Loaded: The Hahn Ready Mix Podcast
A podcast for the employees of Hahn Ready Mix
Loaded: The Hahn Ready Mix Podcast
38. Artificial Intelligence
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
Andrea and Griff somewhat inexpertly dive into the exciting and evolving world of AI. We talk about how it can be used both at work or at home to make life easier, as well as the AI revolution's impact on workforce and jobs.
Also, open enrollment is here!
Welcome to Loaded, the Hahn Ready Mix podcast with Andrea Meyer, Griffin Hahn, and producer Lex. Hey everyone.
SPEAKER_00How's it going today?
SPEAKER_02Really good. Fresh, fresh off of vacation. Yeah. Happy to be here.
SPEAKER_00How was that?
SPEAKER_02It was really good. I love to get out to the warm weather when it starts to get cold here.
SPEAKER_00So we had some You picked a good time to be gone because it got really cold.
SPEAKER_02I I saw that. Uh not uh I was not sad to miss it.
SPEAKER_00Yeah. Yeah. You did it right.
SPEAKER_02What did I miss here besides bad weather?
SPEAKER_00Uh you missed uh people not listening to our podcast from two podcasts ago about cold weather and leaving water in the truck and then that water freezing and losing a load. I won't say who, but uh their name, their nickname um involves a starchy root vegetable.
SPEAKER_02Rhymes with mater?
SPEAKER_00Maybe. Yeah, yeah. All right. Well, what uh what announcements you got for this week?
SPEAKER_02Um I it's time, it's open enrollment time. So when this podcast comes out, the information will be shared on Paylocity. And if you need a printed version, please come in and ask us. But all of the updated um costs and any updated deductibles or anything like that will be published on Paylocity by Monday.
SPEAKER_00Great, great. Another thing you missed, and I'm we probably should have said this uh in our last podcast, but we're bad at looking ahead, uh, is uh happy Veterans Day to all of our veterans and um hopefully we're able to get out. We do a lunch um and and and celebrate there. And and we just appreciate everyone's service that has served the country, this country, and we're we're really proud to have you as part of the team.
SPEAKER_02Good. Um, it is the time of year to update our truck registration documents. So Angela and Lex here in the office are working on putting together new folders with all of the updated documents, and those will be going out next week. And um, so you shouldn't have to do anything to get it into your truck, but once it's in there, we want everybody to double check and make sure that the the numbers match the right the right paperworks in the right truck.
SPEAKER_00Yeah. Yeah. Great.
SPEAKER_02It's a fun project every year.
SPEAKER_00Enthralling. All right. What are we going to talk about today?
SPEAKER_02Oh boy. Well, uh, the topic you selected is AI at work, which is a great topic. I'm interested and excited, but I don't feel like either one of us are AI experts.
SPEAKER_00We're not.
SPEAKER_02Who here is qualified to discuss this topic.
SPEAKER_00We're not. We're not experts. But uh We're users. We're heavy users. Yeah. I think we know enough. We could have had your husband on.
SPEAKER_02He would have been I think that's why I'm hesitant because I feel like if he listens to this episode, he's just gonna roast me about all of my misinformation.
SPEAKER_00Maybe. Maybe. He won't he won't listen. He's too busy and important. He would have been uh I think he would have just got two in the weeds, so for sure. Yeah, for sure. You we would have both been sleeping. I mean, Lex always sleeps, but we would also sleep. Everyone would be sleeping. Yeah, yeah. Yeah. All right. Well, let's jump into this. Like, so artificial intelligence is it's it's the big thing happening, right? It's it's moving the stock market. It's what's causing the data centers that we see being built all over the country being built.
SPEAKER_02Everywhere but here.
SPEAKER_00Everywhere but here. Um it is, you know, it's it's kind of ubiquitous and and its use cases are are just huge, right? And and we don't even we don't even know what all it might be used for going forward. So we we just thought I just thought it would be a good idea to have a conversation about it and what it means for us here at Han Redmix, what it means for our industry, and what are we hoping to do with it or not do with it. And you know, I think there's a a big long discussion we could have on this. I don't know we'll get to everything, but um I think we take a stab at it.
SPEAKER_01All right.
SPEAKER_00Cool. Well, I think the first thing is when people think of AI, a lot of people originally or immediately go to uh what was the the artificial intelligence in in Terminator? What was it called? Something net, right? Skynet. Skynet, yeah. So people think of Skynet and Arnold Schwarzenegger saying, I'll be back, right? That was a good is that a good is that a good Schwarzenegger voice. Uh so anyway, uh and of course, in in that uh AI tries to exterminate humankind. Similarly, in The Matrix, right? That's so you know, Hollywood has this um uh very scary outlook for artificial intelligence, and um, and that's what we know. And and let's be clear, I don't think anybody can sit there and say that's a hundred percent off the table.
SPEAKER_01Definitely not happening. We don't know.
SPEAKER_00Uh but a a the people that are developing artificial intelligence um spend a lot of time on kind of building the guardrails for AI um and and kind of ethics behind AI, making sure that we don't lose control, right, of of what we're building. So um I think in general, we should be looking at more as an opportunity than a threat. And particularly when it comes to work and productivity and and just kind of general life convenience, right? So people think of AI as something that's brand new, and it's definitely not, right? The the uh sophistication of it is exponentially growing, the the use cases for it are growing, but AI has been around forever. I mean, AI what it is is machine learning, right? And um machine learning has has started in like the 1960s with you know mainframe supercomputers. So, you know, this it's not new, and and I think that maybe helps with with some of the fear um around it and on how you can deal with it is this like we're it's not new. So one one of the major uses of AI is making uh logistical decisions, right, that people are doing in in in business. Uh our command optimization model or our dispatch module, that I think that kind of hit the market in the mid-2000s. We were one of the first real implementations um in 2012 or 2013, maybe I think when we first did that. And that's got what's called the Ortech uh machine kind of under the hood of that software. And that's an AI. It totally is. It is learning from what happens with our customers, it's learning from um orders and it's making recommendations on which truck to ticket and um which load should come out of what plant based on past historical data and and predicting what has happened so many times before or taking data from what has happened before and predicting what's going to happen in the future.
SPEAKER_02Yeah, it for sure is. And it's kind of funny that they didn't call it that back in the day when we started with it. And now even things that aren't really AI, they're being renamed AI. So you think it is like the newest, coolest technology, but it's really just the same thing, repackaged.
SPEAKER_00Yeah.
SPEAKER_02So it's working both ways.
SPEAKER_00Yeah. You know, and I I think that there are a lot of different types of AI, right? Kind of what to what you speak of. Um, you know, there's really um kind of narrow AI, which you know, think of like your GPS or speech recognition, right? Or um, I don't know. Um there's a there's a lot of them where it has one use, right? It's one very specific one specific thing that it does, and um and so it's it's very narrow. You have kind of uh reactive AI. So think of like computer games, right? Where they, you know, you may have a character in a computer game or something that will react to what you do differently, right? So it's just reacting. It doesn't really learn, it doesn't really get smart or or or whatever.
SPEAKER_02Is that what is does that include like all the filters and the images and videos that all don't actually look I mean it looks like it's actually the person, but it's not them. Yeah, maybe in that category.
SPEAKER_00Yeah, yeah.
SPEAKER_02All the Snapchat filters, I feel like the middle schoolers don't even know what real people look like anymore because they're always talking to each other with a AI filter over it.
SPEAKER_00Yeah. So then you have like limited memory AIs. Um, so I think this is where our dispatch optimization would would fall into. Um, but that's you know, like self-driving cars. Um they they store some memory, like some some kind of operational or uh useful knowledge, but they're not, you know, like our our dispatch optimization, you know, can't double as our therapist or give us a a food recipe or whatever, right? So it's very limited in scope. Um and and the memory it knows is specific to you know the tasks that are it are doing, right? Um then you kind of have what what's kind of coming on now with what they're targeting is it's called theory of mind AI. And so that is where you have a computer that doesn't have emotions, opinion, consciousness, but it can understand human emotions, opinion, consciousness, right? So it can be um helpful for us by by interacting with us on a really broad um um measure. So it's kind of like I think that's kind of what ChatGPT is angling for. I don't know if they've quite reached that definition or whatever, but that's Chat GPT, Gemini, what's the um what's the Google one? Um Gemini. No, that's is Gemini the Google one? What's the Microsoft one then? Copilot. Copilot, yeah, yeah. So all these are supposed to be, I mean, that's kind of what they're targeting for those is is um uh AI that is um really kind of can can communicate with us across a wide range of topics on on kind of logical and emotional levels. And then like a kind of fictional, you know, end goal is a is a self-aware AI, which does not exist yet, right? So that would be that'd be your sky nets and your matrix, right? Is is an AI that that does have its own opinions and beliefs and and consciousness. So that doesn't exist, but um yeah. So okay, so we've talked about the types of AI. Let's talk about how we use it. I know you're a user of AI, Andrea. What what do you use? How do you use it?
SPEAKER_02My most frequent use of AI is for me to uh unfiltered rant or rave into it, and then it refines my thoughts into a suitable for public viewing email to send mostly to vendors that I work with. So it's it's really good at taking my rage and refining it in a professional way.
SPEAKER_00That's great.
SPEAKER_02I actually have it hooked up to my email so it also reminds me of previous emails that I've sent and gives me more bullet points for my emails that I send.
SPEAKER_00Yeah, perfect. We need to get Jeff set up on this. No, that's great. I mean, so we both have Chat GPT, I know, and um I use it all the time. I so the the thing for me that I think helps me learn about how to use it is to just use it as my search engine, right? So instead of going on to Google and saying, um, you know, how many times have we landed on the moon or whatever random thing that come across has come across my my mind, then just ask Chat GPT and you'll be astounded the um kind of the depth of the results you get doing that versus a Google search where you're you're kind of up to look at multiple sources on your own and it's gonna evaluate all those in the snap of a finger and and really um and gives you good answers. So that that's that's a huge one. I I always always a little careful about having it write things because I think especially Chat GPT has a very distinct oh yeah, it's obvious, obvious AI style.
SPEAKER_02Yes.
SPEAKER_00Uh, you know, it's always has the hyphen in there. There's always something that's got a hyphen. Um and so it is um I I'm I'm a little wary of having it rewrite, but I do think it's great to like proofread. Like, hey, hey, what you know, what went wrong here? But it's it's kind of crazy some of the stuff you can use it for. We I use Chat GPT to build the evaporation retarder or sorry, evaporation um rate calculation that's on our website. So it like wrote actual code and made a web widget that we put on on our website. It's just kind of astounding. Like it would have taken six months and twenty thousand dollars for um for for us to hire someone to do that. And I was able to do it in a day with Chat GPT.
SPEAKER_02Yeah, same thing. I have a couple of reports that we look at, you know, every day where the data's coming from different sources and the the formatting and the number of clicks it would take me to pull it together, you know, one time is fine, but doing it every single day, I don't want to waste 30 minutes doing it. And I was able to use AI to make it so that it it just happens automatically and I get those updates. The water added report, which I know the drivers love hearing from me when I get that report and ask them uh how much water we added and double checking the facts on that. And then also the um the truck inside assignments, you know, it's so dynamic and people are switching trucks and trucks are switching locations all the time and just being able to produce that for people really quickly whenever we want. Yeah, is really nice.
SPEAKER_00Yeah, yeah. Yeah, I've used it, I used it last weekend. I was baking a cheesecake and the the cheesecake was not turning out the way the recipe said. So I just took a picture of the cheesecake and gave it to Chat GPT and said, what do I need to do from here? And it just led me home, right? It just it it's it's kind of you can't do that with Google. I can't take a picture of something in process and say, help. And and AI can, it can help. So um, you know, I know there's there's like I'm not advocating taking medical uh uh recommendations from an AI, but there's been like stuff with my kids where I've been like, ah, what's it what is this bump?
SPEAKER_02Is this an emergency room visit or is this a we can wait till Monday?
SPEAKER_00Yeah, exactly. So um just there's there was a plant in our backyard that I didn't know what it was and took a picture of it and it identified. It's just it's like it can help with things um just wildly across the spectrum to make our our daily lives easier. Um so I think you know that one of the things to note is that it can make mistakes, right?
SPEAKER_02Yeah, it's a little bit of trial and error.
SPEAKER_00Yeah.
SPEAKER_02Where you have to ask the question and then check the answer to make sure that it actually applies or makes sense.
SPEAKER_00My favorite thing is you can ask it to check itself. Yeah. Right. You can say it just like you know, it tells you some answer, some question you you give it and you say, Are you sure? And it'll check. And sometimes it'll come back and say, Nope, actually, I made a mistake. And so um you don't even have to do the work. You just double just take the extra step and it uh yeah.
SPEAKER_02Another thing that we use it a lot for is um like software specific questions. Like if we're trying to use a program, it most of the time has already read the manual. So rather than like searching the manual yourself and trying to make it make sense, yeah. You ask the questions and it kind of leads you through it.
SPEAKER_00Yeah.
SPEAKER_02A lot of times it's right. A lot of times it gives you three different things to try uh before it gets there. One thing that we were trying to use it for that didn't, we didn't ever figure it out, but we were trying to use it for the clock on the river plant, you know, is wrong most of the time. And so we were trying to ask it how to uh how to update it, what the steps were. Yeah. And it had it actually did have a lot of the information, but the specific um mechanism that is there was not wasn't current enough to be able to use their programming.
SPEAKER_00But gotcha.
SPEAKER_02We got pretty close.
SPEAKER_00That's cool. That's cool. We use it all the time to come up with topics and talking points for this podcast, right? That's pulling the curtain back a little bit. But you know, it helps us come up with that. I've used it to develop a specific mix design before, right?
SPEAKER_02That was on my list of things I was going to bring up, but it um when you ask what areas AI is already really working on, uh optimizing mixed designs is like the number one thing that it's already doing. And there's you know, studies and programs through the University of Illinois and we, you know, even vendors that we've talked to where they're basically letting AI troubleshoot mixed designs or improve mixed designs for different areas.
SPEAKER_00So yeah, there's totally you could be replaced any day now. Well, that's a good segue here in a second. But uh, you'd think there is some neat kind of let's call it in-the-box solutions that are coming out now. Um, so companies that are may are are leveraging AI to do something specific, right? So one of those is mixed design optimization. And uh, so there's a company that we talked to where you send all your materials to them and they will run a bunch of tests on it and then feed it to an AI and to make the perfect mixed design based on all these inputs that we have, and um which is pretty cool. Not really like the ROI is not there. We don't use the same mix design quite enough to make that work, but uh the technology is really neat. And then there's others that were kind of taking an aggregate approach to it where you give them the information in the kind of the strength history and any other pertinent information about all your mixes, and it looks at them, takes that information and looks at all your mixes at once and then makes recommendations for certain individual mixes. That's pretty cool too. They even have um, you know, stuff where um you have a camera on the top of your plant taking pictures of aggregate stream coming into the plant, and it will identify gradations of aggregates like in real time um coming into the plant. And that's not the a like the engineering community is not ready for that, but um it's some pretty cool stuff uh about you know what's happening industry specific with AI. And and some of that, you know, costs will come down and those things will be uh more attainable as time goes on. But you know, I think the the kind of the general AI things are where the needle can be moved right now. If that makes sense.
SPEAKER_02Yeah, it does.
SPEAKER_00And I and there's there's definitely stuff uh coming for dispatch. I know one of the dispatch platforms we've talked to, it will know who you're talking to on the phone and automatically pull up on dispatch, like um this is who you're talking to. So these are the projects they've been working on, the likely mixes from the like they're narrowing the focus uh for dispatch right away, uh, which is pretty cool um to help with kind of efficiency and getting off the phone quicker and and all that. Um so a lot a lot of really neat stuff happening on that on that side. So we've talked a bit a little bit about the industry and I think there's a there's a huge fear out there that AI is coming for everybody's jobs, right? And I sat in a um a round table uh down in Muscatine a few weeks ago, and there was a number of companies there, and they all said, you know, some jobs are going to change with AI, but they said there's there's absolutely no way that it changes the number of people that work for them. And and these were some really big companies, way bigger than ours. Um, and I listened to that, and I uh I'll be honest with it, I just don't believe them. Um, I think AI is probably going to replace in in the larger workforce, right? Uh quite a bit of jobs because it's going to make um some of the more routine stuff, you know, a machine could do it. So I don't want to hide from that fact, I think, or or that that viewpoint. I think that that's very possible. The good news for us and for everyone listening to this that's an employee on ReadyMix is twofold. First, our industry is one of the most insulated that there could be from AI, right? AI is a long ways away from driving a ReadyMix truck, from fixing a truck. You know, we we have a lot of very hands-on things we do.
SPEAKER_02Even on the on the set on the administrative side where things are more automated and we use computers for them, it's just still kind of an ancient industry where even the data sources don't speak to each other really well. You know, I mean, we're still handling paper tickets from the quarries.
SPEAKER_01Right.
SPEAKER_02Uh, you know, yeah, kills me every day. But yeah, we are. Um, so going from, you know, sometimes still a paper and handwritten process all the way to AI is uh quite a leap. And we're we're making progress. progress on those things every day, but it's still a long ways away. And I also think another problem, you know, for our industry is uh just like for example, what when we're building something, it's on a road that isn't named yet. You know, like it's it's there's no real data self-driving truck can't get there's no real data to give the machine to learn from to make that next decision.
SPEAKER_00So yeah.
SPEAKER_02It's even when I think about like when we listen to our phone calls, there's so much background noise and uh incomplete data that we're given from our customers that like it takes a real person to think through that and get us where we actually need to be.
SPEAKER_00Yeah, for sure. And even beyond all that, even for um some tasks uh that that can be replaced by AI, I mean uh let's not forget we're a family small business and uh I'll make a commitment now that we are never going to let someone go just because there's a technology alternative, right? What how we'll handle that and how we've handled this in the past, because this is not new. Technology has changed over the last, you know, 130 years of our company, right? That there it's it has changed and we don't have people cutting open bags of cement anymore, right? So that job that used to exist doesn't exist, right? So that things will continue to change and evolve, but we are never going to just let someone go because there's a technology alternative. What we'll do is when someone retires um or leaves us for another reason before we hire someone new, we'll evaluate whether there's a a technology solution that that makes sense and move um uh responsibilities around or whatever. Because it's it's very unlikely that somebody that everything they do is you know perfectly encapsulated in some kind of uh technological solution, AI or otherwise. So so that is that is how we operate. That is how I think most small companies operate that care about their people. We're not just going to say, oh well we're getting rid of all of a certain segment of the company because there's an AI solution. There's just there's just no reality in which that that takes place. But as we incrementally as as you know people move in and out of the workforce um we'll we'll always evaluate what's out there um and how we could get better and more efficient using technology. So I just wanted to make sure I'm clear with everybody on that. So it's it's nothing to be scared of but one thing I definitely wanted to say is that I feel like sometimes people think like using AI is like cheating on a test in school, right? It's like oh I I'm looking up the answers is not so so for for me right now I'm telling you you have permission use AI to help make your life easier to make it better and and learn you know be careful and learn um as you do it of the pitfalls and double check your work and those things especially when you're getting started. But we want you to use AI. We want you to find time to take on other tasks and and time you know find better efficiencies find more capabilities by by leveraging technology. So I just wanted to to give that that pretty clearly that you have our encouragement to use it. Learn it share what you discover get get comfortable with AI. You you don't want to be the person um that's you know in the year 2003 saying this this email thing is never going to catch on and I'm not gonna learn it. I'm not gonna be a part of it right you you don't want to make yourself kind of obsolete by by ignoring what's coming here. So yeah so that's my encouragement on that topic. The other thing I wanted to talk about um was kind of the human side what what AI can't do right so it can't replace empathy teamwork character customer service the human element it can't replace right so we and I think that's a little bit why like I'm not a huge fan of using it for writing to like generate writing because I think the human You don't like communicating and writing that much well I don't text email or email like you would rather talk to someone face to face or on the phone than communicate and writing. For sure for sure. But I mean there's a time and place for emails and and all that and and but I just think you lose you you there's a a genuineness that you lose when you it's completely um you know AI generated or whatever. But you know that those human touches that people want to have that interaction right it's you know I don't see a a place where we're ever like having AI taking orders. I don't think that that would be good for us and that wouldn't be good for the way that our customers interact with us. We want them to have that human touch right so um just just know that you know the the it's the people that make this place and this industry special and AI isn't can't replace that.
SPEAKER_02Yeah I agree I definitely see it as an integration not as a replacement right so if we had AI that could double check the order and be like are you sure these elements are correct like yeah you know these are the things to double check or you know giving reminders or something like that. I think there's tons of opportunities for it to help us be better. Yeah but I definitely see so many issues with the considering it as a replacement.
SPEAKER_00Yeah and so on that topic I asked my chat GPT I said what what do you want to say to the employees of Han ReadyMix? Like what do you want to talk what do you want to say about AI as an AI? What would you say? And I'll read I'll just read off straight here what he what it said. It said AI isn't here to replace you. It's here to work with you. I don't get tired but I don't have instincts I don't feel pride in a good day's work. I don't shake a customer's hand that's what makes you irreplaceable what I can do is help with ideas, words, plans, reminders and patterns that make your job smoother. I can help you think faster, communicate clearer, and see things from a new angle my goal is simple to give you more time for what really matters people, safety, quality and growth. AI is not the future. It's a tool you already have in your hands. The more you learn to use it the more powerful you become the next era of ReadyMix won't be built by machines. It'll be built by people who know how to use them wisely I wish I could speak that powerfully it's a good thing you have it as a tool. Yeah exactly exactly so I just thought that was a really neat statement and and I do think it underlines you know what where we can go with AI. And so yeah I think the the gist of the whole thing is is we want people to to get used to using AI to learn. If you want you know particularly if you're like a computer user right here at work if you want advice on how to get started with it, come see me, come see Andrea and um we'll help you get off the ground a little bit and um I think you'll be kind of astonished by um how it can change the way you work day to day anything else you want to talk about on the subject?
SPEAKER_02I'm sure that this subject will come up again in the future and I will um I'm sure have input from my husband to share all right we'll wait for that next week. All right thanks for listening to Loaded the Han Right Mix podcast. Please remember to subscribe and share this episode with your friends.
SPEAKER_00Thank you.
Podcasts we love
Check out these other fine podcasts recommended by us, not an algorithm.
