Issue link: https://iconnect007.uberflip.com/i/1538540
14 PCB007 MAGAZINE I AUGUST 2025 Charles Koch likes to talk about how we're adding value for our customers, which is something we all need to do. AI is a hot topic and an important tool. Where do you see the impact of AI? AI is a really interesting new tool, but I'm not sure how intelligent it is. I agree that it's artificial and a huge con- sumer of electricity. AI resources gobble up hundreds of megawatts at a time, and whenever they go to relearn, their loads may go up by maybe 20–50%. AI itself is in its infancy. We must look at how we can use computational elements to view huge amounts of data very quickly, either on our phones or through typing, voice, or otherwise. It offers immense value. It really does. Someone shared with me that when they wanted to write an app for a digital wrist- watch, they used AI to do it in a matter of minutes, as opposed to sitting down and learning the pro- gram or writing the code. That's fascinating to me. Sooner or later, the way we're hammering at it now will be self-limiting, maybe by the amount of electricity, until another breakthrough happens. I'm a firm believer that the world's most advanced circuit boards will be designed by AI soon. It's a mathematical computation, a set of parameters. It is rule-based. Do you agree? It may be seen as more of an art than a science. It can be a science using AI if you can model it entirely based on a set of rules and dimensions and spacings and thicknesses, all the stuff that goes into circuit boards. Industry pioneer and veteran Happy Holden worked for HP in the 1970s and is considered to be the father of HDI. He had the rules-based model doing circuit board designs, and he'll tell you, this is nothing new. No, it isn't. As a college intern in 1967-68, I worked on a classified project for the government, design- ing a machine. We had a rules sheet. Now, it's just changed from being done by hand to being highly automated, which everybody just takes for granted. That's tremendous progress in 50 or 60 years. Now, think about 40 or 50 years from now. It's hard to imagine where this will lead us. It could be that a new integrated circuit pops out just like getting a Coke from a machine. What problem do you wake up thinking about? Not because it's urgent, but because it's worthy. When I'm doing things the way I like them to be done, I might get up in the morning, make coffee, and sit down in a comfy chair at home. I won't look at my phone, but instead just sit and be a little quiet. All the noise starts falling away, and then lit- tle signals start popping up in my head. This morn- ing, I thought about this interview with you, then about getting new eyeglasses, and then I thought, "There's a whole bunch of things that came up about our test switches, and we probably use an average of nine of them in every control protection panel we produce. How could we make them bet- ter, cheaper, faster, simpler?" My thoughts continued, "We started some work in the area of early detection of autism based on pupillometry. What do we have to offer there?" Every day it's like that. On a good day, I just try to