Will AI Replace Architects?
--
Spoiler: Not in the near future. But how can it assist us?
Will AI take over architect’s jobs? Thomas Lane in the May 2023 edition of the Building magazine reported that AI can automate 37% of architects’ and engineers’ work. This, however, will likely mean that we can automate more mundane tasks and focus more on strategic and creative tasks. Just as Revit, digital NBS specifications, 3D modelling and real-time rendering software did not overtake architects’ jobs but just changed the way we work, the same is likely to be true of AI tools. New tasks — to do with AI management, for example — will also emerge.
At the time of writing this article we are in the second quarter of 2023 and AI has only fairly recently emerged into public consciousness. It rapidly established itself as the trend of the year, after Open AI’s ChatGPT was made available to the public as a free trial version at the end of 2022. Despite its limitations and originally only having access to data up to the year of 2021, the internet and social media exploded with chatter about ChatGPT and AI in general and the race is on for who can find the best use-case for it.
Many Architects are among the professionals left wondering what it all means for them. Rumours of AI taking over white-collar jobs has caused some concern and reactions range from declining to engage with ChatGPT/ AI, being sceptical about the benefits and being worried about data privacy. Most of us even worry about the potential of Artificial Intelligence to take over the world. How ‘intelligence’ and ‘creativity’ should be defined are debated. Nevertheless, most Architects also approach AI with curiosity, trying to understand whether and how it could be used for our work.
Some of the larger architectural and engineering practices have been experimenting with AI tools for many years already, building their own use cases and customised databases for the AI to work from.
This article has been compiled to begin answering the question of how AI could be used by Architects.
The short answer to whether AI will replace architects is that it’s unlikely to do so in the near future. Reading through the rest of this article will give you an idea why.
The field is fast-evolving and new uses will emerge over months and years but a generic understanding of what AI can do and what its limitations are is starting to emerge from the mist.
TLDR: You can refer to the parts highlighted in bold.
Until AI wins an architectural design competition, we have nothing to worry about.
These are some of the generic categories of how AI could be used to work alongside Architects and how it could benefit our work:
1) Design options based on specific criteria: AI algorithms can analyse vast amounts of data. Depending on the tools available, they can also, in principle, generate design options. Architects could in the future input parameters such as budget, space requirements, or sustainability goals, and AI could quickly produce optimized design alternatives for human assessment.
2) Site analysis and massing studies: AI tools, such as Autodesk Forma, can pull site information from open access databases and help you quickly create massing studies taking into account environmental data, floor areas, building shapes, heights, views, etc.
3) Generative design: AI can quickly generate visual designs, such as photorealistic imagery, based on specified prompts. By employing generative design software (e.g. Stable Diffusion, Midjourney, Dall-e 2, Adobe Firefly), Architects can brainstorm various material and spatial design options at an early stage of the design. We can also use hand sketches or existing imagery as basis for more detailed designs, further iterations and/or alterations. Producing impressive imagery for clients may become much less costly and less time-consuming, even though there will undoubtedly be a learning curve ahead before we can use these tools in a way that results both in high-quality imagery and a thoroughly-considered design.
4) Pattern recognition: AI can analyse and identify patterns in large datasets, including architectural drawings, technical standards and historical designs. It is currently a challenge to compile and format a database in such a format that the AI can readily access and analyse but this will become easier over time.
5) Coding — customised apps, programmes, plug-ins: AI can help you write code and produce your own apps and desktop programmes, which could potentially help you collate your design libraries for database creation, among other things. (Hopefully there will also be open source libraries where architects share information and not just siloed and fragmented libraries behind corporate ownership.) It could also be used to develop plugins for existing architectural software, such as Revit.
6) Energy-efficiency and sustainability: AI can assist architects in designing energy-efficient and sustainable buildings. By analysing factors such as climate data, building orientation, materials and energy consumption patterns, AI algorithms can recommend design modifications that make buildings’ energy use more efficient and reduce their environmental impact.
7) Summarising data: AI can also help summarise information from books and reports for you, so you can find the key information you need quickly, even though you will need to review the original source as well (since AI tools can make mistakes and ‘hallucinate’). Currently some of the free tools don’t allow analysis of ‘long-form’ text, however, but have a word limit. ChatGPT is currently rolling out plugins (test versions) — available to all Premium (paid version) subscribers — which increase your options in terms of data that can be analysed, including live internet data.
8) Building maintenance: AI tools have already been developed to analyse video feed and recognise weaknesses in existing roads and building facades, for example, which will enable earlier detection of maintenance needed in buildings and infrastructure.
9) BIM and project management: Building Information Modelling (BIM) is a digital representation of a building project, incorporating data on geometry, materials, and other relevant information. AI can analyse BIM data, identify potential clashes or conflicts, optimize schedules, and assist in project management tasks, enhancing coordination and reducing errors during construction.
10) Digital twin technology: This technology utilizes AI algorithms to create virtual replicas/ simulations of physical objects, systems, people or environments. Digital twins can represent buildings, infrastructure or even entire cities. The potential of digital twin technology for Architects is too vast to be summarised here but in principle they can help collate, analyse and manage complex data in real time.
11) Virtual reality and augmented reality: Architects can utilize AI-powered virtual reality (VR) and augmented reality (AR) technologies to visualize and present their designs. These immersive experiences allow stakeholders to explore virtual 3D environments at different scales (including 1:1 scale), examine spatial relationships, and make more informed decisions regarding the design and layout of a building.
12) Natural language processing: AI’s natural language processing capabilities enable architects to interact with design software using voice commands or written instructions. This functionality can be used to fundamentally change the design process — with clever and complex prompts you can work with your own conceptual ideas in a new way. Tools, such as ChatGPT, Bing AI, etc. can help you develop new ways of perfecting your designs where the AI assistant thinks with you and expands on your thoughts, critiques your thinking and explores out-of-the-box solutions.
13) Cost estimation and material selection: AI algorithms can analyze historical cost data, construction material specifications, and market trends to provide accurate cost estimations and recommendations for material selection. Architects can leverage this information to make informed decisions that align with budgetary constraints and project requirements. To be able to utilise this, you may need to build your own database and find a way for the AI to access it, unless there is an open-source or subscription database for this information available in the future.
13) Improved design process: If we can design good-enough prompts for an AI tool such as ChatGPT or Bing AI to generate information from, it can help us process, re-process, re-frame and refine thoughts, lists, concepts, definitions, briefs, etc. We can use it to help us think outside a box and to critique our thoughts and conceptual text. We can ask it to analyse text from a different point of view. We can use it to search information quicker and more efficiently than a generic internet search engine would, although it doesn’t reveal its sources (and sometimes it makes things up while insisting it’s correct) so the information will always need to be checked, verified and referenced.
14) Macros: You can ask ChatGPT to generate macros for software you already use, e.g. Visual Basic for Applications (VBA) code for a PowerPoint presentation. The AI will then create the presentation first draft for you, on your chosen topic. You don’t need to know how to code, just where to paste macros on PowerPoint. If you’re not quite sure how to do this, ask ChatGPT to teach you step by step.
Will AI take over architect’s jobs? Thomas Lane in a recent edition of the Building magazine reports that AI can automate 37% of Architects’ and Engineers’ work. This, however, will likely mean that we will eventually automate many of the more mundane tasks and focus our time and effort on more strategic, complex, specialised and creative tasks — those tasks that AI is not able to do. Just as Revit, digital NBS specifications, 3D modelling and real-time rendering software did not overtake Architects’ jobs but just changed the way we work, the same is likely to become true of the various AI tools that will be rolled out over time. New tasks —e.g. to do with AI management — will also emerge.
A recent comment I saw online stated: Until AI wins an architectural design competition, we have nothing to worry about. (Source: Unknown) The contextual understanding of AI is still too elementary for it to be able to replace humans with something as complex as building design. It can definitely assist us, however.
While AI is unlikely to cause mass job losses for architects in the near future, it will fundamentally change the nature of our work and we should keep our skills updated to utilise the new tools and emerging knowledge, including AI, where relevant to our work.
This is the introduction to an article series about the various AI tools that are available for Architects today and in the future. We will also discuss some of the shortcomings of AI tools. Subscribe to keep notified of the next articles and please join the discussion — share your comments below.
