. It reflects our culture where people are deeply engaged with technology, passionate about what they do, and committed to pushing boundaries to see how far new innovations can go.

We believe Timberlearn has near-limitless value to educational institutions. It bridges and reconciles the competing needs of students’ appetite for dynamic learning, while ensuring full alignment with institutional intellectual property protection and standards.

The UK is renowned for leadership in education. We think Timberlearn can play a key role in shaping its future delivery, helping UK institutions both maintain and strengthen their position as global leaders.