For much of the past decade, organisations focused on what AI could do: automate processes, identify patterns, improve efficiency, and generate new insights. Today, as AI becomes embedded within operational systems and decision-making processes, a different question is increasingly being asked.

Over the past several years, discussion around AI has often focused on capability: larger models, improved performance benchmarks, greater automation, and increasingly sophisticated outputs. While technical capability remains important, a significant shift is now taking place across both public and private sectors.

For many organisations, AI procurement used to be treated like any other technology purchase: compare features, review costs, negotiate support, and move forward.

For many SMEs, the idea of adopting AI can feel intimidating. Headlines often focus on large technology firms employing teams of data scientists, machine learning engineers, and AI researchers, creating the impression that meaningful AI adoption requires significant internal scale.

AI is no longer confined to technical teams. It shapes procurement, strategy, risk, and reputation. Organisations are making consequential decisions about AI-powered tools, vendors, and processes every day and those decisions reach the boardroom whether leaders are ready or not.

Digitisation is often framed as preservation. Scan it, store it, and the problem is solved.

But the reality is more complex.

Without careful design, digitisation can unintentionally strip context, flatten meaning, and obscure provenance, creating a form of digital erasure rather than preservation.

As cultural organisations digitise collections and adopt AI tools, data is often framed as an asset, something to be stored, analysed, and reused.

But cultural data is not just content.

It represents histories, identities, and communities. And with that comes responsibility.

XR technology has advanced rapidly. But technical capability alone does not guarantee meaningful user engagement.

Too many XR experiences fail not because they lack sophistication, but because they lack relevance.

Designing XR that people actually use requires a shift in focus: from what is possible to what is valuable.

For years, XR has been associated with headsets, immersive, high-end devices promising fully virtual experiences.

Yet despite significant investment, adoption has remained limited.

The idea of “creative automation” has become one of the most persistent narratives in AI. The promise is seductive: tools that can generate designs, write content, and produce media with minimal human input.

AI is frequently positioned as a shortcut for creativity. Faster outputs, instant variations, endless generation, the implication is that more equals better. But in practice, creative work rarely benefits from unlimited possibility.

AI systems are increasingly used to generate historical narratives, reconstruct artefacts, and simulate environments. But as these tools shape cultural storytelling, a fundamental question emerges:

Digital heritage projects increasingly focus on reconstruction, rebuilding lost artefacts, restoring damaged architecture, or recreating vanished environments in 3D.

For much of the past decade, AI progress has been associated with scale: larger datasets, bigger models, and vast computational resources. But for many UK SMEs, that paradigm was never realistic, nor desirable.

AI sustainability is increasingly discussed in dramatic terms. Headlines warn of soaring data centre emissions and the environmental cost of training large-scale models. While these concerns are valid, they often remain abstract, focused on global trends rather than practical decisions.

For much of the past decade, AI has been framed as an automation technology. The promise was efficiency: fewer manual processes, faster decisions, reduced human intervention.

For much of the AI boom, performance was measured in decimals.

Models competed on benchmark scores, fractions of a percentage point in accuracy, reductions in error rates, marginal improvements in speed. The narrative was clear: better AI meant more precise predictions.

When AI projects fail, the explanation is often framed in technical terms. The model wasn’t accurate enough. The data was incomplete. The system didn’t scale.

For much of the last decade, progress in artificial intelligence has been framed as a race toward scale. Bigger models, more parameters, and ever-growing datasets have been presented as the inevitable path to better performance and, eventually, general intelligence.

Across the UK, organisations are experimenting with AI at an unprecedented pace. Proof-of-concepts, funded pilots, and internal trials have become commonplace, particularly among SMEs, cultural organisations, and public bodies exploring automation, data analysis, or digital engagement.