I had the pleasure of supporting an LA-based event this month and discussing how AI-based workflows will affect office designs as we move into the future.

Most talk of AI in the AV industry has to do with sensor based automation, like personalization and camera tracking and switching in meeting rooms. Most of the buzz around AI in the greater world is around generative AI with tools like ChatGPT and Midjourney.

I think we all see the potential for generative AI as it applies to anyone in the business of content creation. Writers and graphic designers are grappling with what generative content may mean for their careers and industries. The truth is that generative AI will affect a much wider group of people, researchers, paralegals and code writers all come to mind, but it’s hard to imagine a workflow that won’t be impacted by AI at some point.

My co-presenter in LA, Steve Weikal from MIT’s Center for Real Estate and Industry Chair for their Real Estate Transformation Lab, shared that AI will eliminate 83 million jobs, while creating 69 million new ones. I’ll add that it will also change tens of millions more. I will argue that AI has the promise of doing a lot of the tedious machine-like work that people have been “forced” to do in a pre-AI world, but is actually better done by machines anyway.

Think about the average construction project. We often have to create our schedules backwards, starting from the occupancy or completion date, backing out time for construction, then equipment lead times, then the design phase and seeing what’s left for discovery with the client.

This process often truncates the most important part of the construction project, being the intent and outcomes desired by the owner. Many projects suffer from this lack of time in the front end of the process, resulting in a building and workspace that is technically sound, but not ideally suited to the people occupying the space or the specific workflows used or work being produced.

Lack of access to the right stakeholders is often prohibited by the timeline, meaning that we get a limited view of the organization’s needs filtered through the point of contact for the trade we happen to represent. If there is one thing I often hear from programmers or those creating technology User Interfaces, it is that they are designing in a vacuum, unable to talk to the people who will actually use the system.

However, the promise of generative AI is that it can truncate the design and production phases of the project, giving that time back to the front end for better discovery on the programming phase, netting a better end result.

If you’re creating a UI, instead of spending hours typing and debugging code, imagine having time to do stakeholder interviews, and then use natural language to prompt the code to be generated in the appropriate programming language. Or if you’re a design engineer, providing equipment lists and single line diagrams, and letting AI do rack layouts and fully engineered drawings based on the equipment list and I/O specs.

The human is still at the center of this process, choosing or curating the proper large language model, creating natural language prompts to generate the outcome, and reviewing the output for mistakes and/or subtle changes that may make the user experience better, even when the generated designs are functionally correct.

Of course I’ve drawn an example as to how AI could positively affect technology outcomes, but the same principles and efficiencies have similar implications across industries.

Given that AI will definitely affect our workflows, and our physical spaces should be designed to support the human occupying them and the work they’re preforming, there are obvious implications to the types of spaces we should be designing as well in an AI enabled future.

I’ll tackle some of those implications in my next blog.