Where AI is Headed in Workplace Design

By Jennie Morton | May 22, 2024 | i+s Magazine

No industry is immune to the current wave of artificial intelligence, including architecture and design. The general consensus is that while AI for workplace design has great potential, it needs more time to mature. But the inroads are there, and the industry is headed for a significant pivot.

Applications with Results

Today’s most viable AI applications are for visualization, especially repetitive tasks or skills that are complex to learn.

“We’ve been stunned by the AI plugins for Revit and SketchUp to create visual renderings and wireframes based on a prompt. We can see how this will quickly supplant traditional rendering processes,” reported Jason Rosenblatt, senior design strategist for MillerKnoll.

“AI could be like a copilot, supercharging a designer to do more with their time,” forecasted Faizan Zaidi, Spectorgroup’s director of design technology. “For example, AI could generate 30 layout iterations, then a designer selects the best to move forward.”

SGA has tested 200 AI tools so far, and only a handful of image applications have risen to the top.

“We’re using AI to quickly develop ideas that are plausible enough to get client feedback. The photo-to-video and animation tools are pretty interesting, helping to put a face to a space,” said Michael Schroeder, SGA’s chief technology officer.

Kimball International is even having AI improve communication for their CEU presenters and sales team.

“We’re using PowerPoint’s rehearsal coach. You can record yourself and get real-time feedback, such as on non-inclusive language and areas that need more context,” according to Kaelynn Reid, certified futurist for Kimball International.

It then creates renderings without having to assign materials and textures to each component.

In-House Models

Since off-the-shelf products aren’t tuned to a particular design ethos, AI will likely gain the most traction as in-house models. Firms already have data from past projects for training. Until this becomes more economically viable, look for simple tools that can be customized.

“We paired AI with our internal wiki–you can ask it a question and receive an answer in seconds rather than manually searching,” Zaidi explained.

“We have our own design app that creates a programming document,” added Rosenblatt. “You can input types of spaces, square footage and headcount, then calculate how many 6×6 workstations will fit or adjust to increase private offices by 10%.”

This Midjourney-generated rendering showcases a hypothetical project for a tech company as envisioned by an AI Design Charette at SGA.

Continuing Education

AI will need to become a core competency. Whether through a formal program or individual time, encourage staff to experiment with tools and share insights.

For example, SGA launched an internal charette, where six teams created a project exclusively with AI and presented results to the entire company.

“I framed it like the Shackleton expedition to the Antarctic–challenging without many resources,” said Schroeder. “These aren’t prime time tools yet, so expect the unexpected. But it was important that we start seriously investigating AI.”

This Midjourney-generated rendering is from an AI Design Charette at SGA. Employees prompted this generative tool to deliver a hypothetical project for a tech company.

Ethics and Privacy

What happens if a firm trained a model using only images of a well-known architect’s work? What are the ramifications of generating images from AI that scraped photos regardless of copyright? It’s a Wild West at the moment for privacy and ethics. It’s possible the best way for firms to protect their IP is using a closed model.

“For example, in-house AI could safely use all your project photography, read proposal language or work with confidential client details,” Rosenblatt said.

A private model might also help with bias, ensuring the information doesn’t have predispositions toward certain occupant types or hinder efforts for universal design.

“I’m an optimist of what AI can be in the hands of a good steward. However, we as a design community need to be champions of transparency and accountability,” stressed Reid. “We need to ask hard questions about AI’s equity issues so we don’t build environments that are oppressive to any groups.”

The Near Future

This moment feels similar to 2007, when smartphones hit mass adoption and everyone rushed to launch a mobile app. AI is on a similar trajectory, with hardware, UX and UI catching up to the hype.

“But everyone should prepare for a future where things have significantly changed,”
emphasized Schroeder. “We’ll need to think of AI as a teammate, not a tool, because its cognitive capabilities are orders of magnitude greater than ours.”

To those concerned about job security or the impact on creativity, AI can be viewed as a learning opportunity.

“Remember we’ve already been through a tech shakeup when AutoCAD replaced hand draft. Digital tools only make the industry grow, not shrink,” Rosenblatt reassured. “I could even see where the client becomes more involved, starting the design process with a basic prompt: business type, aesthetic and square footage.”

We might even cross an uncanny valley where AI is fully embedded into the built environment, become responsive to its occupants.

“It’s not going to be as obvious as a robot, but AI will be integrated into a building’s daily operations,” Reid projected. “Designers increasingly will manage data sets to create tomorrow’s workplaces.”

About the Author: A former i+s editor, Jennie Morton is a freelance writer specializing in commercial architecture, IoT and proptech.

To view this article in its’ entirety, click here

Written by Anne Holden