Chester Thompson, WSP assistant vice president and San Francisco lighting studio lead, shares his insights and questions on the use of AI in industry development.

As application of artificial intelligence (AI) technology continues to expand, professionals across different industries are asking themselves: How will these advances impact our industry?

It is also a conversation that my lighting industry colleagues and I have been having recently. When it comes to our specific discipline of architectural lighting design, we have started to see AI used as one of our many tools.
For example, AI image generators can quickly create numerous design concept images, like how large language models like ChatGPT, Bard, or Copilot are able to generate text-based content. These generated concepts can then lead to richer conversations with clients because they can guide us to what they like, and what they don’t like.
Right now, we are at the forefront of this development, and there are an increasing number of questions and concerns regarding how AI could affect our project deliverables. But I believe that AI is a tool that can support our endeavors through iteration, task management and cost savings.
If we combine this AI support with our own personal, experience-driven creativity, then my colleagues and I could improve our efficiency at delivering quality work.

HOW AI ACCELATARES THE CREATIVE PROCESS
As it stands now, the use of AI is limited to applications that operate like engines fueled by relevant, available data. That’s the case with image generators like Midjourney, DALL-E or Firefly, which are trained on images of existing lighting solutions in order to generate new options.
These engines can iterate at a rate that humans aren’t able to match and in doing so, this can help accelerate the creative process for a project by inspiring new ideas and possibilities.
It’s particularly useful in cases where there’s a design that needs to comply with a very specific set of design standards. An AI tool can work within those parameters to create a list of compliant solutions for a designer, and then the designer can use those options as a springboard.

However, a human still needs to be in the driver’s seat throughout this process. AI can produce different iterations faster than a human brain can, but these creations are both limited by the training data provided and ‘good lighting’ is often subjective.
More importantly, AI lacks the human experiences that are necessary for the most inspired works, such as truly outstanding lighting designs.

PRESERVING THE HUMAN EXPERIENCE
As a design professional, I believe that our experiences inform our work. When you walk into a high-end restaurant and find yourself in a dimly lit room with rich reds and oranges overhead, it speaks to your memories of sitting cozy by a campfire. It’s that thoughtful response to the human experience that we bring to the table as designers.

You can program AI to mimic certain colors, sure, and you could go deeper into biological response evaluations to better program AI, but you’re still missing the creative spark that comes from a designers years of experience honing their craft.
Furthermore, today’s AI technology uses historical data for training, which inhibits the technology’s capacity to respond to our ever-evolving culture and trends.

CAN AI REDUCE LIGHTING COSTS?
In many ways, AI is simply a new version of automation. In cases where you’re following a design concept with known lighting solutions, AI applications can quickly recognize the patterns in those solutions and complete these repetitive tasks automatically. This can save both time and money on any applicable project.
There are also opportunities to rethink our current workflows and toolsets to capitalize on AI’s capabilities; for example using a conversational interface to direct application of prototypical solutions like wallwash, graze or downlighting, and automatically generate objects in a design model.

It also presents a promising opportunity for me and my colleagues to bring quality lighting solutions to more people. As it stands today, only a small percentage of the world is afforded lighting solutions that enhance day-to-day living. By reducing the cost of these solutions using AI techniques, more people would benefit from this basic and essential quality of life improvement.

Source: https://www.wsp.com/en-us/insights/2024-artificial-intelligence-for-lighting-design