Scaling AI in a Resource-Constrained World: What Business Leaders Need to Know

Reflections from the Frugal AI Hub panel discussion at Cambridge Judge Business School

When capital and computing power are limited, how do you build a thriving AI ecosystem? It's a question that matters deeply to the UK, and to every organisation trying to navigate AI transformation without Silicon Valley-sized budgets.

Last week, ChangeSchool Managing Director Viren Lall attended a panel discussion at the Frugal AI Hub at Cambridge Judge Business School that brought together voices from policy, academia, and business to explore this challenge. What made the evening valuable wasn't grand declarations or easy answers. Instead, it was the honest acknowledgement that the sector is navigating genuinely uncharted territory, and the willingness to grapple with questions that don't yet have clear solutions.

A Maturing Ecosystem Takes Shape

The breadth of initiatives underway across the UK became clear throughout the evening. Tim Flagg from UKAI described how the organisation emerged from conversations about what UK businesses genuinely need, evolving from the "Wild West" of a year ago towards a more structured industry with guardrails taking shape.

Jakob Mökander from the Tony Blair Institute emphasised building trust through transparency. Businesses increasingly want to demonstrate their commitment to sustainability and ethical practice, not as box-ticking exercises, but as genuine competitive advantages that build consumer confidence.

From a policy perspective, Gareth Lomax represented the government's Artificial Intelligence Incubator, highlighting efforts to make e-services more accessible whilst navigating the tension between moving quickly and getting things right. Prof. Kamiar Mohaddes brought the economic lens, exploring how AI-driven productivity gains might translate into broader economic growth, whilst acknowledging the uncertainties around energy consumption and sustainability.

What resonated was how these distinct perspectives consistently circled back to the same fundamental challenges: access to talent, the sustainability of data centres, public trust, and perhaps most pressingly, the future of work.

The Future of Work: No Easy Answers

One exchange during the Q&A captured the evening's honest tone. When asked about AI's impact on employment, the panel acknowledged a crucial point: nobody actually knows how this will unfold.

Historical patterns (where technological revolutions ultimately created more jobs than they destroyed) may not be reliable predictors for AI's trajectory. This intellectual honesty matters. Rather than offering reassuring platitudes, the discussion acknowledged that AI represents a qualitatively different shift from previous technological advances.

The speed of change, the breadth of impact, and the fundamental questions about what uniquely human capabilities need to be preserved: these remain genuinely open questions. The panel explored whether the UK would need a coherent political narrative encompassing pro-technology, pro-globalisation, and pro-migration positions to truly scale an AI ecosystem. There's no consensus yet, but the conversation itself highlighted how intertwined these challenges are.

Sustainability: Beyond the Headlines

The discussion around data centres and environmental impact revealed important nuances often missing from public debate. Yes, energy consumption is increasing. Yes, water usage in some jurisdictions is significant. But the picture is more complex than many headlines suggest.

Several panellists noted that the UK's climate (cooler and with higher ambient moisture) means data centres here use substantially less water than those in arid climates. One data centre operator reportedly uses approximately a swimming pool's worth of water annually in closed-loop systems, a fraction of what's required in desert environments.

The shift towards renewable energy in the sector is accelerating, though not always publicised due to concerns about "greenwashing" accusations. What emerged was recognition that whilst the evidence base remains somewhat speculative (with projections often outpacing robust empirical data), local communities' concerns are very real. Technical sustainability is only part of the challenge; social licence to operate matters enormously.

Trust, Talent, and Sovereignty

The concept of "AI sovereignty" wove through multiple conversations, not as narrow nationalism, but as a question of capability, resilience, and self-determination. Can smaller economies develop AI capabilities without matching the scale of US investment, which reportedly drives 40% of GDP growth? The frugal innovation lens suggests they must be more strategic about where to invest, which partnerships to build, and which capabilities to develop domestically.

Trust emerged as both a challenge and an opportunity. Businesses increasingly recognise that demonstrating ethical practice, sustainability credentials, and genuine societal benefit aren't peripheral concerns. They're core to building the consumer trust that underpins adoption and growth.

Access to talent remains a persistent constraint. The panel touched on workforce development, the need for education systems that prepare people for an AI-enabled economy, and the reality that talent mobility will be essential. Yet immigration remains politically contentious, creating a tension between economic logic and democratic pressures that won't resolve easily.

What Leaders Should Consider

For those navigating AI transformation in their organisations, several themes deserve attention:

For business leaders: The competitive advantage increasingly lies not just in adopting AI, but in demonstrating ethical practice, sustainability, and genuine value creation. Building trust through transparency will separate substantive organisations from those merely riding the hype.

For policymakers and educators: Workforce development can't wait for clarity about the future. Building adaptability, critical thinking, and uniquely human capabilities (judgement, empathy, contextual wisdom) becomes more important when specific job requirements can't be predicted.

For board members and advisers: The questions to ask aren't just "Are we using AI?" but "How are we preserving human judgement in decision-making? How are we building capability rather than dependency? What are we doing to ensure our people develop alongside our systems?"

For all leaders: The sustainability debate requires nuance. Blanket statements about AI's environmental impact miss important regional and technological variations. Engagement with evidence, not just headlines, will enable better decision-making.

The Challenge for Executive Education

The evening's discussion has particular resonance for those working at the intersection of education, innovation, and leadership development. At ChangeSchool, the team has been exploring these questions through its AI brain trust, which examines use cases and develops workshops for innovators globally. The approach prioritises ethical, non-biased applications that enable organisations to scale whilst maintaining human judgement at the centre of decision-making.

Recent academic work, including research on Structured Social Learning from Julian Birkinshaw, highlights how AI in education risks short-circuiting the cognitive struggle necessary for deep understanding, displacing skills through substitution, and decontextualising judgement by removing it from social and historical contexts.

This reinforces ChangeSchool's commitment to discovery-based learning and the case method in its executive MBA and executive education programmes: approaches that preserve the essential human elements of learning even as AI's capabilities are harnessed.

The challenge isn't whether to use AI in learning and development, but how to design experiences that build capability rather than dependency. How can educators help leaders develop the judgement to know when to rely on AI-generated insights and when to override them? These are questions the ChangeSchool team is actively exploring as it scales programmes internationally.

Moving Forward Together

Events like this matter because they create space for honest dialogue about genuinely difficult questions. Rather than providing easy answers, they help participants ask better questions.

Several remain particularly pressing: How can organisations prepare workforces for a future that can't yet be clearly envisioned? What governance frameworks can keep pace with technological change? How can leaders balance the efficiency imperative of AI with the need to preserve essential human capabilities? Can frugal approaches to AI genuinely enable smaller economies to compete, or will capital and compute constraints prove insurmountable?

These are the questions that will shape how organisations (and economies) navigate the AI transformation ahead. At ChangeSchool, the commitment is to helping leaders develop the capabilities to answer them in their own contexts, building judgement and wisdom that no algorithm can replace.

Viren Lall is Managing Director of ChangeSchool London. This article reflects on insights from a panel discussion at the Frugal AI Hub, Cambridge Judge Business School, co-hosted by Jaideep Prabhu, Elizabeth Osta, Nicola Buckley, and Christian Neubacher. 

Be part of our growing community of innovators and leaders

Follow us on LinkedIn