
1 minute read
Thomas Paterson
DIRECTOR, LUX POPULI MEXICO CITY
Paterson studied AI as part of his mechatronics degree at the University of New South Wales in the 1990s. He believes the lack of meaningful data sets will hold back its effectiveness in lighting design.

‘AI is going to have a huge role in small optimisations, in taking the principles of a design and working out the right densities of lights in a system, working out the right timings etcetera.
But I think we're further away than people think, because the things that are dazzling people now like ChatGPT are fundamentally working in a space where there is a massive data set which is highly evaluated. So you can tell what's good writing and what's not by how much it sells, by click counts and so on.
But how do you actually evaluate whether a lighting design is good? Otherwise you're just steering towards an average of what's been published.
In terms of actual design thinking, it's millions of miles away because there's no data set that actually qualifies whether a lighting design served the client's interests, whether it worked with integrating with the facade or the building systems, with any of those aspects.


It’s too complicated, and there are just no data sets for lighting that are meaningful. Who can feed in 100,000 examples of projects with their documentation and with evaluation of what worked and what didn’t? Those data sets don't exist and a lot of that information is confidential.
So I think it's going to be a useful tool. You should be able to identify which walls are wall washed and with what type of product, and what type of downlights, and so on. And in principle, AI should be able to lay out your plans, do all of your load scheduling, write your spec. There's a lot that it's going to be able to do.

I'm not particularly excited about it any more than any other tool, but I don't fear it. At our end of the lighting design market, we're doing truly crafted, truly detailed lighting design. People come to us because they want us to do innovative thinking. And innovative thinking means understanding stakeholders’ interests, understanding the history of a place, understanding how something is built, understanding how it will be constructed, which are two different things.
That huge breadth of knowledge, plus the knowledge of psychology for brand and identity and all these sorts of things are all dependent on much research and knowledge, and we exist because people want that. Otherwise, I’ll go to a sales rep lighting designer. So a lighting designer who is threatened by AI is not a very good lighting designer.’ ■