Latest / You've Been Heard / 373- AI Saves Five Million w/ Benny Zhang

Transcript

373- AI Saves Five Million w/ Benny Zhang · You've Been Heard

• 0:10
  Speaker A: Everyone in the world, I would like to introduce you to Benny Zhang. So welcome to. To Dissecting popular IT Nerds. Benny, we had such a really fun conversation last time. I'm really hoping that I can do it justice this time. But why would you start with give? Give us your story, man. How'd you get into it? How'd you get started? What was your first computer? How are you changing the world or making a difference in the world right now?
• 0:38
  Speaker B: So my first computer was a Commodore 64. So my dad was a visiting scholar for Cornell University. When he came back to China, he actually bought me a Commodore 64. So that was the computer I played video games and also learned how to code.
• 0:59
  Speaker A: Okay, so let me get this straight. Dad came back from the United States with a Commodore 64 box as a gift?
• 1:07
  Speaker B: Correct. Yes.
• 1:09
  Speaker A: That has got to be one of the most pivotal, pivotal, most memorable moments in your life. I mean, that's like, probably one of the coolest things a dad could ever do.
• 1:21
  Speaker B: Actually, it was pretty complicated at the time because it took him a while to figure out how to put it together. Then took me a while to learn how to use that properly. I think just about the same time I also read a book called Silicon Valley Fever. So it talks about the early days of intel, hp, how they got started, and eventually a long story about how Apple got started. Started. So. So both of those events are my, or were my catalyst for me to. To get into it.
• 1:59
  Speaker A: Okay, so Silicon Valley Fever. How old were you when you got that?
• 2:04
  Speaker B: I can't remember my age. Like that.
• 2:06
  Speaker A: You know how they say, like, don't, like, you should never compare yourself to other people? That's a lie. We all compare ourselves to other people.
• 2:12
  Speaker B: I want.
• 2:13
  Speaker A: I want to know it, because I was. Yeah, I was wasted. Waste. If I knew what I do now, I wouldn't have wasted so many years. And I don't know, but that's. That's how life goes. So Silicon Valley Fever. What's the. I. I haven't read it. I need to read it. What's the general. What's the general. I guess thesis and rundown of the book.
• 2:34
  Speaker B: So basically it talks about how the Silicon Valley entire IT industry started right in, like, through the Stanford University, the Lend Lease program, and also Stanford's commitment to entrepreneurship. So that's the theme of that, right? Then there are a lot of stories about, like, John Shockley. She's the inventor of the transistor. I mean, back then, the transistor is the size of like a football or something. So. So it's pretty amazing. Now we can pack like what, 3, 4 billion into a small chip no bigger than a small waffle, right? So, yeah, so that's truly amazing.
• 3:16
  Speaker A: And that led you to say, I have to be in it?
• 3:21
  Speaker B: I think that's the.
• 3:22
  Speaker A: IT didn't even exist back then, though. Yeah, go ahead.
• 3:25
  Speaker B: Yes, that didn't exist back then, but that was my inspiration to get into technology then. What you guys typically don't know is in China, when you, when you graduated from high school, like we as children don't have a say like what major we want to study or what career paths want to take. Our parents typically dictate to say, okay, you got to study this, right? Because your score is good. And also it's a good profession. You got a good job, right? So the topic of passion never actually came up. It's all about, hey, you can have a good job, then probably lead to a decent life. That was it, right? So I actually been pushed or guided by my parents to study chemical engineering, but. But my passion always lied with it, right? So then I studied in, in, in China, then eventually in Canada for many years. Then I went to work as a chemical engineer. But I was always the IT contact for the company because we have multiple sites. So I'm one of the IT contact in one of the sites. Then one, I think during one month period of time, all of the company's IT people back then with like eight of them, they all quit. So the company was left with nobody, right? So I raised my hand, I said, hey, maybe I can do the job then. I was actually still very grateful. They actually let me try it.
• 5:10
  Speaker A: It's that that story is so common. It's. I was working at the cafeteria, but I knew a lot about computers and they would have me fix the printers every now and then. And now I'm the cto. That's the extreme. That's like the extreme answer, but it's right. Okay, so that's awesome.
• 5:30
  Speaker C: Now at dissecting popular IT nerds, we expect to win, and we expect our IT directors to win. And one of those areas where we know that we can help you win is Internet service providers. As an IT director tasked with managing Internet connectivity, few vendor relationships can prove more painfully frustrating than the one with your Internet service provider. The array of challenges seems never ending. Dealing with ISPs can try one's patience even on the best of days. So whether you are managing one location or a hundred locations, our back office support team and vendor partners are the best in the industry. And the best part about this is none of this will ever cost you a dime due to the partnership and the sponsors that we have. Behind the scenes of dissecting popular IT nerds, let us show you how we can manage away the mediocrity and hit it out of the park. We start by mapping all of the available fiber routes. And we use our 1.2 billion in combined customer buying power in massive economy of scale to map all of your locations, to overcome construction fees, to use industry historical data. And we leverage aggregators and wholesale relationship to ensure you get the best possible pricing available in the marketplace. And on top of all of this, you get proactive network monitoring and proactive alerts so that you're not left calling 1-800-Gold-POUND SAND to enter in a ticket number and wonder, why is my Internet connection down? In short, we are the partner that you have always wanted who understands your needs, your frustrations, and knows what you need without you having to ask. This all starts with a value discovery call where we find out what you have, why you have it, and what's on your roadmap. All you need to do is email Internet popularit.net and say, I want help managing all of my Internet garbage. Please make my life easier and we'll get right on it for you. Have a wonderful day.
• 7:31
  Speaker A: The China thing fascinates me. It's actually going to be very useful for you. Now, I'm assuming you speak. What do we speak Mandarin? What? What's. Yeah, I speak Mandarin. English.
• 7:40
  Speaker B: Right.
• 7:41
  Speaker A: So that's has to be like really, really helpful if, I don't know if you've ever worked in a company where you had manufacturing in mainland China or Hong Kong or something, but it's got to be very, very useful. I'm just guessing is that. Does it come up ever?
• 7:55
  Speaker B: Not actually, not really because I spend most of my time in Canada and the U.S. so that actually never came up. Maybe I should have explored and just.
• 8:04
  Speaker A: I run into like, you know, we've had global, we've had global shows before where you know, the CTOs like I gotta travel to Hong Kong and China and you know, it's like the different language barriers is, you know, like always comes up and time Z is another thing. But okay, really cool. So that's awesome. The back then it was a cost center. It's. It's arguable that it is still a cost center, but I think with the changing of the guards AI which there's this massive frenzy which I would, after talking with a lot of people, I would say it's most definitely a bubble with tons of money being invested into it and not enough money coming out of it yet. But I would say we're definitely moving away from the cost center piece. What's your experience with it as a business enabler as enabling?
• 9:00
  Speaker B: I give you a story, right? So unfortunately, even today a lot of cities fos that's typically it reports to you, right. Still consider it as a cost center, right. So their goal, when you have a dealing with a cost center, then like our goal would be, hey, how can I cut the cost? Right? So I tell you a story. So I was hired for this new company. I won't name name obviously then, then they were introducing like me to, to each department, then our accounting department actually telling me, hey, they are doing this major analysis project, right? So taking like, I don't know, 10,000 wells, then analyze the data from those wells, right? And the data is somewhat structured and somewhat unstructured. So it's overwhelming amount of data. They budget I think was $5 million. Then they plan to finish that in about a year and a half, right? So they hired a pretty sizable consulting company to do the work. So I talked to them. Then I also looked at the data. I thought, okay, I think I have a way to do this better, right? So I called up a developer I knew before. Then we get together with the business and study their data for a little bit. So they're doing this whole manual work, right? So they have a team accountants, then a team of analysts, plus the consultants. So they're doing them like manually, like, well by well, then open the Excel sheets and CSV files, then pulling data, right? So after we studied the data, we it side and proposed a combination of a little bit of AI, a little bit of like a program and logic. We did it with $40,000 and within three months the project was done. So just imagine the amount of money we save, amount of time he saved for the company. So they were just like shocked how this even remotely possible, right?
• 11:26
  Speaker A: This is a great, a really great AI use case. Now to get to the heart of the matter, what it's not that you probably didn't have the ability to do that, that a lot of people listening to the show right now would not have the ability to do that even if they had to tinker around and figure stuff out. Most of the people are very curious about how things work. And I would say that most of the people could figure this out that listen to this show however, what they might not have is that bridge of communication to how you would even know that that is going on and to begin with. So maybe dig in a little bit more as to what even got you in to the circle of trust, so to speak, so that you would even be allowed to play around with that, let alone have the time. Because a lot of people are putting out fires all the time because the majority of these, I don't know, mid market IT leaders, manufacturing IT leaders are kind of overwhelmed, in the overwhelmed zone of putting out fires all the time. So how did you have time to do that to begin with? How did you have the. How do you build those connections to even get in there to even be able to show your stuff?
• 12:39
  Speaker B: So I think that there are a couple of things to remember here, right? So I remember my old days when I first studied my Microsoft, the certified professional engineer. Then the very first class, I didn't know anything about Microsoft products at then. Then the instructor came in and he said, asked the first question to say who here thinks you're going to work with computers? So obviously all of us raise our hand, right? Was that okay?
• 13:11
  Speaker A: Of course.
• 13:12
  Speaker B: I'm an introvert. I want to work with computers so I don't have to deal with. So the instructor said, you guys are wrong. Behind each computer is a human being. So that sentence stops, struck me and also stuck with me for a long time, right? For a very long time. Then what I get to understand is it's not just like a black box. You have to go across the aisle and shake hands with other people. You have to interact with other people and listen to their complaints and sometimes criticism and understand their pain, then gain their trust. Otherwise you're not going to be able to get those projects. They won't even tell you. Or if they have idea, they're just going to go out do their own things. I think that's a very fundamental thing in the IT professionals is to understand behind each computer is a human being.
• 14:23
  Speaker A: And I think also we've gotten to the point in history where most people will agree with that whether they do it or not is. Or do it well is a whole nother question. I think there's this. There's a lot of what I've noticed after talking with so many people for years now, IT leaders is that there's the general complaint or the general kind of grumbling is that it isn't heard enough or like we're there. We have a seat at the executive roundtable. We know our KPIs but there's a, there's this fine line that can be drawn between are we, are we really a valuable seat at this table or are we just liked or we just a necessary entity. In other words, we need to, we need to bring all the ERP systems together or we need to upgrade this and when everything's broken, we love it and we need to. Can you make sure we don't get hacked, please? And everyone knows that that's like, that's probably one of the number one things that keeps IT guys up at, at night. Because it's not a matter of if, it's a matter of when and how bad it's going to be. And I, I, I think that IT leadership often might not feel respected or enough and get enough respect. And I think on the flip side, on the CEO side or the CFO side, I think that they're saying it is not hardened enough or they're not, they don't really understand maybe the business drivers, they're kind of like a secondary support piece to the business drivers. And I think there's a level that it might actually have to take on even though they have too much work and might be stressed out and putting out fires enough. I don't even know if it's at the level of the stress that maybe a c, that a CEO is even handling because I think they have to drive dollars and cents and I think we could emerge as a new, a new CTO could emerge that's willing to put a number on their head. That's like I will drive revenue, not, not just understand it and not be a cost center and be a support structure, but I'll actually take it to the net, not just roi where I'm making sure that, you know, we're, we're as efficient as possible and there's a return on investment and all that, but I'm actually willing to put a number on my head to help drive revenue. I don't know if I'm asking too much, but I think that that's where the gap is, is maybe they don't really understand the stress that the CEO is under. They're more complaining about the fact that the CEO is not giving them the respect that they deserve or it's a mix between the two and there has to be this way of understanding. Do, does that make any sense to you?
• 17:21
  Speaker B: It does. Right? So I would argue respect has to be earned. Then nobody just come out deserve respect. Right? So if we, you think about it, right? You're asking other people to invest in you. Right. So in the executive table. That's what happens, right? Where we as IT leaders would ask the business to say, hey, put the money in me then why would they do that without seeing values out of the, out of our department? They wouldn't. I mean, nobody would. I mean you wouldn't. I wouldn't put a money on the table to say, hey, go do the thing then without even seeing the value. Right. So it is truly upon us. Not. I wouldn't say, not just IT leader, the entire IT team to demonstrate our values consistently to earn our respect before they can give us our respect.
• 18:22
  Speaker A: Let's. How can we do that? Very almost specific. Can we be even specific? Is there something that we can give out to the audience that is next level, next level beyond roi or do you have any suggestions? Or it's just how you do it or how have you. I mean you had this amazing case study with it, with the AI and IT really. How much was the consulting gonna cost? And you did it for 40,000. How much was the consulting gonna cost?
• 18:51
  Speaker B: I think the total budget roughly about 5 million. So the vast majority already goes to the consultants. And that consultant actually is pretty funny because he talks so slow like, like every, because every hour is. Or any minutes for him. It's a lot. Yeah, it's like so super slow. And it does things super slow.
• 19:15
  Speaker A: I hate that I'm so. I need things fast. I really am a fast gambling even when I'm coaching, coaching my kids. Does it need to take that long to empty the trash? But I should have the kids on someday. We should dissect me the. Okay, so he talks really slow. He still got paid the 5 million.
• 19:35
  Speaker B: No, no, that's all.
• 19:37
  Speaker A: He can't pay.
• 19:39
  Speaker B: We fired him within a month or so because once the business see they can get it done so quickly, they immediately fire them. So they hate me. Obviously losing that much revenue to them is a huge deal.
• 19:53
  Speaker A: Did they get paid anything? Well, I got paid something. They probably got paid more than you got paid more than you.
• 20:01
  Speaker C: More than.
• 20:02
  Speaker A: Your entire project cost. They probably got paid for more than your entire project. Okay, so what was the like windfall after that? Like what, what was the kind of like talk after that? Were you like, hey, cfo, you know, hey John, what do you think, man? Pretty cool.
• 20:20
  Speaker B: So yeah, so that, that actually opens the door for. Because I was a brand new new manager in, in the company, right? So. So I was the head of it but that opens the door to. To a lot of new projects flowing into it. So they would ask us to do things. Right. So they ask us to do another thing is consolidate because we collect, because we're all a gas company. So we collect like tremendous amount of data. Can you imagine how much data there is? It's a gigabyte per second kind of data, right. So they asked us to collect those data and analyze and put it into a proper database so their analysts can come in and do their analysis. That was a major endeavor, right. So because before they were relying on contractors providing the information, which you can understand there's a major cost when you do that. So like I said, we have to earn the respect from the business before we can actually ask for, even ask for anything. Right. Start with something small, then, then, then once they see we can make an impact, then they'll be more willing to working to work with us.
• 21:44
  Speaker A: You had a really great analogy. It was probably like the, one of the best analogies of all time. Not that this analogy hasn't been used before, but I felt like it was the best implementation of said analogy.
• 22:00
  Speaker B: So.
• 22:01
  Speaker A: And I think it was the, the car, car mechanic analogy. And how, how can we use that?
• 22:10
  Speaker B: So I always argue, right. So the way I define it is we have two missions. First is to protect the company. That's very easily understand, right? So, so the second is to empower the business. That point means we IT need to help the business not only just gain a competitive edge. We want to be so far, so far ahead of our competitors and it's going to take them years to catch up. So. But how we're going to do that, we have to change our mindset because a lot of our IT people's mindset is still in this car mechanics mindset. Right? Oh, here's a car that breaks down. So I fix it, I fix it really quickly. Then customer's happy. In reality, the customers are not happy if their car breaks down very often. They're not going to be happy. So one of the things I ask that when I hire a service desk manager, I always ask them the question to say, hey, does having more service tickets means better or worse? So the answer is closing down. Yeah, you did a good job. But pretty soon somebody is going to want a different car. Right? Until we change that mindset then. And we cannot demonstrate the value to the business because keep fixing the car is not valuable to business. They want a car that's reliable and.
• 23:53
  Speaker A: Fast and a sports car or a no, no, I'm the one that can't be okay, One that can't be broken into or if it's stolen, can be tracked and traced. And I mean it goes on and on and on.
• 24:08
  Speaker B: So.
• 24:10
  Speaker A: It sounds simple. We've had this conversation before. It sounds really simple. But there is a fine difference there between that because yeah, I do believe that most people are dealing with antiquated silos fixing broken what we got. We know that. I would say 50% of the people listening to this call probably have an IBM what as is it as 400 sitting somewhere in the business running something. Not that that's a bad thing, that's probably a good thing for IBM. But it. How can we fine tune that? What would be the next step? How can we fine tune the IT department to take it to the next level? Should we, should we be survey surveying the sea levels? Should we be, you know, whether you're remote or in house. Should we be. What kind of questions should we be asking? What kind of steps should we be taking to further continuing along with the car mechanic analogy.
• 25:12
  Speaker B: So the way I always advocate is changing this whole IT like a black box into like a human being, right? So remember like I said, behind each computer is a human being and behind each IT computer is a human being as well, right. So for example, the practice we typically use is we don't use a service desk account to send out interactions with users, right? We send it out from our personal account. So you will see an email coming from me like say Benny says something instead of from service desk. The point is to changing this black box of service desktop to each individual. So that way we have individual connection with a lot of users. Right? All the users actually. Then for me at the leadership level, I have never shied away from any speaking events. Actually I volunteer to speak in many events. In our leadership training events, in our board meetings, in, you know, any kind of public speaking possibility. Put myself in front of the employees then talking to them. Then there's another way I think is probably the most effective way is doing regular lunch and learners. So it will organize an event with HRS help, right? Then we do that monthly for all the users. From my own experience, that's the most impactful things we can do. Not on the part we teach people how to use technology or we're sharing some information about security, stuff like that. Not about that. It's about introducing it to a lot of people and open people's eyes to see more opportunities. Then the users will come to us for new stuff. I give you one good example. We did this chatgpt how to use the Generative AI then to help our work. Right? Then I got a call from this VP to say, hey, we in our business we have to read more than 1000 page long specification document, but we only need probably 20 things out of those thousand pages. Is there a way you can help me to read this quickly? Because right now I have to spend two people, three days to read this through thousand pages and extract the information we need so we can put together a proposal for the customer. And this happens over and over again because there's a lot of activities then if we can help them save time. And we did it through an AI tool to extract the information they want, I mean that's a huge, huge thing for them.
• 28:30
  Speaker A: This was. They had to read like trade journals or what was it specifically?
• 28:34
  Speaker B: It's a recent specification, right? It tells you how to deep doc.
• 28:37
  Speaker A: Like okay, spec doc on, on like some tool or equipment or something or.
• 28:44
  Speaker B: New or typically on a project.
• 28:47
  Speaker A: Okay. You've given us two beautiful AI use cases on one podcast. This is great. But all of them led to human interaction which should give people a clue that AI is a human assist. It is not a replace of a human.
• 29:18
  Speaker B: You can euphanize as much as possible, but in reality it will displace a lot of jobs. Then there's a major saying like AI won't replace people, but people using AI will. So for us, I love it. We must learn the skills, otherwise would well be left behind. And the pace of changing actually is dramatically increased. In the old day you were talking Moore's law, that's about 18 to 12 months. 18 months to 12, four months. But now we're talking about six to seven months is a generation. So we must stay, I think using Steve Jobs word, stay hungry, stay foolish.
• 30:07
  Speaker A: I don't know where to go from here. There's so many questions I want to ask you, which is pick one, hold the Steve Job sting. I want us to give back something like, almost like a mini, like a mini infographic, something that we can build for the people. Because Evan's like, okay, great lunch and I'm going to go do it. Well, you've done a bunch of them, so why don't we ask you what is the step by step process to delivering a good lunch and learn and speaking to the humanoids that we've hidden in the server room forever to not talk to. Right, but what is the. What is the best way or what's like. I mean I think that how to use Gen AI or ChatGPT to do work better is a great one. That's the first one that people love that's going to lead down the rabbit hole of well, how do you make sure people don't leak company data and do all kinds of other stupid stuff using it? And I, I beg the question on that one. So why don't we just start with how to lead a lunch and learn on X topic. Let's just use how to use Gen AI to do your work better. I think it was a great one without breaking all security protocols and I don't know, something like that.
• 31:24
  Speaker B: Sure, sure, sure. So the, the centerpiece or the center like a theme should be. I think we all need to understand every time we do any kind of presentation, it's actually we're providing entertainment. So it's extremely important we make that entertaining for the audience. So I always suggest we make it like a very interactive. Right, right. With the users. Then you can ask questions or you can ask them to ask questions or you can have them participate with us. Right. So for like Gen AI, a typical tactic would say hey, what topic you wanted to ask the chatbot. Right. Or what do you want to do? Then have them participate. So remember, it needs to be entertaining. If it's not entertaining or it's just full of IT jargons, then nobody can understand. It's useless.
• 32:34
  Speaker A: It's the little things. The little things. It's the small little things that make a massive difference between success and unsuccessful. And I think we've probably talked numerous times about how to give public presentations and how not to be nervous and how to do all these things. Well, you've just summed it all up in one word. Entertaining.
• 32:55
  Speaker B: Yep.
• 32:56
  Speaker A: How do we be entertaining? Okay, so interacting with the Q A participation. What do you want to do? If you were to do an outline right now for this lunch and learn, how would you outline it? How long would it be? What would we start with? What would we eat? Would it be a round table? Is it a group of lined up of chairs with a presentation up at the front? Where do you do it? What's the best format?
• 33:18
  Speaker B: So it depends on the company. Right. So if you're small enough, obviously in person will be most ideal then. I'm allergic to pizza, sandwiches or hamburgers. Unless it's a lobster sandwich. Then I'm all for it. So yeah, you can organize that through HR or some other like an admin function. Who will handle the food? And surprisingly, once you have food on the table, more people will show up then some other companies. We have to do like during COVID period for example, I had to do those lunch and learns remotely, right? So we did quite a few lunch and learns remotely. And it's also a different trick you have to to apply to keep people entertained and also engaged, right? So when you do things remotely, one of the things we would ask is everybody has to have their camera then you have to ask a question. If you don't, we'll ask you a question. So to keep them engaged, right? Then I personally like this way is to say hey, I have a quiz at the end so make sure you pay attention, right? And sometime I would say yeah, this topic is going to be on the quiz. Make sure you remember that. So keep their attention up all the time. Then obviously I don't have a quiz for them. Sometime I do have a recap, but not truly a quiz per se. I think that's some of my tricks I use. So if it's our presentation always contains at least couple videos then a few interactive questions or interactive games we play with users real life examples. For example, if you talk about a generative AI, the best way is to tell people okay, here are few interesting use of of AI. So give you one example. So this lady, she's using ChatGPT. She told the ChatGPT to pretend to be her overbearing and always criticizing mother in law then so she can practice how to interact with her real mother in law with ChatGPT. So this is a really caring person.
• 35:59
  Speaker A: This is the really. This person must really care, okay. She must provide loves her husband.
• 36:06
  Speaker B: I can tell you that much innovative way of using ChatGPT right? So it's very entertaining as well when you tell that to people. So.
• 36:18
  Speaker A: That'S awesome. Did he give her some good feedback?
• 36:24
  Speaker B: I don't know. It seems like very helpful because if you tell ChatGPT to a meme it can be pretty ne.
• 36:32
  Speaker A: Okay, the wow, I've never thought of that. These are all great things. What else should we do with ChatGPT now I've just gone down the Chat GBT hole. What's your biggest secret to using AI right now? Coaching end users?
• 36:47
  Speaker B: Well, the huge thing I have to tell everybody is like AI does hallucinate. That's a huge problem. I mean I get to a point is I don't trust anything that tell me nowadays I have to confirm that with a couple of different AIs. No, they all pretty much sit similar or have the same difference. Then sometimes that can't even be trust trusted. Then you have to verify that. So I think That's a huge thing especially come to work. Right. You do not just produce whatever AI tells you and present that as the truth. Because very often there's hallucinations. Then I think that's a huge warning. I think you also mentioned the privacy thing or the data security is huge deal. People just dump whatever on chatgpt then without knowing and thinking about it. Okay. When I put this contract in ChatGPT, it would become a public document. Then for public traded company, sometimes they can be a huge problem. For example, you disclose a client information that's in violation or now disclosure agreement. Yeah. So these two things is something truly to be careful of.
• 38:11
  Speaker A: So what can we do?
• 38:16
  Speaker B: Education. Right. So we actually have this like I had this discussion multiple times with the C level effective. Right. So because you probably notice this as well, in some major companies, they just flat out block all the AI tools. Because I can do that in my security tool in like two clicks. Right. To block them all.
• 38:41
  Speaker A: Yep.
• 38:42
  Speaker B: So then I had this argument to say, okay, you got to think about the AI tool is no different from Google. It's just a new way. We do Google now. Now with Google, do you block Google then? Obviously, okay, nobody actually blocks Google. But why would you block a new tool then this is going to be better than Google. And frankly, actually I'm, I'm using like AI tools way more than Google nowadays. And surprising to me because I grew up with people, played with Google for many years.
• 39:18
  Speaker A: It's just kind of annoying now to Google is like, should I. It just seems like why would I go use Google unless I'm looking for like a restaurant's phone number.
• 39:25
  Speaker B: Correct? Correct. Surprising. You say that I may need to adjust my stock portfolio based on that comment. So that's the really hard conversation with the CEOs, right. And also especially the chief compliance officer because he's very or she very often concerned with data leak. Right. Then I said, the best way we can handle this is through education, educate our users and achieve two goals. One is we can have the data compliance issue resolved. Second, the more they know how to use AI, we will become more efficient. It's a huge benefit to our company. So I had this conversation with many CEOs now.
• 40:16
  Speaker A: Do you track, do you track what they're doing on AI, like their prompts and how they're using it?
• 40:24
  Speaker B: We don't honestly, because like when you get to a certain size, it becomes very difficult. Obviously we have Companies sponsored like ChatGPT accounts. Right. We have our team accounts. Then we also use COPILOT quite a bit and there's quite a few other AI tools we use.
• 40:45
  Speaker A: We have a, we, we had there. It's being sponsored by the podcast that's just. I'm going to plug it anyways. But we've got a Devs AI dashboard that plugs in really all of the major LLMs and a bunch that I don't even know. It's all in one dashboard. You can. So when you're, when you're building different, you know, bots or projects, you can say, okay, I want you to use Claude, I want you to use OpenAI, I want you to use Gemini on this one. Right? And I want you to use the Gemini now from an administrative standpoint, it can go in and see all the, all the inquiries and everything that all their end users have done from one dashboard. So I was just thinking off the top of my head, someone could probably write a. They could probably write a pretty good long prompt that incorporates your AI usage policy and it could probably flag things just for an education standpoint, when an end user is kind of pushing the line a little bit. And you could probably give a reminder. I'm sure there's probably something already out there or some security product has got to be building this somewhere. But I would think that could be something that could be built pretty quickly. Almost like a training or security kind of usage policy AI thing.
• 42:10
  Speaker B: That's a great idea. That's a great idea.
• 42:13
  Speaker A: This is what we do. I mean, this is what we do on this show. So we've got the dashboard that plugs it all in. I think it's $30 a user. If anyone's interested, I'll put it in the chat of the episode or something like that.
• 42:27
  Speaker B: I will take a look. I'll take a look.
• 42:29
  Speaker A: So back to. Back to this beautiful Edge uses. I had so many things that we wanted to talk about today. We went through kind of like the zero who applies zero trust to something like AI. It's gotta be difficult.
• 42:42
  Speaker B: I think so one is probably stump the engineer. One good thing about the generative AI, at least for the time being, it doesn't connect to our environment just yet. But it could be potentially like Copilot could potentially a major lead, right? Because typically we have more trust in Copilot than any other AI tools. Then if we are connecting our internal data with a external AI, then it become a huge, huge problem. Right? So yeah, because those major companies can be hacked, then there is a much bigger target that than all of us, right? Can we do like A concern can.
• 43:29
  Speaker A: We have like a data warehouse, like separate kind of off the grid with a host, our own host, our own, you know, open source version of, you know, some generative AI or some LLMs or something like that. It seems like a lot of work to do for, I mean it's going to make sense when it comes to like a single source of truth and wanting, I don't know, salespeople to be able to data dip into the company database or something like in that standpoint I think it would be useful. But is there anything that actually do.
• 44:00
  Speaker B: Right. So most of the companies, I'm aware actually all using the similar model like that they have data warehouses build and they also have specific AI data warehouses for AI training, all of that. So most companies actually do that. I think the concern, it still comes down to the leaking of the data because sometimes the actual data itself is more valuable. So for example, if a public traded company like their financial information got leaked when there are tons of money to be made on that information and they could become a major criminal investigation. If somebody got a hold of that and authorized people got a hold of.
• 44:46
  Speaker A: That, that's another deep long hole that we could go down.
• 44:50
  Speaker B: But I think that the way I foresee at least is we going to have those, those large language models in house and we're going to start training our own like AI. And that's actually a lot of companies already start doing that then instead rely on external AI system like ChatGPT. Right. Because they have the APIs you can pipe into. Instead of doing that, we already start using our own training our own AIs. A lot of companies do that and I think that's our way to counter the data leak for the rip for the once it's under our control it has a lot less like possibility to leak the data.
• 45:44
  Speaker A: Do you have a good process for doing that that you can share or not?
• 45:49
  Speaker B: I can't share that.
• 45:51
  Speaker A: I don't think so. Classified information, that's okay, but it's good to know you can do it. It's good to know.
• 45:58
  Speaker B: I'm not saying do it, I'm saying like a lot of companies have been doing that already. Some have been doing that for years at this point. So it's actually a trend. The way I foresee the AI is going to become AI everywhere, everywhere from each department is going to have their own AI agent or tool trained up. Then even each individual may have our own AI tools.
• 46:28
  Speaker A: I know we're running out of time. This has been a great conversation we didn't even get to coaching and developing IT talent, which is something that you're very, very. That's my Jew.
• 46:35
  Speaker B: Yeah, that's my, yeah, I mean I.
• 46:37
  Speaker A: Know you want to give us the one, you want to give us the quick rundown. We got five minutes, let's go. Let's do coaching and development IT talent in five minutes. I take the go.
• 46:45
  Speaker B: So the, the first thing I always talk about when I talk about coaching is what is your passion? Tell me your passion. Right? So, so, so there's one, one person I interviewed, right, Asked, hey, what's your passion? Then he said, data analysis. He's applying for a system analyst job, right? Then I said, okay, can you show me how that works? Then he pulled up this whole, like a dashboard he developed on his own to analyze all his fantasy football game, every player statistics and then all the, the fancy like charts he, he used to analyze those things. He, it immediately tells me it is his passion. He passionate about data. He's passionate about analyzing those data and then getting results. Right? So, so I, the first thing I always talk to people when I talk to people is okay, what is your passion? When you, when you know what your passion is, then you can quickly grow up and become probably one of the best people in the field.
• 48:01
  Speaker A: Okay. And then obviously sometimes we're passionate about things that don't fit into the job role correctly. So how do you get the right people in the right seat on the bus, so to speak? I mean, how do you know when you don't have the right person, how do you develop? You know what I mean? So we find out what they're passionate about. Sometimes it's hard to find a fit.
• 48:22
  Speaker B: Sometimes it is, yeah. So I personally always want people who are passionate about it, right, to join the team. So the second question I typically ask you in the interview process is are you an A player? Are you a top performer? Right? Then okay, everybody says they're a top performer, obviously. Then the follow up question is okay, what are the tangible metrics you have to substantiate, substantiate that claim, right? So then like vast majority people would fumble, right? So I said, okay, if you ask Lionel Messi, he's a major soccer player, what he's, he's a metrics, he will tell you he scored 1,000 thousand goals in his career.
• 49:13
  Speaker A: By now what his, he'd probably tell you what his 100 meter dash is. He'd probably tell you exactly what his, his weight was in January of this year. He would tell you all Kinds of different things. I love it. I love it because I tell my guys all the time, I say, if you're not keeping score, how do we even know who's winning the game?
• 49:34
  Speaker B: Correct. If you are not and if you.
• 49:37
  Speaker A: Don'T get on the seal of play. And then if you don't. What if you don't even get on the field of play? What if we're always practicing? You sure you don't want to be in sales? Should be in sales.
• 49:47
  Speaker B: I'm in for work.
• 49:48
  Speaker A: So our introverts doing lunch and learns. And I realized that I have to talk to people. But you know, there's a whole. There's a whole group of introverted salespeople that are like some of the most successful salespeople in the world. Most people think introverts can be in sales, but that's not true.
• 50:03
  Speaker B: Yeah. So here's the trick, right? So I always, always use this as my to my advantage. So when human interact, the one talks most is actually liked least. So the more you talk or the more we talk, the less we will be liked and the less we will be heard. So as introvert, we should use that to our advantage. Then we tend to listen more and people tend to like us when we listen more. I'm not talking about saying just sit down and listen, but it's more active listening, asking questions when is appropriate. Do let the people tell you their story before you interrupt them.
• 50:49
  Speaker A: You have given us so many gems. I am so thankful for you being on the show. I know you've got to go. I know you've got a meeting coming up here. I would love to ask you to even just do a lunch and learn for our private community if you're willing to do that. I think if you ever have time, if you want to do that, I'm offering that invitation to you.
• 51:10
  Speaker B: Sure.
• 51:11
  Speaker A: And I think it would be very, very, very useful. Very thankful and we'll continue to be in touch. Thank you so much for being on Dissecting Popular. Thank you, nerds. Anything, everyone. Have a great day.