Factory of the Future Company Story: Dunmore

Ronn Cort: We are a traditional like Dunmore is a very traditional B2B industrial manufacturer. We're talking, when you think like classic manufacturing, that's us. Really kind of like, you know, um, not highly automated, not a lot of robotics, very much the noble work of people on the on the machines doing manufacturing. We, we produce things that, that keep this world safe. You interact with our products every single day. You'd never identify us. But if you popped open a bag of coffee today, the, the package that it came in, we make those materials. If you opened up your washer or dryer in the last 24 hours, and there's a label there which says, Remember to clean your lint, trap. We make the materials that make the labels really not very sexy stuff, but Things that protect people every day.

All of a sudden there's this, there's, this technology that starts to, to, to, you know, to show up. And here's how this happened for me was, when I went around on my organization and just started to talk to people and say, hey, how do you use it? What I heard was they were like hesitant to use it, and the most common thing was like, this is gonna take my job, this is gonna take my job. There's a lot of that kind of conversation. That was like the narrative. And, and Ed said it in an earlier conversation we had had, which is like, AI does not take your job. Someone who knows how to use AI effectively takes your job.

What we have done is we've brought our AI agents in-house. We're not using a public, you know, we're not using, we, we've got our own licenses. We do not use our data to train the LLMs. Right, we, we restrict that. We're very specific about which agents we allow in, like, we will not use DeepSeek because when we look at the requirements around it, it, they own, they own whatever we put in there. OpenAI and Anthropic, which is Chat GPT and Claude, both of them have very robust, protocols around control of data and how they use the data. And we don't allow, we don't allow those models to be trained by our data.

You know, we've been struggling with like, how do you get all this knowledge and transfer it? I've got this generation that, that only knows things verbally, but they've also got an incoming generation of new workers, 19, 20 years old, who expect everything to be delivered in video content, short form, 2 and 3 minute videos. That's how they learn. The dichotomy, the gap between those two learning styles is huge. So, one of the really interesting things that we've done is we are making up our production workers and sending our younger workers out there with an iPhone and a video as a video camera. And all the production worker now has to do is explain what he's doing step by step while we have video. You take that video file, you dump that into an AI agent that's well structured, talks about how to write your SOPs. I want to get a detail, and you take that audio file and you that, that video file, and what you get back is you get back a written standard operating procedure, and you get a script. And how to actually do that, how to run that piece of equipment, and you actually have a video edited video with voiceover that is all generated by AI.