When Nicolas Negroponte speculated about architects empowered by the new found agency of computers in his 1970 book The Architecture Machine, [1] he envisioned powerful tools that could depict, analyse, and help construct a design. And while his contemporaries were experimenting with the rudimentary precedents of what had recently been deemed ‘artificial intelligence’, he could not possibly have anticipated a world where computation, digital information, and connectivity were both extraordinarily powerful and ubiquitous.
The digital tools of yesterday – CAD, BIM, even the internet – will soon give way to autonomous AI agents that may drive our architectural processes from project conception to industrialised construction in ways far beyond that book’s wildest speculations.
2034: the automated future of AI
While the tools of today extend the architect’s human agency, those of 2034 are likely to be less instrumental and more autonomous, shifting the role of the architect even further from singular author to orchestrator of both human and algorithmic processes.
The anticipated shifts are numerous: from supervising drawing production to prompt engineering; from exploring three-dimensional implications from perspectival rendering to using immersive environments partially authored by AI agents; and from generating construction details from assembly diagrams to transferring assembly instructions to manufacturing platforms.
The two-decade transition to BIM occurred within the bounds of architectural practice, strengthened by advances in computer speed and graphic resolution. Ironically, the graphic processing units (GPUs) that made BIM possible are powering the move to AI, suggesting the pace of change is accelerating and making 2034 feel as if it is far in the future. But best to begin preparing now.
Four scans, four perspectives
The four perspectives offered by the horizon scans in the Technological Innovation theme anticipate this future while prescribing an interconnected series of recommendations that are a template for a proactive practice strategy.
Innovation Strategy counsels that the inflection point offered by the digital turn is not to be missed by an industry sorely in need of less friction and higher value. It suggests that the entire ecosystem of delivery – designers, builders, and asset owners – collaborate by capturing data and connecting it through new systems and infrastructure like digital twins.
The Digitalisation of Design suggests that the novelty of AI will give way to pipelines of information and autonomous agents that will smooth – and in some cases, replace – traditional modes of service and delivery.
Automation in Construction envisions the construction site of 2034, not bristling with robots and drones, but refactored by principles of information-enabled industrialised construction, where certain assemblies are created by designers and manufacturers offsite, and the resulting building is optimised by sensor controls.
Finally, Architecture in the Age of AI asks four important questions about professional knowledge, the meaning of professional judgement, business efficacy, and professional responsibility and risk as measures of the potential for AI to transform the building industry.
Through a purely technological lens, the contributors each frame the implications of a design and construction industry that is inevitably digitised and where liquified information flows freely between the various participants in the supply chain.
Collective efforts will be required – to define data standards and workflows, to set new automated process protocols, and to design new mechanisms of risk and reward – that the technology makes possible, but does not guarantee. If the original vision of BIM as frictionless digital collaboration was never realised, is it likely that AI-enabled, high. resolution, data driven, computationally intensive processes will reach that goal?
Retaining professional responsibility
Here lies the critical logic behind framing this section as "Technology Innovation" rather than "The Future of More Cool Technology in AEC". The title implies a responsibility of the architectural profession and their collaborators to not simply demand and occasionally implement every new tool offered by the emerging AI-powered technology. They must also rigorously investigate how the built environment is produced, determining which processes to embrace, reject, or modify.
That effort, at the heart of an industry-wide innovation strategy, must broadly consider an array of questions involving issues as disparate as design ethics, digital assets, education, risk and reward balance, new models of delivery, and social equity, just to begin the lengthy list.
The demand for industry innovation catalysed by – but not solely reliant upon – technology is nowhere more apparent that the tragedy at Grenfell Tower that occurred 17 years before 2034, these scans’ cut-off date. The fire resulted from an epistemic failure of modern design, construction, technical standards, product testing, regulatory oversight and asset operation, where critical decisions were disassociated from a coherent understanding of their implications and, even worse, an enforceable model of responsibility.
From the most charitable perspective, data about the circumstance might have been available, but not in a way that the danger could be detected. The systems to connect, evaluate, and predict behaviour of the deadly assembly were impossible, made inoperable by the means and methods of modern project delivery. Everyone on the Grenfell team had access to plenty of technology, but with no chance of using it to prevent disaster.
The year 2034 might see multi-modal process automation tools where this data will cohere in digital twins and predictive technologies will evaluate performance prior to assembly on automated platforms. But, as the AI scan contemplates, all will turn on an industry consensus about the meaning and implications of professional responsibility in a world where the line between computation and human work is increasingly blurred.
Other components of RIBA Horizons 2034 are architecture’s role and responsibilities for externalities that will also shape the profession – climate, economics, and population. Each sets out a fundamental challenge that architects can address with its skills. Our ability to translate those skills will turn, however, on whether we can convert the enormous opportunities of newer technologies into tangible, relevant, and critically needed results.
A fascination with the instrumental possibilities of technology is not enough to achieve this end. It demands an innovative and collaborative stance that can challenge and dramatically improve the way the industry works. Dame Judith Hackitt’s recommended Golden Thread, from the independent review of building regulations and fire safety, [2] is not just an information theory, but rather an innovation strategy to be applied across the entire delivery spectrum. Perhaps technology-driven innovation can help.
About the author
Phil Bernstein is an Associate Dean and Professor Adjunct at the Yale School of Architecture, where he has taught since 1988. He is a former vice president at Autodesk, where he was responsible for the company’s Building Information Modeling strategy that included the development of the Revit platform.
Prior to Autodesk Phil was an associate principal at Cesar Pelli & Associates where he managed many of the firm’s most complex commissions including projects for the Mayo Clinic, Washington National Airport, and Goldman Sachs. He writes and lectures extensively about issues of architectural practice, project delivery, and technology. He is the author of Machine Learning: Architecture in the age of Artificial Intelligence from RIBA Publishing.
References
[1] N. Negroponte (1970). The Architecture Machine: Toward a More Human Environment, p. 153. MIT Press
[2] Department for Housing, Communities and Local Government - J. Hackitt (2018). Building a Safer Future: Independent Review of Building Regulations and Fire Safety: Final Report, Cm 9551
