The Good, the Bad, and the Ugly of AI: Reflections from Our Thames Lido Round Table
As artificial intelligence accelerates at breakneck speed, the design, property, and construction industries find themselves standing at a crossroads. For some, AI presents boundless opportunity - a chance to work smarter, deliver better outcomes, and open new creative frontiers. For others, it raises very real concerns about job security, data privacy, and the integrity of professional judgement.
At our recent round table event at Thames Lido, we brought together a diverse group of professionals from across the region to unpack these tensions and explore one central question:
What does AI really mean for the future of our industry - and how do we prepare for it?
What followed was a lively, honest and, at times, provocative discussion that revealed not only the power of AI, but also the importance of keeping human experience, values and judgement at the heart of practice.
This article captures the key themes and insights that emerged.
AI as the Naïve but Powerful Assistant
Our session opened with a presentation by Bart Janaszek of Digital Trading, observing that AI is best approached not as a master, but as a naïve, exceptionally fast trainee - capable of producing vast amounts of content in seconds, but lacking instinct, lived experience, and the practical sense that underpins professional practice.
Participants agreed that while AI tools can radically speed up tasks such as meeting notes, design iterations, research compilations and client visuals, none of this removes the need for human oversight. AI “hallucinates”; it gets things wrong. And when those errors find their way into reports, design statements or planning documents, the liability sits firmly with the human professionals, not the machine.
This reinforced the idea that critical thinking and professional judgement will only become more valuable - not less - in an AIdriven world.
Opportunity: Efficiency, Creativity and New Ways of Working
One of the most compelling themes was how AI is already transforming daytoday practice. Several attendees shared examples, from rapidly generating design visuals to speeding up documentation processes, improving consistency and freeing up architects and consultants to focus on meaningful work - rather than admin or repetitive tasks.
AIgenerated imagery in particular sparked discussion. Tools such as Gemini and imagecreation models are increasingly capable of offering clients quick conceptual sketches or stylistic options, enabling richer conversations earlier in the design process. These tools don’t replace design expertise; instead, they act as accelerators - helping teams explore ideas at pace.
Others highlighted the potential for “AI agents”: personalised models trained on the voice, methods and outputs of individual professionals. These could turn meeting transcripts into structured reports, analyse planning constraints, or recommend content themes based on past projects - all while learning and improving over time.
Used well, AI could enhance the quality of junior learning by reducing time spent on lowvalue admin and increasing exposure to real design and project challenges.
Challenge: What Happens to the Next Generation?
Alongside optimism, the group openly explored concerns around employment and training pathways. If AI takes on the traditional “low-skilled work” once assigned to junior staff, will the next generation still gain the experience needed to become experts?
Several participants voiced concerns about the shrinking availability of earlycareer roles. If graduates no longer enter businesses to perform foundational tasks, do we risk undermining the progression pipeline - not just in architecture but across marketing, consultancy and planning?
This challenge is nuanced. While some argued that fewer entrylevel roles could emerge, others felt this simply shifts the nature of training: instead of learning through repetitive admin tasks, juniors could be exposed earlier to design thinking, client work and real projects. AI might compress the learning curve rather than eliminate it.
The consensus: we must remain intentional about how we train the next generation, ensuring AI enhances - not erodes - professional development.
Data, Confidentiality and the New Rules of Information
A recurring theme was data privacy and the fear of “training the AI for the competition”.
Participants expressed anxiety about uploading client information, proprietary methods or sensitive project details into public AI tools, especially when confidentiality is a contractual requirement.
Yet, paradoxically, there is also growing pressure to allow AI platforms to learn from our content - because these are fast becoming the new search engines. As one attendee put it: SEO is evolving from keywords to answers. If businesses want to be discoverable in the future, their expertise must be reflected in AIpowered search results.
This creates a difficult balance between exposure and protection. Some approaches discussed included:
separating internal and external AI use
using secure enterprise tools rather than public platforms
controlling which documents feed into model training
developing internal AI agents trained solely on firmapproved datasets
This is not just a technical challenge - it’s a strategic one. And every business will need its own clear policy.
The Bigger Picture: Energy, Infrastructure and Societal Change
Towards the end of the conversation, the group zoomed out to consider the broader implications:
AI is not simply a shift in tools - it’s a shift in societal structure.
Questions emerged about the huge energy demands of data centres, the race between countries to dominate AI development, and the strain on infrastructure. Participants also reflected on how AI will reshape education, entrepreneurship and even the perceived value of traditional career paths.
Some attendees saw the potential for AI to drive a new wave of small, agile, niche businesses - where individuals can scale services rapidly with the help of AI agents. Others highlighted risks of inequality, misinformation and overreliance on systems that are fundamentally fallible.
Across the board, however, one message was clear: the future will reward those who adapt - not avoid - AI.
Where Value Truly Lies: The Human Difference
In closing, the group recognised that while AI may transform processes, it cannot replace relationships, intuition, or the trust built between professionals and clients. In a world of automated outputs, human judgement becomes the differentiator: the ability to ask the right questions, interpret the nuances, and guide clients through complex decisions with clarity and empathy.
For architects, planners, consultants and designers, AI should not diminish our value. If anything, it spotlights it.
Our role is not only to design buildings, but to understand people. Not only to deliver information, but to interpret context. Not only to produce work, but to advise, reassure and lead.
And that remains - unmistakably - human.
Around the Table with iB Architects were:
Bart Janaszek Digital Trading
Kate FinchKate Finch Interiors
Duncan Flynn Cratus Group
Alex Morgan RIBA
Matthew Battle UK Property Forums
Mark Batchelor 4TY Planning
Doug Johnson Mesh Energy
Nick Cobbold Bell Cornwell
Nicky Brock Carter Jonas
Richard Stacey Evoke Transport Planning
David Gordon Luxury Sustainable Homes
Joanne Bridges Bridges Communications
Peter Sheppard ION Consulting
Nikita Lad iB Architects
Penny Dixon iB Architects
Ian Blake iB Architects