Luke Macnab, architect at Feilden Clegg Bradley, says the potential of AI to transform design has to be balanced with ethical and environmental questions.
Heraclitus observed that change is the only constant in life. Throughout history, humans have created and refined an increasingly sophisticated array of tools to adapt and thrive. Today is no exception; it’s not an overstatement to say we are living through the Artificial Intelligence Revolution.
AI can be traced back to Alan Turing’s work in the 1950s, but it commanded widespread public attention when OpenAI launched ChatGPT in late 2022. Since then, its opportunities and risks have reverberated around the public sphere as global society grapples with the pace of change. As an architect, I have been watching these developments closely and find myself paradoxically excited and unsettled about their trajectory.
It is difficult to predict the full impact of AI on the profession. However, what matters most now is encouraging honest conversation, so that we move forward with clarity and intention. With that in mind, here are a series of ‘provocations.’
A useful starting point is a recent study called “Labour market impacts of AI: A new measure and early evidence” (Maxim Massenkoff & Peter McCrory) which attempted to quantify the tasks that large language models could theoretically accelerate across job sectors, and compare this against real-world usage data. Two observations stood out. First, architectural and engineering professions are not yet using AI to the extent of other studied sectors. Second, and more significantly, they carry some of the greatest ‘exposure potential’ of all, i.e. the potential of AI to replace elements of those roles.
There are caveats, and this is not definitive. But if even modest efficiency gains take hold, and practices begin operating more productively, it opens the door to reconsidering how architects spend their time and how they articulate their value, a theme Keir Regan-Alexander of Arka Works has been exploring thoughtfully in the context of AI-driven practice.
This is worth examining against the current scope of the architect’s role. One of architecture’s enduring attractions is its multifaceted nature; the articulation and visualisation of concepts, championing a client’s vision, the collaborative production of technical information, and the physical manifestation of projects in the built environment. That combination, across varying scales and typologies, makes full automation difficult, at least for now. But it also means there are numerous points at which AI tools can meaningfully enhance specific aspects: visualisation, report writing, compliance checking, environmental simulation, carbon analysis, and more.
What interests me most, though, is not efficiency alone. Collectively, these tools can foster an exciting culture of experimentation. The bar to entry for sophisticated AI applications has dropped substantially in terms of cost and ease of use, and architects who engage will find new ways of designing along with new ways of thinking about the profession itself.
Opportunities
As automation absorbs more production tasks, what remains is increasingly human-to-human connection, which is arguably what was always most important. The relational skillset of architects is well placed to lean into this, both at the individual client level and at the scale of community engagement. Meaningful co-creation with communities has traditionally sat in tension with production time, given stretched fee structures. If technology compresses production time, architects can talk more and draft less, liberating them to apply their skills to genuine social value and creative ambition. The work of Barefoot Architects offers a compelling model for community-led projects, and I hope that the liberation of time through AI can enable more like-minded schemes.
Architects have a legacy of innovation, whether typological, material or systems based. Now, AI also enables new business structures that simply could not have existed before. A promising example of this is the startup Livedin, where the team use architectural skills combined with AI tools to connect landowners, self-builders and designers, a proposition that uses large datasets and the entrepreneurial vision to harness them. This kind of architectural entrepreneurialism is likely to grow.
Looking further ahead, the UK construction industry’s persistent productivity gap is worth considering. Robotics is progressing more slowly than LLMs, but it is developing. As these systems mature, it becomes increasingly sensible to approach the design and production of the built environment more holistically, such as incorporating the opportunities of Design for Manufacture and Assembly. Companies like XKool and Dataform Lab are already exploring the built environment as a system into which AI plays a defined, human-led role, which is a model worth watching.
Responsibilities
There are therefore grounds for optimism, but alongside the opportunities there are responsibilities and challenges which I think architects, along with the global community, need to take seriously.
AI computing demands significant energy and water, a factor that cannot be ignored given the carbon intensity of the construction industry. For example, Meta’s Hyperion data centre project is projected to consume the equivalent of roughly half the electricity of New York City. Data centres are already competing with communities for resources, and this will intensify. But it need not be purely extractive. Research by Prof Petter Terenius and others explores how waste heat from data centres could be productively reused. Architects, with their position at the intersection of infrastructure and community, are well placed to reimagine data centres as community assets rather than isolated burdens.
At an individual level, MIT research has explored how over reliance on AI, delegating thinking wholesale, has measurable negative effects on cognition. There is a balance to be struck. The most productive relationship with AI is conversational, engaging with it critically, challenging its outputs, rather than simply accepting them. For a profession built on judgment, this distinction matters. Likewise, to satisfy their professional due diligence and quality control procedures, architects should be wary of delegating tasks to AI which they can’t do without it.
Artificial General Intelligence
We are at a pivotal moment that extends well beyond architecture. What distinguishes this technological revolution from previous ones is not only its speed, but that intelligence itself, not merely tools, is being developed. The stated aim of the major AI companies is Artificial General Intelligence. The investment reflects this: according to OECD analysis of venture capital data, AI firms accounted for 61% of global VC investment in 2025, or $258.7bn, representing more than double AI’s share since 2022. If AGI is achieved, the implications extend far beyond how we work. They touch how we form our identities. That demands intentionality, not just adaptation.
Whatever path this revolution takes, I believe architects have a unique set of skills for navigating it, including spatial thinking, lifelong learning, problem-solving, creative thinking, teamwork, resourcefulness, and a commitment to social value. They are exactly what this moment calls for. The task now is to engage deliberately, and to help shape a future that benefits not just the profession, but everyone.
Luke Macnab is an architect and researcher at Feilden Clegg Bradley