Speaker 0 | 00:02.708

Welcome back, everyone, to Dissecting Popular IT Nerds. I’m your host, Doug Kameen, and today I’m talking with Paul Falby, Director of IT at Cassens Transport in Edwardsville, Illinois. Welcome to the show, Paul.

Speaker 1 | 00:14.996

Hey, thank you. Nice to be here.

Speaker 0 | 00:19.459

So we’re talking about pronouncing last names before we got on the recording here. And everybody has, this is always one of those big things that happens with, you get somebody. And then you have to spend a minute looking at it to be like, how do I pronounce this? Is it going to be some interesting pronunciation or not?

Speaker 1 | 00:41.614

At least I say it. My last name is Falby. You don’t know how many emails I get to Fable and Talby and everything else. So I’m so used to, like, my brother says our name different. He says Falby. And I say Falby. Most of us in the family say Falby. More of us. higher a or whatever. And it’s just interesting. And like I said, it gets to where, you know, I’m getting, you know, getting older and I don’t care anymore. Somebody says Fable or something like that. You’re not going to aggravate me.

Speaker 0 | 01:17.722

So, so my, my best friend, his, he’s from Serbia and his last name is spelled. I’m going to spell it first and then, and then I’ll go through the, how it’s actually pronounced. M I L A S I N. O-V-I-C. So this is not how it’s actually pronounced. Mila Sinovic is how it reads. Absolutely not how it’s pronounced. Mila Sinovic, right? You would get that, right? Everyone would get that, right? I mean, I’m Doug Kavine, but everybody wants to say my name is Kamin. I will randomly run into people who will be like… Oh, hey, it’s your, oh, Mr. Kameen. And I’ll be like, wait a minute. How’d you know to say my name? Right? Like, it’s like a thing when they actually know. And I can, and I’ve always, in the back of my mind, I’m convinced, like, depending on where it happens, if it’s because like, their hospitality system or something has noted that, you know, like when I go to, when I go stay at a hotel, and they’re like, oh, welcome, Mr. Kameen. And I’m like, did Hilton somewhere put in the system that like, it’s pronounced this way so that they can impress me when I show up?

Speaker 1 | 02:28.858

I bet you they do that kind of stuff. You’re right how names can be really different than what they’re spelt or anything. When I was growing up, down the street was a man, and he had sons that I played with, from Lithuania. We said his name was Gribinus. But there would be, if you go to, we used to call them Lithuanian luals, you know, they’re parties.

Speaker 0 | 02:58.134

Oh, that sounds fun.

Speaker 1 | 02:59.435

And some of the people would be… calling him Mr. Rabenis. We said Rabenis because his last name started with a G. People would call him Rabenis. They’d just drop the G and do it that way. It was crazy. Those Lithuanians, they were something else too.

Speaker 0 | 03:20.106

They know how to party, right?

Speaker 1 | 03:21.867

Yes, they did. We lived about a quarter mile away and you could hear the accordions in the background and them singing. and stuff because they were drinking alcohol and stuff like that. It was fun.

Speaker 0 | 03:34.971

Oh, man. Those types of memories. That’s great stuff right there. So, yeah. So, I mean, this is one of the challenges of, you know, these are very, we’ll call them high-class problems, but living in America with all these different people who have all these different names and from all these different cultures and how does one culture pronounce it versus another? When did it change? You know, a lot of our names have been… Americanized over the years as our families have lived here for successive generations and stuff like that. So, yeah, that’s just, it’s always a crapshoot whenever you start looking at somebody’s name. So, but we’re here to talk about you and leadership and IT leadership in particular. So I’m gonna, I’m just gonna point out, you have a long career at Cassins. This is not… This isn’t like a new role for you, if you will. I mean, you’ve been there a long time. Can you tell us a little bit about this? This is going to be great to talk about for somebody who’s really spent a lion’s share, almost all of a career in one spot.

Speaker 1 | 04:40.458

Well, the Kasson’s family is just a wonderful family. They really are. This is a family-owned business. So titles here aren’t that big of a deal because I’m never going to be like… the president of Kassens because there’s family members and stuff like that. And that’s good. They are wonderful to work for. They’re very worried about their employees and stuff and the benefits and such like that. So the balance between work and home life has just been great. I had kids later on in life, so I wanted to be able to be there at ballgames and coach their ballgames and such. Cassins was a really good fit for me that I could be at home and run a 5 o’clock practice in the afternoon and such like that. So that’s why, like I said, great people to work with and that kind of situation. That’s why I’ve been here this long. And the trucking industry sometimes gets a bad rap. They don’t think it’s very sophisticated. But. Trucking is actually, it’s like a, it’s a lot of, especially in ours, we, we move cars and it’s a, it’s kind of like the grocery business where it’s a high revenue, low profit margin business. So you’re always scratching to figure out how to save a little money. Yeah. So you’re always looking into, you know, we’ve been doing EDI as soon as, as long as I’ve been here since 1991, it was over like 1200 baud modems back then. And now, you know, we did a study years ago that we talked to our main customers, which are the Chryslers, the GMs, the Fords, 40 different ways sometimes. And it’s like sometimes we send the same data to them in different formats. And it’s just you just have to adapt and figure out what tools to use to do it. And it’s such like that. So it’s been really great. You know, it’s just the other day I was looking at a Perl script I wrote. back in 92 or 93 and i can see some and then i saw that some code in there i was like yep i put that in for year 2000 yeah like we could build something in the 19th century and our 20th century and you know so it’s something that had to be had to put a 19 in front of it and when they pay it it might have had to put a 20 in there because they didn’t send the century in the format and that’s the way they still do it if you can believe that they never have changed their feed to us from this manufacturer that they don’t even tell you what century you’re in you know so so you’re your own y2k resolution operation too yeah you know if my if i if i stay you know working long enough i’m going to be working uh work through y2k and the in the uh seconds the epoch thing so which is looming i forget is it 2034 20 yeah when is the epoch

Speaker 0 | 07:59.075

Wait, it’s consulting the internet, that series of tubes to figure it out. Oh, and this is this 2038.

Speaker 1 | 08:08.223

Yeah, there you go. Conceivably, I could still be working then. It’ll be something. I think that one will actually could be worse than Y2K. We prepared everything. And, you know, it was, you know, for Y2K, I remember driving up here. I live about 10 miles away. Almost got wiped out by a deer walking across the road right in front of me before I got here. But yeah, we got in here, you know, after midnight and went around everything. Everything was working good. Went back home again. But the Epoch thing, I think, could be a little more serious, especially with all this IoT stuff. They’re not building 64-bit, you know, making sure that they’re doing 64-bit operating systems. It might be a little more problematic.

Speaker 0 | 08:55.406

So. On the leadership front, 33 years so far at Kasson’s. You mentioned it’s a great company, a great family company to work for. Knowing that the family runs the business, so that’s not going to be your future there, but you still get to be a leader to a lot of other folks in the organization. I mean, Kasson’s not a small company. I did some cheating and homework. I looked it up. You’ve got sites all across the northern U.S. Yep. And, you know, so this is a pretty big, sophisticated operation, lots of employees or staff that are looking to you, staff on your team. How have you led teams? You know, maybe I’ll be more specific than that. You’ve been there 33 years. So my question, I think, is first, what traits have you really taken from working in a family business that you think are the most important for leadership?

Speaker 1 | 09:52.061

Getting to know your employees. You know, there’s… Management is a lot of getting, you know, like you got a set of expectation level. And, but on a smaller team at the same time, you can’t, there’s a lot of camaraderie, you know, you, you end up going out for a drink or two, especially when you’re younger and stuff with everybody. And even though you’re, you know, you gotta, you gotta get into the people, you know, so you get a lot of bonds between the people. At the same time, you need to let them know that, you know, when you go out, out, out that front door, we’re. on the same level we’re friends or whatever but when you come inside work is work out you know play is play you know just because i may play softball with you that doesn’t mean you can get away with x or y or z you know work at work friendship outside is friendship outside that’s what i’ve always tried to impress on people and you know it you As long as you set that out and you’re consistent with it across everybody, you don’t have so much problems. You know, you got to listen and you got to listen to everybody, you know, no matter if you don’t know, because, you know, you might miss something. You know, you might think, oh, this person never has any good ideas and you don’t want to listen to the other idea. Well, you need to because, you know, inspiration comes from everywhere, you know. So sometimes, you know. that one time that person has the great idea might be the biggest thing to do, the best thing to do. So you’ve got to be listening, set your levels of what is good and what is not, and stick with it. And I really think getting to know the people is important. Try to know their husband’s name. Try to know their kids’names and stuff like that. Now, like I said, Cassins is a big place. big company, but our IT, it’s not so big that you can’t know it. Now, this formula may not work for everybody, but I’ve been in charge of different people ever since I’ve been 17 years old and basically treat everybody like you want to be treated and get to know everybody and have your level set has seemed to work for, I don’t care if it’s, when I… worked in the bakery and was on the cleanup crew, led the cleanup crew, or in college when you ran the tutoring department, or Cassons, or running a baseball team, or running a men’s softball team, which can really be something. As long as you treat everybody with respect and stuff like that. The easiest thing all the time is to rule autocratically, to be in charge and listen to other people’s opinions and do that. It’s a heck of a lot harder, but it has a lot of benefits because it just seems that people are more cohesive that way.

Speaker 0 | 13:02.300

So I think about a couple of things. I’ll just add some color or commentary on that you said. So first, you said it might not be the way for everybody to run things, but that’s the joy of having so many different businesses out there is that there really are different. Each business has its own culture and its own lane and what works for it. So one time when we talk about a business that, you know, has, they get together and have, I’ll call them, I’ll say this somewhat facetiously, but feeling circles, you know, they get together, they’ll talk about, let’s talk about what’s happening to our problems today. Let’s talk about these types of things and let’s get it all out. That’s not going to work for company B, you know, and that’s fine. You know, their culture is just, it’s different. So every organization has got its own way of doing things. uh you mentioned that you’ve done a lot of things over the years of different leadership there’s a friend of mine who he was in the military and he he told me that when he he goes even even the people that are the way they teach leadership there is when they’re even when they’re swabbing the floors they put two people in on the job because one’s in charge it all it all points in time and leadership doesn’t start like even the smallest jobs there’s still a leader somewhere in some way and You know, you’d mentioned that even at 17, you were nominally or in factual application, been in charge of somebody in various roles for a very long time. So there’s a lot of experience. And then I think the other thing that I noted, when you were talking about outside and inside the office, and I work at Mid-Size Enterprise myself. So your comments, I think, resonated with me is the idea that the department is not large enough. That we become somewhat anonymous, you know, like you can’t you can’t just be like, hey, there’s the guy who runs the place or the woman who runs the place. And then there’s like a bunch of staff and they don’t really communicate or talk to each other. We’re small enough that that. We do. And you’re going to end up talking and communicating outside of the outside of work. And, you know, there needs to be that like distinction to be like, hey, particularly if you live in a smaller city or smaller town where you might just literally interact with these people, because maybe we’re on the same kids on the same little league teams and just other stuff like that. They go to the same school because we live in the same community and you have to be able to make those distinctions to serve. I would survive isn’t the right word, but, but to thrive in that environment, you have to be able to set aside, like, this is the work stuff and this is the outside, the work stuff.

Speaker 1 | 15:40.377

Yeah, exactly. You know, like I run a, uh, I’m the president of our, of my town’s baseball and softball league. And the other day, one guy come up to me and goes, do you remember me? And I’m like, I’m sorry, sir. I don’t. And he’s like, well, I interviewed you with you back, you know, 10 years ago. I didn’t get the job and all that. well, I hope I was nice. And he goes, well, yeah, yeah, you were very nice and all that new stuff. I just wanted to tell you, I remember, and I remember that thing. I said, oh, you know, cool. Great. You know, and you know, that brings, you know, like, and just more and more leadership thing is like, you know, like I have a baseball team and it’s, and it’s the same here too. You know, I teach the kids. I say, you know, play like you want to be the person that even if you’re in right field, you want to be the kid that somebody goes. man, who is that out there? Because you didn’t, you know, the way you’re playing and, and just like with your, if you’re a coder in, you know, I, I T and stuff like that, you want somebody to say who wrote that code? That is, you know, it’s, it’s, it’s succinct. It’s, it’s neat. It’s easy to understand. You know, it’s not, if you know the language, it’s easy to understand, but I mean, it’s written in a way that somebody can come in and, and figure it out that they have a decent background. You want to be that person, you know?

Speaker 0 | 16:59.153

So. But of course, nowadays, I’m going to take a right turn here. AI is really good at writing code. Of all the things that AI is, like, there’s, you know, we have a lot of conversations. I used to be like in IT about the coming of AI tools and the change it’s going to make and how, you know, oh, God, it’s going to replace these jobs are going to do this. And there’s a lot of hyperbole about, you know, stuff that. its claims will change but is not really changing at this point maybe in the future it will and that’s that you know the future is the future but like in today’s environment it’s not replacing you know these jobs but one thing that ai is the current generation of ai seems to have been particularly well suited for is becoming like a coding assistant for folks i don’t know if you you know you mentioned because you’ve done some coding so i’m just not aware yeah i still

Speaker 1 | 17:54.910

even though I’m leadership, I try to do 50% or 75% coding still. I like, even though I don’t do some of the newer languages as much anymore, I like to know what’s going on. I like to know how, you know, that what looks good and what doesn’t and what works and what doesn’t. And I, you know, I actually, nowadays I don’t, I don’t use VS code or anything like that. I use right now. I’m, I used VI for many years and I’ve been using, I use them. And now I’m, I’m going into NeoVim and it’s got that copilot now with Kodium. And some of the languages, it’ll suggest and stuff. It’s like, it’s getting better and stuff, but sometimes you ask the AIs to program something for you. And if you really know the language, you’re like, yeah, that’s not the best way to do something, but it may work. And that’s what I’ve been saying is that those coders like that, it’s really the trainers and the data that… the samples they have to work with. Like if, say, I’m not a big Python programmer, so if you trained a model on my Python scripts, you’re not going to get as good a code as, say, Guido, the guy who wrote Python. If you trained it on his scripts, you’re going to get a lot better quality suggestions than if you do it on mine. So you always got to take that for granted.

Speaker 0 | 19:20.264

I think we trained it on everybody’s scripts. That’s the whole point. that’s the that yeah that’s its own that’s its own problem is they go vacuum up everybody’s work and then they use it to train the models which you know it’s like you know maybe we should all be compensated for that in some way right but exactly yeah you know it’s a whole different discussion there but i the i think the it’s been one of the more uh i don’t say called robust use cases is that these these ai tools are are good partners for for code so that you know so i’ll be careful not to say like hey it’s going to replace all these coders because it isn’t like it makes mistakes it it doesn’t you know it doesn’t logically figure certain things out it requires review and it requires you to look over it but relative to the effort it took to write the code you can save yourself a bunch of time by getting something to start with and then work backwards from there which which for you know a lot of times could be easier

Speaker 1 | 20:08.042

Oh, yeah, I will do. I’ve done that myself, too. I’ll do them and then I’ll say, oh, yeah, I never thought of doing the syntax that way. I’m like, yeah, OK, I don’t try to remember doing it that way. Yeah, I’ll take that. You know, and it’s kind of like when you use some of these tools, sometimes it almost seems like it will learn kind of how I my my traits of how I do it. You know, I guess it’s kind of getting kind of stiff. sophisticated in okay he like i do a lot of query you know and databases and i like my stuff formatted this way because i learned you know doing a lot of maintenance like that’s how i like to see queries written so i can understand what it does and i can spot bugs real easy and it seems like it’s kind of picking that up and formatting the way i like you know i think so i i do yeah I do kind of for a little bit, I, it didn’t work cause my token wasn’t work. It wasn’t right. I got, and I was like, I kind of, cause I started doing it for a week and I’m like, I don’t know if I’m going to do this. I said, no, just push on and see what happens. And that I did. And so I got it, like started working three or four weeks with it. And then the token, you know, messed up and I’m like, Hey, I’m not getting my suggestions anymore. And so then, you know, I went back and I fixed it. This cause I kind of missed it.

Speaker 0 | 21:34.177

So, so. I’m going to ask you to pontificate on the future a little bit here. But so, you know, using some of these tools, where do you think some of these AI tools are going? And where do you think it’s going to be useful in our work here doing IT support and IT work?

Speaker 1 | 21:50.858

I think it’s going to be for us. We’re looking at, you know, in my time here, it’s going to make a lot more things a little more efficient. I’ll make an example. When I first started here at Cassins, we had an eight accounts payable department and there were seven ladies in there entering in invoices and such. Now we have one and a half just because of EDI and stuff like that. And now I think with the AI, we’re going to be able to cut that, not the people down, but it’ll be because.

Speaker 0 | 22:27.554

You can grow the business with the same people, you know, like become more efficient. Yeah.

Speaker 1 | 22:32.137

Cause now, you know, then. We’ll cut the remaining paper out as much as we can. Because now we mostly only handle the things where we get somebody emails us a PDF for a bill or they still do mail it to us and then we scan it and we move it around. Now we’ll be using things like Doc UI and some of the other stuff like on Amazon also to, okay, I know I’m getting this PDF emailed to me and I can lift the invoice number and I can lift. this part out so somebody doesn’t have to enter that and get the rest of like i said but doing edi type things we move from seven people to one and a half and now we’re going to move down to one ish and move it around and such so where that one person that we won’t we won’t get some of her time will be split or his time whoever is there to another department to help something else out that you know we really you know be able to use our people even more efficiently, you know, and looking at Trent, you know, instead of just banging out, getting the stuff paid. Now we can, we’ll be able to use those people, look at trends and see what we’re doing right. And what we’re doing wrong with purchasing and stuff like that. That’s where it’s going to, you know, and some of that may, I can see AI even stepping in there where it might be able to help you that said, Hey, you bought this widget last year for 50 cents. And now it’s 75 cents where it might are two, you know. three years ago, you know, some kind of time where you, it’ll say, where, you know, prices creep up on you. You don’t even know it. And they might say, you know, you know, this is costing you 50% more and you might, the AI might actually go out and, you know, search the web for you and say, Hey, you know, you’re paying 75 cents, but this, you can get this for 60 cents from these people. You know, I could see it doing that kind of stuff where it kind of analyzes what you’re buying and then goes out and. gives you suggestions of, hey, you know, you might want to, you know, bid this out, you know, type of thing.

Speaker 0 | 24:36.108

Wow. It can assume too much, right? You know, you bought, be like, oh, I bought, I bought this roll of duct tape and I brought this other thing. And then it’s like, it’s like, it looks like, I’m sorry, Paul, it looks like you’re planning a kidnapping.

Speaker 1 | 24:50.714

That was scary too. You know, it’s really amazing. You know, that kind of stuff is kind of scary. You see it, I’m sure, in your Google Maps and stuff like that, that, you know, I take a walk at lunch and I happen to walk by. you know taco bell or something like that and it’s it’s assuming that i went in there you know and how do you rate it or you get a uh google survey you know sent to you saying you know did you go to this place today did you buy it with the credit card and all this different stuff just because their ai is put together that you went past a taco bell there’s

Speaker 0 | 25:31.449

always it’s funny so i was having a conversation with my older son yesterday about uh kind of a well it was ai and cars so self-driving cars he was he was um he’s he’s 10 and he was uh let me think about how he framed this i if i recall correctly his portion of the conversation was something like you know wow it’s like test like he was talking about tesla and he’s like i don’t think we’d ever want to drive one of those it’s really dangerous you know nobody driving the car and i’m like well how the there’s And of course, these statistics are there’s there’s some squishiness here in these statistics. And I’m not a Tesla fanboy by any measure. But there’s a reasonable case that computer logic is probably better than humans at driving vehicles. You know, like there’s probably, you know, the claims that there are less accidents if the computers were controlling the cars is probably accurate. The the biggest challenge, though, comes with to me. I see two big problems there. One. And that’s what I explained to him. I said, there’s two really big kind of like ethical or moral problems that you face. One is, how do you get the computer driving the car to be trained enough to know what’s right and what’s wrong? Because, you know, we’re people. So it’s assumed that we’re fallible and we make mistakes. And if I get into a car accident because I was talking to my son and not paying attention to the road for a moment and I rear-ended a car, you know, that’s… I’m obviously at fault, but that’s considered part of human failing. On the other hand, if a Waymo vehicle smashes into the back of a vehicle that’s computer-operated, then suddenly it’s a million-dollar lawsuit. Because there’s this assumption that the computer should have been infallible, and that you let something onto the road that was not. But at the end of the day, how does these things have to learn? We allow people to make mistakes that don’t result in, I’m not talking about crash results of somebody dying or something like that, but I’m talking about like, let’s say the car bumps into somebody at 20 miles an hour in a red light and smashes up the back of the vehicle, but otherwise doesn’t really injure anybody there. Now, if that car is like a Waymo car, they’re probably on the hook for a million bucks, or at least a lawsuit for a million bucks, because somebody is like, oh, I know, that’s deep pockets and you let this technology out. So I’ve explained it to him that. The technology, the challenge with this technology that we have is how do we get it knowledgeable enough? And at what point do we determine that it needs to learn? If that’s, you know, like these, these large language files and stuff like that, they have to absorb data to learn new stuff. So you have to let it do something. And you also have to let it make manageable mistakes in order to understand what use cases or edge cases could happen, you know, and There’s, I think, another example that I told him about, which could be problematic, is the choice conundrum. So if I’m driving down the street and I’m going to rear-end a car or I’m going to go head-on to another car, or I can swerve off the road but I might hit a pedestrian, what do I choose? You’re going to hit somebody. Now, you and I would choose hitting the car. But how does a computer logic, is it going to know fast enough and make a determination that it’s… that it like in a probability sense it’s wiser to hit the car than it is to mow down a pedestrian who is almost certainly going to get killed um you know like those are those are real ethical quandaries that we face with technology today and how you roll it out yeah

Speaker 1 | 29:11.307

that’s that is really going to be something because you know when they get to that point you know they i know they’ve had you know cars that you know are self-driving and doing that but from what everything i’ve read is They’re having problems with the AI interpreting what, like you said, that pedestrian, what they mean when they’re standing along the road, like they’re trying to wave you down or that’s some of the problems they’re having. And yeah, like, you know, your example is spot on. You know, it’s like, do I do I go ahead and, you know, have the head on accident? Do I slam on the brakes and just hope that by doing something that might be instilled, that that allows that other guy to do something?

Speaker 0 | 29:53.260

or like i said art i just totally try to avoid it and you know and and and hit the pedestrian you know that’s in the in the car the programming is probably going to have a bias for action you know so like you could actually that’s a you brought up a great thing that just thought made me think about this you said do nothing so that is an option so doing nothing is actually an option and you and i may choose to do nothing because the alternatives are worse but a computer program is unlikely to choose that because it won’t have it’ll have a bias towards action typically it’ll it’ll believe that it needs to pick something and it may not identify that doing nothing is also picking something and i i’m you

Speaker 1 | 30:38.246

brought out that do nothing because when i was in college i was going down a country road and and i had a friend with me and a car coming the other way was pulled out and past that car and was flying and we’d coming over the hill and it was like boom what am i going to do it’s coming straight at us and my and what i decided stay in my lane don’t do nothing and that car that was you coming my going straight at me it whipped itself into the the ditch and went around us oh wow so nothing happened so i had a split that i remember thinking what should i do should i try to you know should i go into the ditch then i thought he could be doing it just go go straight be predictable you know and that’s what that person did and whipped around us and my friend was all upset and was like you did you weren’t even upset i’m like i didn’t have time to be upset you

Speaker 0 | 31:35.674

I’m more upset than you thought, man.

Speaker 1 | 31:36.994

Don’t check my underwear.

Speaker 0 | 31:41.116

That’s right. I’ve mentioned when it comes to an animal running in front of you, like a woodchuck runs across the road, whatever. A lot of people are like, oh my god, don’t hit the woodchuck, and they swerve out of the way. I’m like, no. I will hit the brakes as long as it’s… reasonably safe and it won’t cause another problem. I’m not going to swerve off the road and go potentially smash into a tree to avoid a woodchuck scurrying across the road. So I’m sure I’ll get hate mail from some animal rights activists.

Speaker 1 | 32:13.693

Well, that was an example I was thinking about when you were talking about that was an animal. You know, it’s like, you know, they’re going to have to program that in. If you’re going down the road and you could, you know, you’re looking forward like you’re supposed to, and you’re checking your rear view mirrors and everything else. and you see something run out in front of you and there’s nothing out that you’ve checked everything i might i’m gonna swerve and miss it if i and if i see something coming i think there’s any chance that me trying to avoid it’s gonna make me lose control of the car i’m gonna hit it you know that’s right yeah and that kind of thing but they’re gonna have to get these things to do yeah

Speaker 0 | 32:49.028

yeah sometimes it has to make the choice to be like so you’re going back to the self-driving car the programming the ethical quandaries of this You and I could make this rationalization, and it’s just accepted that as people, we have the self-determination to make that rationalization, and people accept that it was our right to make. But if you tell a car to do that, who ultimately gets to decide? And I’m not saying that we shouldn’t decide, but there’s a real question about who is the decider, and that’s probably part of what we have to figure out in order to make this stuff work. Who gets to not be held? accountable um from and the accountable is maybe isn’t even the right word but just to use the example if i hit a car i’m unlikely to be sued if nobody gets hurt it’s just an insurance claim everybody’s cars get fixed we go on our own way it’s frustrating if a waymo vehicle smashes in the back of another car they’re likely to get sued for a lot of money for for for not um programming that car correctly so at some point we have to make a determination as a as a like a society if you will that like that that It is acceptable for these vehicles to make an error that would be similar to the error that a human could make. And if we’re going to put them in human conditions, we have to expect that there will sometimes be human-like failures that come from doing it.

Speaker 1 | 34:17.254

And then how are these things going to be regulated? Because, you know, say these models are self-training. So you set the model up with accepted parameters of these decisions, and it somehow retrains itself and it starts doing DBH from the way you had it trained. How is that going to be regulated? Because we have tons of lawyers in the United States and a lot of everything’s codified. So how do you, if this AI is, is there going to be that you have to have these AIs be trained? tested you know is it still conforming to what it was because that you know these things are certified and put out there and then come back as it don’t have to be checked later that it’s still It’s kind of like your brakes in your car. You’ve got to check every once in a while to make sure that they’re still good. Most of us, a lot of us, when you start hearing the scraping and carrying on, but what about the AI? Is that going to be checked to say if it’s still working within the confines of what it should be?

Speaker 0 | 35:23.384

It’s only when the AI goes, I’m sorry, Paul, I can’t do that right now. And you’re like, uh-oh, I think I’ve got a real problem here.

Speaker 1 | 35:35.492

because you think about it a little more still is like every three or five years or whatever like you gotta go take a driver test again yeah yeah yeah absolutely how do you recertify this stuff recertify so if you have a car and you drive it for 10 years at the pathway point are you gonna have to or they don’t have it that you have to take it in and say that uh somewhere and say that they can check and say yeah your ai is up to specs otherwise it’s like no your your ai is not up to specs you can’t we’re not going to let you drive that. I could not say drive that vehicle, but sit in that vehicle while it drives you. Yeah.

Speaker 0 | 36:10.687

So I want to take us back to another topic. This is more of the fun topics. Not that we didn’t have just a bunch of fun talking all about AI, self-driving vehicles and the ethics around them. So good thing we solved that problem for everybody. If they just listen to the podcast, they got cliff notes on what to do.

Speaker 1 | 36:28.535

Exactly.

Speaker 0 | 36:29.615

Um, I wanted to ask you about, oh, yes, your history in computers. So, like, this is one of the things that we ask guests a lot, to just get a feel for the era of computing that you came of age in and where you were. So, what was your first computer or your first, like, big technology experience?

Speaker 1 | 36:52.574

My first computer that I got to use, my brothers back in the 70s or early 80s bought a… trs 80 oh it’s 80 nice we used to call it trash 80 you know yeah oh yeah they were actually pretty good machines except their floppies would deteriorate really fast and it ran bait you know like it was a you know ran basic and such like that so that was my first and my my brother has always liked uh my older brother that bought it always liked uh stocks and and stuff like that so he bought it to crunch numbers for buying a stock

Speaker 0 | 37:29.612

and stuff like that so you might so that in the early 80s he was using this to crunch stock numbers yeah yeah nice that’s pretty cool

Speaker 1 | 37:38.974

I would he would go to me you know when I was in high school you know they had the Apple twos and you had the little programs that you had right that make it you know make a little box and in squirrel back and forth with the box and stuff like that so I knew a little bit about Apple basic so then he wanted me to work on his machine and you know learn that basic and i and doing things like you know you know can you draw make it draw this this chart and stuff like that of a stock okay okay you know and sat around trying to get it and banging my head against the wall and stuff like that and then one night i was watching doctor who it used to be on tv on on pbs at 10 o’clock on a sunday night and i’m sitting watching doctor who it was just mindless drivel i hope i don’t offend anybody you could I would just watch it. And all of a sudden I’m just watching it. I’m like, that’s how you do it. And about how to start to stock. And I ran back and got a piece of paper. Cause I didn’t, the computer was at his house and just write it all down the code. And then I was excited. The next day I went over there and coded it all up. And sure enough, I had a nice, you know, a nice chart with a line, you know, representing the stock close price on there and going, ah, yeah. Okay. Yeah, I can do this. And,

Speaker 0 | 38:55.328

Well, thanks to PBS. Well, thanks to PBS. Your public dollars at work here.

Speaker 1 | 39:00.290

Public dollars. It was one of those things where you’d watch the show and your mind would just release going, you know. I’ve always found that sometimes you just got to let your mind release and the answer will come to you. In the shower.

Speaker 0 | 39:14.576

There’s a shower.

Speaker 1 | 39:18.358

My people here at work always laugh. And if I come in, I say, You know, I was in the shower this morning. I was thinking, oh, no, what has he come up with now? And that inspiration type moment like that. So, yeah, that was like my first using computers and stuff. And then when I was in college, I have a master’s degree, joint master’s of science in economics and mathematics. Oh. And I was doing an independent study on doing some box Jenkins analysis of stock prices again. You know, there’s a theme here. And, you know, back then you couldn’t use spreadsheets or anything like that because the decimal places didn’t go out far enough. So I was. write in fortran and do it i had to pick up how to do fortran and it was doing it that way and you know that was kind of neat so i got a little experience i ran all my information stuff like that and you know kind of got into it you know i i you know i was in economics and mathematics so i wasn’t like a programmer or anything and then i got a job and i was in the government and this is like before the first gulf war i was gonna get uh rift which is reduction in force and there’s we were looking through the job openings and stuff and back in the early 90s if the job market wasn’t like it is now you know it was like it was sparse you know and i saw this ad asking for somebody which knew informix 4gl was the language back then and i grabbed a book and i looked at it and i opened it up and i read through the book i go yeah i can do that so i make it so i make it here so i i called you know responded to it that you But, you know, that and had the interview, got the job, read the book three times and started and started programming. And kind of that was the change in my career, you know, back then was, yeah, I can do that. And went down that line. I really liked it because, I don’t know, it’s a lot of programming and some of the things I do is really math based. So, you know, I had a really good structure for it, I feel. And, you know, I’ve always kind of liked that because. For the most part, you know, math is black and white. You know, a lot of problems that we deal with in programming, there’s an answer to. It’s just if you can find it. There’s some that you’ll never find. But it’s just, you know, that’s it. I just kind of fell into it. You know, and for my master’s degree, like I said, I have two master’s degrees. For my mathematics, it’s STAT OR. And for my master’s thesis, I had to write, again, some Fortran programs to do these calculations of these.

Speaker 0 | 41:57.612

average run length distributions and just kind of liked it you know and and just kind of fell into it you know yeah that’s that’s awesome that’s a great story like in the 80s you were writing stock chart programs like that’s yeah that’s pretty that’s pretty fancy right there that’s a great story so we’re coming up in the end of the of our episode here and before we go uh i want to i want i want to ask you for our listeners uh especially the ones you you’ve You’ve been in even a long career here in one spot, which is really awesome and unique in its own ways. What advice would you share for somebody who is coming into the business today in IT and IT leadership? What would you give them as advice? So I put you on the spot for a second here, right?

Speaker 1 | 42:45.111

Patience and adaptability. You know, a lot of the things we did back 30 years ago, I wouldn’t do it today. something that we still use and it’s still working. But, man, it’s going to be in the future here. You’re going to have to understand that a lot of projects will really never really be done. With the use of toolkits like Angular and things like that, where you create a progressive web app, the more sophisticated it is, it takes a little longer to… it takes you a year to create this thing or a year and a half. You get that thing and you get it deployed, you’ve got to turn around, turn your team around and have them upgrade it because over that time now you’re two versions of Angular behind. And if you don’t have your team constantly touching that code and such, and if you don’t touch it for a year or a year and a half, it gets so far behind and then they don’t remember to the crucial things that it takes to upgrade it. So it’s like, it’s really, it’s, you know, the time of having a, a green screen app, a program that, yeah, I wrote that 10 years ago and you don’t touch it ever since. I mean, you’re going to more and more, you got to deploy these apps, have your teams constantly in that app. Just so when angular says, you know, this. method or something like that’s going away, they understand what’s going on so they can replace it and keep the thing running. Because otherwise you’re going to have to, you know, what’s happening then is you’ve got to try to, and then you have to have all these Docker boxes and everything else that you can run and keep it at an operating system that’s, you know, two or three years old, running older versions, just so you can keep running it. Otherwise, if you want to keep running on the newest cell phones and everything else, as soon as you deploy something, you’ve got to keep… get in there and just keep it on a development cycle and and keep on there and because like i said the times of these i wrote this program 20 years ago and they use it every day it’s gone

Speaker 0 | 44:57.848

If you’re not picking up your code from 1992. Yeah. All right. Paul, thank you so much for investing your time with us on the podcast today.

Speaker 1 | 45:08.154

Yeah, no problem. Great to talk to you.

Speaker 0 | 45:10.916

That’s a wrap on today’s episode of Dissecting Popular IT Nerds. I’m Doug Kameen, and we look forward to coming to you on our next episode.