UnHacked

Justin: 00:14

All right, guys. We're live. Welcome to episode 19 of unhacked. We are live on Facebook and YouTube. If you're listening to this later, man, it's already come and gone.

Justin: 00:27

It's, in technology, things move fast, so you better get on and and watch us live. Alright, guys. Let's do a quick introduction. I'm Justin Shelly with Phoenix IT Advisors. I'm going first, by the way.

Justin: 00:36

I always go last. We we support clients in the Dallas Fort Worth Metro and then also in Northern Nevada. Coming soon, Dallas Fort Worth Metro and then also in Northern Nevada, coming soon, maybe in the Idaho TBD. Anyways, this, this podcast has been something that I think we all kinda just love doing every week, but also it it keeps us sharp. So that's there's kind of a hidden benefit to it there.

Justin: 00:59

So, that's that's me and my story. Mario, why don't you tell us a little bit about who you are and how to say your name properly because I always screw it up.

Mario: 01:08

Mario Zaki with, Mastech IT. We are located in, North Jersey servicing, you know, the tri state area, Jersey, New York, all the boroughs. Tomorrow will be our 20 year anniversary.

Justin: 01:23

So Oh, congrats.

Mario: 01:25

Thank you. And I mean, we service out the Tri State area, but, you know, just like these guys, we can do nationwide if we need to.

Justin: 01:35

Alright. Alright, Mario. Congrats on the 20 years, man. That's that's pretty Yeah. That's something, man.

Justin: 01:41

That's something. Thank you. Brian. Brian, tell

Bryan: 01:44

us something. With, B4 Networks located in Niagara Falls area, Ontario, Canada. And we'd like to say that we help business owners remove all the frustrations that come with dealing with technology.

Justin: 01:59

Alright. And last but not least, Birender back for what is this? Your second appearance on the show, I believe?

Barinder: 02:06

Yeah. 2nd yeah. 2nd appearance. Last time, was fun. So I'm back.

Barinder: 02:11

I heard you guys talking about AI. I'm like, I can't miss that one. Yeah. So we're in the right of, Vancouver, Red Rhino Networks. We serve the the lower mainland of British Columbia here, but then, of course, our clients all across the country.

Barinder: 02:24

So, yeah, having a good time.

Justin: 02:27

I heard at least 2 of you that have just become competitors of mine, so I'm not sure how I feel about that. But, anyways, so guys, today we are talking about artificial intelligence, the benefits and risks of using AI in your business. Before I go any further, I would like a show of hands of everybody and anybody that used AI to prepare for this episode. No, but okay. We've got Mario.

Justin: 02:54

Nope. Brian Brendler, you didn't use AI in prepping for today?

Barinder: 02:58

Not to prep for today. No.

Justin: 02:59

Hypocrites. Oh my god. I did, by the way. I did. So let's let's talk.

Justin: 03:06

I wanna open up today with the definition of artificial intelligence. And, you know, I I said that as we were getting started, was it you, Mario? I was like, everybody knows what AI means. Like, yeah. Well, sure.

Justin: 03:18

Everybody knows what the cloud is too, but define it.

Bryan: 03:21

Right.

Justin: 03:22

Yeah. Put a put a lay user, an end user in in front of a camera and a microphone and say, hey. What's artificial intelligence? That'd be, I would be in in fact, put, tech geniuses in front of an a camera and microphone like we're doing right now and go. What is artificial intelligence?

Justin: 03:39

Brian, on the spot, what do you got?

Bryan: 03:41

Well, I'll tell you what artificial intelligence is not and what most people think it is is automation k. And it's not. So just because something is automated doesn't mean it has artificial intelligence. My view on artificial intelligence is something that is capable of making decisions on its own and be able to learn and grow and, inhale new information and be able to process that and and and embed it into its algorithms. So automation, by definition, can make decisions, but it's static.

Bryan: 04:10

It doesn't change without human input versus AI that could make changes based on new information and, change its programming essentially on the fly to allow, that that change to happen.

Justin: 04:23

Alright. Brenda, you got any thoughts on this?

Barinder: 04:26

Yeah. Similar to what, Brian is saying. It's definitely not automation. What AI is, artificial intelligence that enables computers or machines to simulate human thinking and problem solving. And I I think there's 3 elements that make AI AI.

Barinder: 04:41

One is some ability for generalized learning. It has an ability to continually learn. It has an ability to reason, mimicking human thinking, and also an ability to problem solve. And I think those three elements really differentiated from other kinds of technology. Right.

Justin: 04:57

I hate to break it to you guys, but you're all wrong. Have you seen the movie I, Robot? Have you seen I, Robot, Will Smith back in the day? When was this?

Bryan: 05:06

Yes.

Justin: 05:06

I don't know. At least 10, 15, 20 years ago. I don't know. I'm an old man.

Bryan: 05:09

Yeah. He's still sitting at the bottom of the ocean right now.

Justin: 05:13

Is that not the real definition of AI? Is I, Robot, the movie? So if you haven't seen that movie, go watch it, and you'll know what AI is. Yeah. It's the the misunderstood, unseen, computerized force that will eventually take over the world and annihilate all human existence.

Justin: 05:32

That is what we're talking about today.

Bryan: 05:33

Thanks, Justin. Alright. I'm gonna hang up now.

Justin: 05:40

Alright. Your your definitions were technically more accurate than mine, but I like mine better. So, today, what we're gonna do is we're gonna talk about what it is, what it isn't. We've already got a a head start on that. What do machines do better, and what do humans do better?

Justin: 05:54

That one I'm really interested in in breaking down with you guys. And then, of course, this is a podcast about cybersecurity. So we do have to at least touch on how does AI impact cybersecurity, both the good guys and the bad guys. I wanna I wanna hit both of those. But we're really gonna I I would like to and and this is why I say this might Brian, I told you earlier, this might be a little bit longer than normal or Brenda I was talking to, because while we're gonna talk about cybersecurity, I kinda wanna get into this podcast is directed at business owners.

Justin: 06:27

How does this really matter to the owner of a business? If I put myself I mean, I own a business. What, what are the benefits of using it? What are the risks of of using it? And then really, what are the risks of not using artificial intelligence in my business?

Justin: 06:42

So I wanna get into that and then maybe kinda build a a template or a road map or a a checklist or whatever of how somebody could go from knowing nothing about AI, never using it anywhere in their business, and getting started. I'd like to leave people with that kind of a process. Alright? Does that work, guys?

Bryan: 06:59

Mhmm. Yep.

Justin: 07:01

Okay. So hopefully, you're ready. You've done your homework. You've got your notes prepped, and we are gonna jump into what do humans do better and what do machines do better. Let's, let's start with this topic, and and I'm not gonna moderate here.

Justin: 07:13

You guys just jump in and and start duking it out. What do we got?

Mario: 07:18

Well, I mean No. Go ahead, Brian. I was just gonna talk out of my ass probably, so go ahead.

Barinder: 07:25

I I I think sometimes when people think of AI, they they think of AI and compare it to the end game of a human intelligence. I think we're so far from it being a general human intelligence still, and I think that's a difficult benchmark to hit. So the way you frame that is what is what is AI and machine learning, what is it really good at? A good term that I've heard in the past, especially applied to machine learning, but I think also applies to deep learning, which is, you know, what AI is based on is they are prediction machines. And they're pattern recognition machines, prediction machines, and then now using generative AI as a subcomponent of that, they can actually problem solve for you.

Barinder: 08:09

So if we think of that categorization of learning and problem solving, whereas pattern recognition, that's where machine learning AI really excels.

Justin: 08:19

Okay. Yeah. Mario, you were

Mario: 08:23

Yeah. Well, to add on to to that, what I think humans will will do better is actually thinking of what the problem is that they want the machine to solve. So you tell it, like, give me a, you know, give me this, and it will output something based on information it's gathered previously. You know, where humans will kinda think of something that hasn't been thought of before, you know, where I feel like AI will not make that decision at the moment. You know, it will not come up with something out of the box new because it hasn't been done before.

Mario: 08:59

It hasn't gathered that information to present something new.

Bryan: 09:04

Yep. Okay.

Justin: 09:06

Any any other takes on this? I've

Barinder: 09:08

I've been thinking about, like, what the future looks like in this AI world a lot because I've got 2 young daughters. And what do I want them to grow up learning? What skills do I need to teach them? Because if it's something that AI is gonna be able to accomplish in the future, what is the role of the human? Right?

Barinder: 09:23

And if the future is gonna be 2 types of work, one is you're told by AI what to do, or the other is you get to tell AI what to do. I want my daughters to be on the telling AI what to do side of that equation. Right? And I think the elements that could come into creativity and integrated learning, I think that's always gonna be the most challenging aspect for AI to mimic, versus, you know, any of the odd predictable, outcomes, like, you know, that are spreadsheet driven, for example.

Bryan: 09:56

Right. I think Yeah. From my perspective, I'm seeing that AI a lot of people thought it would not be good at doing the creative side or, you know, the music, the arts, and things like that. And to an extent, it's still not, but I think it's surprising how many people are are actually, able to utilize AI to generate, and or at least soup starters. I like to call it soup starters for for on the art side.

Bryan: 10:24

I think where AI will shine the most is in, going through a massive amount of potential iterations of something, and give narrow down to a very small subset where we, as human beings, should focus. So, for example, in the medical industry is being able to go through, you know, 100000 different compounds and narrow down where which compound scientists and or, researchers should focus their efforts on first to get the better the the best results or potentially being able to use, AI to narrow down what type type of treatment from the 100 of 1000 or the thousands of different treatments available for a patient down to, like, the 3 or 4 that are the most likely to succeed for that patient based on their their makeup, which would take a human being too much time to do it. It just wouldn't be realistic. So I think that's where AI will be a lot better, than a human being is just in in handling the mass amounts of iterations that that we would have to go through to resolve. What would seem for, like, we wouldn't do that work for 1 individual or for one small thing, but with AI, we can now.

Barinder: 11:34

Yeah. An AI powered model in 2019 was able to detect lung cancer more accurately than doctors.

Bryan: 11:40

When

Barinder: 11:40

it when it comes to the health care, system, like Brian saying, it's gonna be reinvented because, AI and machine learning is so good at recognizing patterns, and a lot of diagnostic work is pattern recognition with large sets of data.

Justin: 11:55

Did you guys ever watch back to Hollywood. Right? Have you ever seen, House? Yes. The medical show House?

Barinder: 12:01

So good.

Justin: 12:02

Doctor House was the OGAI machine. Right? Nobody could solve problems like him because he could take all that information, that obscure data and and spit out the answer, but he wouldn't do it just like, hey, guys. Here's the answer. He's always gotta, like, kinda fuck with people along the way.

Justin: 12:19

Right? I love

Bryan: 12:19

that about this show. For it.

Justin: 12:21

Yeah. So, I I think, I think AI is best. That that is really what the machines do best, right, is is take massive amounts of information. And like you said, Brian, consolidate it down into maybe the the 2 or 3 points that we really need to look at to filter through. So, the first question that I think everybody but Mario failed that I asked was, did you use AI in preparing for this episode today?

Justin: 12:47

I did. I punched in what do humans do better and what do machines do better where AI is concerned, and guess what the machine came up with? Exactly what you guys just said, almost word for word, and I did zero research on it. That's the power of AI in my mind. This is a lazy man's

Bryan: 13:03

in my brain. This is

Justin: 13:04

a lazy man's dream. Oh, you've been doing you've been doing presentations. You've how many presentations have you done on AI now, Brian?

Bryan: 13:10

At least true. I mean,

Justin: 13:10

I think Yeah. Everybody in this room has done presentations on AI. I I mean, I'm just gonna expose myself. I haven't. I'm the only one here who hasn't.

Justin: 13:18

This is my first, show or presentation or whatever you wanna call it on AI. And I got all the answers right just by popping it into chat g b t. Question. Free version.

Mario: 13:29

For for all those other presentations, did you use AI to prepare for them?

Bryan: 13:34

Oh, me? Absolutely. Yeah. Because I was researching AI, so I would figure I'd I'd figure out what you could do with AI so I could show people what they could do with AI. Yeah.

Bryan: 13:42

Of course. Okay.

Justin: 13:44

Good. Good. Good. Here's a here's a piece of trivia. How long has AI been been in the works in development?

Justin: 13:54

Because and before I before you answer that, how long have people been talking about it in like, this is ubiquitous right now. Everybody, every platform that I use, every product that I buy, everybody is bragging about AI. We are AI. We use AI. Everything's AI.

Justin: 14:13

It's everywhere right now. I don't know that it's even that's why I'm like, let's define it because it's like saying the cloud. Well, what is the

Barinder: 14:19

cloud? AI is at max hype right now. I mean, I see the the stupid summarize my LinkedIn article with AI button. It's like, you click on that once and be like, oh, I don't need that feature. But everybody's trying to integrate AI because it's at max height.

Barinder: 14:33

I I but 2020, I think, is when shot GPT 3 hit. And that's when everybody said, oh, damn. This is good. This are the large language models and generative AI. When that hit during COVID, it it got picked up by mainstream news, and it it skyrocketed.

Barinder: 14:50

And a lot of things came together, whether it was a video cards by NVIDIA and the GPUs or, the the processing capacity. Right? And then all the the research work. But I think I think AI is a field of study has been around much like, you know, it's like a field of study, like chemistry or physics. I think it's been around for a very long time, like 40 years for machine learning.

Justin: 15:10

So I I will, date myself. I do that a lot. But, when I was in in college, for computer science, my this is CS 101. So it was like my or maybe 10. And I was like 102 or 1.

Justin: 15:23

Anyways, it was it was 2nd or 3rd semester, programming, c plus plus programming, and my professor, and this was back in 90, 95, 6, 96, or 97, and my professor was frequently talking about artificial intelligence. Back then, I'm like, I'm trying to figure out how to do, you know, just variable x plus plus and what that means. You know? And, yeah, print l and, you know, just stuff like that. And he's like, artificial intelligence.

Justin: 15:54

This is I'm like, dude, you are crazy. I don't know what you're talking about. That's not a thing. I mean, I didn't really think that, but I it's just like it was so far outside of my brain, and and what is that? Almost 30 years ago, and and it was it was a a main thing on his mind, on my professor's mind, which means he'd been studying it for quite a while.

Justin: 16:14

So, we're hearing about it. Like you said, Brenda, we had I think we had an inflection point in point in 2020 or thereabouts, but this has been researched, developed, studied, for decades. So just a just

Barinder: 16:29

a You know what this tells me, Justin? You paid way more attention in university than I ever did because I don't I don't remember them talking to AI and I went to school after.

Justin: 16:38

Okay. Well, well, story number 2, and this isn't as, really something to brag about, but in that same class, not very much, not very long after that, my phone started ringing and it was a customer because I had just started my computer repair business. And I, and I hit my own little, it was not an inflection point, but it was a, it was a pivotal moment where I had to decide, am I going to stay in school or am I going to take these calls? And I got up and I walked out of class, and I never finished college. So and and and I used the logic of, well, Bill Gates did it.

Justin: 17:14

Bill Gates is a college dropout. So clearly, I can pull this off.

Bryan: 17:18

Yeah. You might. Clearly.

Justin: 17:21

I I actually miss it. I I regret it in that I love programming. I love programming. I'd I've done it for a hobby for years. There's no way that a single programmer can make money.

Justin: 17:32

No nothing that I found because it's so labor intensive, Which is why though, when when we talk about something like this, it blows my mind, the the programming, the actual coding that it takes to make a machine do this kind of stuff. When you put computers against humans, it's crazy what our brains can do that we we can't even scratch the surface of it with technology.

Barinder: 17:57

Well So I mean, I talked to my guys, just before this, this, AI session. I'm like, hey. What are you guys using for AI? Because I know we use it on the marketing side, and that's how I use it, plus a couple other examples I can share. But they use it to dev develop code, and do Yep.

Barinder: 18:14

Their, their PowerShell work. They're actually using it for cogeneration. So AI based cogeneration has come a long way.

Justin: 18:21

Yeah. But somebody who still has to write the code to listen to you and then write the code that you want written. And and that's the stuff I knew a guy. I I knew one of the programmers, the developers that worked for I don't remember if it was Tesla. It was the the autonomous driving.

Justin: 18:38

It was either Tesla or it was, was it Uber that was the other major major player, and they're not anymore. They got out of it. They killed too many people or whatever. But no. I mean, for reals.

Justin: 18:52

Right? This was going on. I I thought self driving cars were imminent and that we were gonna be we wouldn't need driver's licenses within 10 years. I mean, the way that was going, until the, you know, people started dying and the lawyers got a hold of it. That was a a crazy technology that now is I mean, who knows what's gonna

Barinder: 19:10

come up with that? Wasn't the Uber situation that they, hired a guy from a competitor who brought over, materials from the other company he was formerly working at, and it was, you know, basically theft of, intellectual property. Wasn't that the thing with Mayway, Waymo, and Uber?

Justin: 19:27

I don't remember that detail. I I know that there were, there were fatalities on all sides of it. Yeah. And and that really messed things up. That, you know, just But

Mario: 19:37

a couple a couple of those, fatalities were actually not I remember reading something about Tesla, the and they they proved that it wasn't actually them that was the problem. It was that people kept looking at the car driving itself, and then that that caused an accident because the person wasn't paying attention to the road because they were staring at the car driving by itself.

Justin: 19:59

Well, as a as a Hollywood expert, as I previously illustrated, I watched a documentary on Tesla. And from what that documentary said, they were cutting they were they were shortcutting right and left. They were breaking regulations. They were they were doing a lot of things that they shouldn't have been doing according to the documentary that I watched, which was years ago.

Bryan: 20:21

Just recently, the self driving features of Tesla were not AI driven. They were they were program programmatically programmed into the car. Like, it was it was a procedural system versus now it's generative AI, and it's it's AI to my knowledge, that they just switched over. But it it was right up till, like, not too long ago, it wasn't.

Justin: 20:45

Well and, you know, back to our conversation of what humans do better and computers do better, this is one of the things that AI spit out at me was, navigating complex environments. Computers are not good at that. Right. You know, computers see things in ones and zeros. There's no tree.

Justin: 21:00

There's no person. There's no semi truck coming at you. It's a 1 or it's a 0. That's the only thing a computer knows. And so to to take these patterns, it can do great work at assimilating a whole bunch of data and sorting through it and and spitting something out.

Justin: 21:14

But then a human like you said, Brian, a human then says, okay. Of these 3 out of a million, these are the 3 it gave me. What what is the best option? But when when one of the choices is, oh, that's a clear blue sky. And the other choice is that's the side of a semi truck, and the computer made the wrong call and a guy got decapitated.

Justin: 21:36

Right? So that that's one of the cases, one of the situations that, this documentary talked about, one of the fatalities. So computers are still a ways off from being able to do that kind of work.

Barinder: 21:48

But if we if we think about it, like, while it's been 40 years in making for AI and research, it's really only last 4 years or so that in the mainstream, it's taken off.

Bryan: 21:57

Right.

Barinder: 21:57

And if we think about that mainstream LLM generative AI overlap in that Venn diagram, the modern tools like ChachiPT, Gemini, and etcetera that we have, We're at the dawn of it. We're at

Justin: 22:09

the beauty

Barinder: 22:09

infancy of modern AI. Where we're gonna be in 10 years is gonna be, you know, extraordinarily different. And so we have to be prepared for that versus just judging it on this linear growth model. AI is not Oh,

Justin: 22:25

it won't be linear at all.

Bryan: 22:25

It won't be linear. In fact, one of the one of the best quotes I heard was that, what what I think it was something to the effect of what we've accomplished over the last 100 years with technology we've been able to replicate in the last 10. With AI, we're gonna be replicating that every year or something like that. Right?

Justin: 22:41

So Wasn't that Jimmy from Hats dot ai?

Bryan: 22:43

Yeah. Yeah. Anytime anytime you you do, like, anybody who's not getting into AI now, I believe they're gonna regret it, because they're gonna be so far behind that, they won't have a chance to catch up. It's like I think he he even said that too. If it's like somebody who, you know, didn't hop on the Internet initially, there was there's there's there's gonna be a lot of growth potential, a lot of money potential, a lot of things that you can do to earn and and and accelerate your business.

Bryan: 23:09

And if you don't do it now, you're you're gonna miss out because your competitors are gonna be the ones that are gonna take advantage of that.

Justin: 23:15

So I wanna I I I wanna wait. Hold on. I wanna I wanna interject or pause because this you guys are jumping ahead about 3 bullet points. Because we're gonna come back to this, and I wanna dig into it. So so, be before we do that, let's talk about cybersecurity.

Justin: 23:30

How does AI impact for good and bad, the world of cybersecurity? And then we're gonna talk about the risks of AI in business, the risks of not using it, which is kind of where you guys are going, and then we'll give the path for how people can actually use it productively. Because there is this this is an episode where we're gonna talk about more than just cybersecurity. We're talking about, and I mentioned this when we interviewed Jimmy from Hats dot ai that, you know, this is a case where we can help our clients actually make money or become more streamlined, more productive using technology, where cybersecurity is just, I mean, largely, it's a sunk cost.

Bryan: 24:05

It's Preventive. Right? Yeah.

Justin: 24:07

Yeah. Yeah. So, so let's go back to cybersecurity for a minute. And how is AI impacting this? And, Mario, I want you to lead because I looked through the slide deck of the presentation that you gave recently, and you kind of you got into this in a fair amount of detail.

Justin: 24:23

And then, Barinder, I don't know if you did or not because I didn't see your full slide deck. But, Mario, if you'll go first and then and then Barinder, and we can kinda just all all spitball here. But what what is the good and the bad of AI in cybersecurity? So the bad, you you know, essentially, what hackers did, like,

Mario: 24:40

a few years ago and, you know, people will remember, they would get an email that, half, you know, half the email was misspelled or just didn't make any sense or didn't look, you know, real. And they would know that this is not legitimate. It's not really coming from Microsoft and people would delete it. Now hackers are using because a lot of these hackers are from other countries. They're using AI in, like, chat gbt to, you know, fix their grammar, make the email look, you know, you know, like, hey, it's really coming from Microsoft.

Mario: 25:16

So those misspelled emails and, you know, mistakes that they used to make are no longer being made. Then when they would actually get into somebody's system or, like, for example, somebody's Microsoft email account, they used to spend some time in you know, I I remember hearing that, like, sometimes up to 8 months before they would actually do something about it. Just Now they're able to yeah. They would now be able to use AI to learn exactly what's going on in your mailbox. They know who you're interacting with the most, what kind of business you're dealing with.

Mario: 25:53

Like, what not only what your business is doing, but the type of business is that you're working with, who their contact person is that goes in, you know, the accounts receivable and accounts payable. They're using AI to go through your entire mailbox and within seconds or, you know, maybe minutes, be able to get a full list of exactly what they need to know. You know, where in the past, they would literally have to go 1 by 1 reading, you know, waiting for you to get a new email. Certain times of month, they'll know, you know, when you're receiving payments or when you're making payments. They're now using AI to generate all this stuff and be able to access the information that they're looking for within within a couple clicks.

Mario: 26:38

So now hackers are using this to make their job a lot faster, a lot easier, and they they're they're it's helping them pretty much know how much you're worth, you know, the company's worth, how much you have at a in bank account at a certain time of the month, you know, when you're receiving payments and stuff like that. They know exactly everything about your company, sometimes more than what you know about the the company.

Justin: 27:04

Okay. Brenna, do you have any thoughts on that? And maybe the other side, if you've got information of how how AI is helping us fight this battle.

Barinder: 27:12

So I can start with the AI fighting, and then I'll jump to the more interesting topic of all the threats because that's always, you know, the fun stuff. So,

Justin: 27:19

it's a good time.

Barinder: 27:21

Yeah. It's it it

Justin: 27:21

Helps me sleep at night. I can promise you that.

Barinder: 27:23

Oh, yeah. Yeah. As an MSP, we have nothing to worry about.

Justin: 27:27

No.

Barinder: 27:28

When it comes to protecting, obviously, the big cybersecurity vendors, the CrowdStrike, the Sentinel, of the world, they are using all this data and telemetry and using AI for pattern recognition to block the malicious, efforts. Right? And that's where, shifting to large amounts of data for that pattern recognition. Again, we are very good at that. So all the cybersecurity mainstream cybersecurity vendors all integrating some aspect of AI into, their protection mechanisms.

Barinder: 27:52

That is this person acting in a way they should be, or should we block it? Microsoft's doing the same thing. So that there's a lot of, integration into those technologies. But when it comes to the sophistication, and the opportunities for malicious actors to get into environments, I mean, they're so much more empowered now, like what Mario said, they can, like, look at so much data, interpret it through these AI models, and react faster. But then there's a social engineering side of it where you can mimic somebody's voice using AI, video using AI.

Barinder: 28:26

There was the example Yeah. Recently of the Hong Kong Bank where a finance worker in this large organization, was on a video call with her CFO and other colleagues, but Right. Both all the CFO and the other colleagues were actually AI generated, looked like her colleagues, sounded like her colleagues. She paid out $25,000,000 to this malicious actor because it was social engineering using AI, video deepfake technology. Also, she was on a video call with it was unreal.

Barinder: 29:02

And I think the types of threats that the average person and the average business is gonna be, subject to, I I don't think we're ready for that from the social engineering side of it. Because we have so much data that's out on the web already about us, That, like Mario said, it's easy for these malicious actors to buy it, integrate it, and then be able to act upon it, that that we're just not ready for. And then, of course, there's the internal threats for the organizations that haven't learned about AI yet. Are you and maybe I'm jumping too far ahead, but this is really important topic. When it comes to us as our organizations or our clients, if you're a business, your staff are already starting to use AI.

Barinder: 29:46

I bet if any of the listeners went and took a poll of their their employees, I bet some of them are you just going on to chat GPT because it's free, for basic use. I bet they're able to punch in some things, punch in some information, and it's gonna spit back some results. Well, what if they wanted to get a leg up and put in some sensitive information? Let's say it's an accountant and they or a lawyer, and they copy paste what their current content is. Hey, make it, a little bit different, or can you provide your feedback on this report?

Barinder: 30:16

Now they just gave up some sensitive information, confidential information into an AI system that all of a sudden is now public, AI. It is not their content anymore. It's used to train the language models further. They've lost control of that data, and they've essentially leaked it out unintentionally out of their protected ecosystem into, this, you know, public OpenAI Gemini type of ecosystem. Right.

Barinder: 30:42

It has no protections around it. That's where Microsoft Copilot comes in and then creates this mode around it, but we don't have to get into Microsoft Copilot just yet if we don't want to.

Justin: 30:53

Alright. That was that was a lot. Brenda, you guys have any thoughts there?

Bryan: 30:56

Yeah. I can jump in real quick because, one of the things that I'm seeing, through my research is that the criminals are using AI to not only do the, you know, the pattern recognition and going through somebody's inbox once they're in, they're also using it to iterate 100 of thousands of different ways of getting into people's systems. Right? Different types of emails. Right?

Bryan: 31:18

And to the point now like, I'm gonna jump over to the advertising side because you might you might be able to see what I'm where I'm going with this a little bit more on a cybersecurity when we're talking about marketing and advertising. They're we're they're going to be they're they're currently using and able to use AI and will continue to do so in a way that will customize marketing and advertising for your specific tastes, your color palette, the way that you think, the way that you believe, your your belief systems, your political motivations, all of it, and they're gonna all be able to tailor marketing and advertising to use specifically on the fly as you're watching a video. Right? And so I'm not just talking about, you know, like, they're gonna create, you know, content for a specific subgroup and then send it out, and another subgroup will get some no. You, specifically, as an individual, Mario, for example, will get a specific recording or video.

Bryan: 32:11

And as you're watching it, it could get your haptic feedback, your feedback from the way that you're looking, and change the video on the fly as they go based on on different metrics, like whether you like it or whether you don't like it, right, based on facial things. That kind of thing will be playing into AI, especially when we're talking about what Bernadu was just talking about where, the the the bank got hacked or whatever via via social engineering, that's where AI is going next. Right? So now imagine, what used to take people hours and hours to craft a a message to try to fool somebody, they're able now to send out, like, you know, 50, 60, 70 of them all at once to every single person in your organization. And maybe they won't know who will click on it or who will touch on or who will will jump in on the meeting, but somebody will because, you know, it's like it's like in marketing.

Bryan: 33:05

It's like the more you do, the more likely you're able to get results. Right? And it's sort of gonna be the same thing with AI. The more that you put out there, the more likely somebody's gonna interact with it. Okay.

Justin: 33:18

So guys, let's let's talk about this is a lot of scary stuff. You know, the marketing, I I don't I don't have a problem with marketing. If No. No. I just somebody, like, hacks my brain and makes me buy something that makes me happy that I like or that tastes good or whatever, go for it.

Justin: 33:32

Knock yourself out.

Bryan: 33:33

Right. Now move that into cyber security is what I was getting at.

Justin: 33:36

If you

Bryan: 33:36

can understand the marketing and how they could use it for marketing, now shift that over to cyber security.

Justin: 33:40

Now they use it to

Bryan: 33:41

fool you to do something you're not supposed to do. So it's like marketing, but for you to click on something you're not supposed to and give them something you're not supposed to.

Justin: 33:50

So this is the bad side. I said the good and the bad. I'm hearing a lot of bad. I'm not hearing a lot of good. And we can take AI out of it if we need to, but what what are we what are the countermeasures?

Justin: 34:01

Right? So this is it it's getting to the point where if we just rely on the human brain, this task of securing data and money and inform you know, it's almost impossible. And and you might even argue that it is impossible. So what do we do then in the world of cybersecurity to protect against these deep fakes and this, you know, basically the hacking of the human brain? Do you guys have thoughts on that?

Justin: 34:26

Because otherwise, we close out the podcast and was like, well, shit. We're screwed. You know? Like, let's go live under a a box of rocks somewhere and and go off grid. What what's the answer to this?

Bryan: 34:37

The funny thing is that there's always the the the criminals are always more innovative at the beginning than than people who are on it. So you'll always see criminals have a leg up at the beginning because they're they're looking at this going, I have a huge potential to make money, versus the other half of the side is, like, I don't wanna spend money to prevent somebody from hacking me. And so I I wanna spend as least as I possibly can. And so the innovation is less quick on that side because it's a cost versus an income. And so

Justin: 35:06

I think define the problem, Brian. You just define the problem of cybersecurity. There's money to be made by the hackers, and there's only money to be spent by the the ones protecting. That's a terrible situation. Go on go on.

Bryan: 35:20

It is. And I think not a whole lot of solutions right now. And there are a lot of solutions. Let me get me wrong. There's tons of people who are using AI to protect us, put it in new systems.

Bryan: 35:29

They're just a lot slower to do it. Right? I can see a day where, you know, if you go to make a payment, there'll be an AI background system that's gonna go, okay. You're making a payment to this account. Who actually owns that account?

Bryan: 35:40

Let's go in the back end system. Let's go let's go connecting to the bank and go, hey. Is the name on this account correct? Is it really this person? And is this the, you know, is this a recently opened account, or is this an offshore account?

Bryan: 35:51

Right? There'll be some AI back and forth handshaking to verify that the information is accurate and correct, and we're not sending payments where we're not supposed to. That doesn't happen right now. People just get an account number. They go, okay.

Bryan: 36:00

Punch it into the computer and hit send, and that's it. Right? Like a bunch of dodos. So I think we'll see a lot of this this AI coming in the next year or so, but it's always gonna lag behind in my opinion.

Justin: 36:13

So I would say, another another way to address this, because you're right, the the technology of security does lag behind the technology of the hackers. I mean, it has to, right, because they come up, they find the vulnerabilities, they find the weak spots, and then we have to detect that, address it, and modify our behavior or the technology. Right? But in the end, they're always trying to get money. That's always the end result of cybercrime.

Justin: 36:44

And so I I would I would say that our our best defense really comes down to proper training procedures, processes, especially where money is concerned.

Barinder: 36:58

So I think the fundamentals don't change. It's it's Right. The technology changes, your fundamentals are do your do your core cybersecurity protections. Make sure i's are dotted, t's are crossed because with AI, any vulnerability is gonna be exploited faster. Identity verification is gonna be a challenge with all the the social engineering, whether it's voice or visual that you can do.

Barinder: 37:18

Identity verification is top of mind for anybody at the IT space right now. But even beyond the IT space as a business person or a bank, how do you verify that the person who you just de sign that document is truly that person or whatever. Right? You can't even jump on a call with them to verify confirm. Right?

Barinder: 37:32

I think those are some fundamental security awareness training. It's gonna change, but it's still security awareness training. Everybody has to have it. Take care of the fundamentals in your business because cybersecurity is about reduction of risk. You can't eliminate it, but let's make sure that we reduce the risk, as much as we can.

Barinder: 37:48

And that's all we can really do.

Mario: 37:51

Yeah. Yeah. I I agree. I think in every one of our security stacks, the top of the stack is always, employee education. You know, they they have to you know, that's where they still have to use their brain.

Mario: 38:04

They have to use, like, should I do this? Should I click on this? Should I go to the Walmart down the street and buy these, you know, these gift cards that supposedly the the, you know, the s the CEO is asking me to surprise the the staff with. You know, they still need to, you you know, people still need to use their brain. Mhmm.

Mario: 38:24

That that's the first and, you know, first thing that everybody should be doing. And that's the first part of our stack is educating our employees, educating our customers' employees, and having them make the right decision.

Justin: 38:38

See. And now we're back to what do humans do better and what do machines do better. So machines do repetitive tasks well. They do, you know, gathering a whole bunch of data and and picking up on patterns and whatever, and they can where where humans are much better though is just sensing that something's off. Right?

Justin: 38:59

A computer has to be taught every possible scenario and what to do about it. But a human can just we just have a gut feel. Now I I could do a whole episode on that too on why we have these intuition or or gut feelings or whatever we wanna call it, which is actually just, us doing, you know, looking at information and taking it in and on a subconscious level, you know, our brains kinda spitting it back up and going warning warning because it's seen something. Right? But we're better at that than machines are.

Justin: 39:28

So, and this is where we have to when we're fighting this, almost impossible threat of artificial intelligence, we we really have to get back to, yeah, the the training, the education, the always be on your toes. Be watching for anything that just looks off, and then it comes back to culture. We did a whole episode on culture. You know, if our culture is, like the the case I keep talking about in the city of Fort Worth where they fired guys, I mean, turn into lawsuits and and it's like, hey. All all that happened was, at least from the article, the the IT guy goes, hey.

Justin: 40:01

There's a problem over here that we need to address, and that culture punished and fired and and lawsuits came about instead of saying, oh my god. Thank you. Now let's let's all band together, and let's all figure out how to solve this problem. So, I I believe that will always be our number one, weapon in this war on cybercrime is we have to bend together. We have to use our our human brains.

Justin: 40:25

We and we just keep iterating through our processes and procedures to to protect things. So, any other thoughts on AI, the the good and the bad as it pertains to cybersecurity? If not, then let's move on to using AI in our business. And and we're gonna kinda start with the risks of it, but we're moving out of cybersecurity. Alright?

Justin: 40:51

We're we're moving more into and we've touched on this when we had Jimmy on from Hat's I AI where, you know, data leaks and and stuff like that. And, Brian, I know this has kinda been one of your you've you've talked quite a bit about us. Do you wanna take that one?

Bryan: 41:07

Yeah. Sure. I'll I'll just start with the fur the very first thing, which is, the very first thing anybody everybody all all businesses should do today right now, if you have don't already have it, is have an AI policy. And your first part of the policy should just be like, hey, until we have a full on policy, do not use AI. And then go through and figure out how, you're you're planning on using.

Bryan: 41:29

Because if we don't give people the tools in order to work, they will they will find their own tools and and and use those. We call that an IT shadow IT. There's there's even in in in the clients that I manage, there are people who are using tools that are not authorized by the company. They're personal accounts. They're uploading files to those personal accounts.

Bryan: 41:50

And no matter how many times I have a conversation with a client, it still happens. And so we gotta identify what those what those those tools are. But when it comes to AI, the the danger is even greater because you're essentially giving up access to your critical key and data and basically providing it, like Boringer said, to the AI model to train. So I like to use the example of, let's say, you as an organization wrote the a really great book, right, a book on how to do what you do the very best. And you you you have somebody who goes, you know what?

Bryan: 42:20

I was tasked with with spell checking this whole document. And they send it to AI, a paid or a free version, you know, chat gpt, whatever the case may be, to do a spell check, and the AI in the background goes, this is fantastic material. I'm gonna just, like, store that right here for now. And, you know, I'm gonna I'm gonna and maybe I'll use that later. And then somebody, you know, in another country or somewhere else goes, hey, I'm I have a question about this, and they ask the the AI about what you wrote, and it's gonna go, hey, I got something on that.

Bryan: 42:48

And it's gonna grab the information. It's gonna spit it back out to that person. And now your your information is being utilized by someone else without your consent, without your authorization. Your copyrighted information is no longer copyrighted. Right?

Bryan: 43:01

And so if you don't have an AI policy in your organization that says what you can and can't do, what you can use AI for, and which tools you're allowed to use, do it now. That's that's my entire message.

Justin: 43:12

Or or worse, you said your your, your intellectual property that's not copyrighted. What how did you say that?

Bryan: 43:20

Your your your your intellectual property

Barinder: 43:21

Are you

Bryan: 43:22

you're is no longer

Justin: 43:22

copyrighted. Information.

Bryan: 43:24

Right?

Justin: 43:25

Or worse, somebody else takes it and copyrights it

Bryan: 43:28

Well, there you go.

Justin: 43:28

And does the better job of it. You're not It

Bryan: 43:30

it might not even be your data to begin with. Maybe you're storing somebody else's data, and now that data is out there. Right? So there's a lot of implications here.

Barinder: 43:38

And and there's lawsuits. Right? There's there's court cases right now with the authors suing, these, the GPT companies because their content is now you can you can write yourself a story. Tell it tell me the story in the voice of, Stephen King, and make it scary. I'd be like, it could do that.

Barinder: 43:56

Well, how can it do that? Because it's read every single Stephen King novel. It's been uploaded and knows about it. Well, is that theft intellectual property? And and if the the the mainstream, authors are having this issue, I mean, we have to get this solved.

Barinder: 44:10

But until then, yeah, don't upload your content. That's your intellectual property or your client's intellectual property into these models.

Mario: 44:18

Well, I'll give you a funny story. A friend of mine, is a lawyer, and he was telling me a story of another firm. And I believe they're they're out of New York. They had an associate, actually, he was tasked to, like, create, you know, put together a case against, like, you know, somebody that was for a lawsuit. And the associate went online and said, give me, you know, you know, give me an appeal of, like, something or whatever.

Mario: 44:46

And it gave him, like, a list of, like, 7 different cases that he just copied and pasted and put it into the the lawsuit. And then the lawyer is like, listen. You know, where you you or sort of the judge called them out on. I was like, I don't remember any of these cases. Show me where these cases are.

Mario: 45:05

It turned out

Justin: 45:06

They were all fake.

Mario: 45:07

It turned out it it they were all fake.

Barinder: 45:09

Yeah. I mean, there's made up an answer.

Mario: 45:12

Yeah. And it actually went into detail. It was like, you know, American Airlines doing this and whatever, blah blah. And it it just the whole thing was made up.

Bryan: 45:21

Yeah.

Justin: 45:21

And the

Mario: 45:22

guy ended up the the getting debarred, and there's you know, they they got they got fucked

Bryan: 45:35

And And and that's all on how you use AI. Right? Because you can it can hallucinate if you've got it on creative mode. But if you've got it on specific, it might still, but less so. So, yeah, it's a little complicated.

Barinder: 45:47

Yep. But I think that's where the opportunity now, like, comes in for, for businesses. Yes. You have to have an AI policy to say don't use AI until we have a policy, but how do you build a policy? You have to go learn.

Barinder: 45:58

And I think there's an

Justin: 45:59

AI, what do you mean how do you build a policy? Type it in the chat GPT. It spits 1 out. Come on, Brandar. I started this episode off by asking if you guys are doing this, and you're you're doing a whole podcast on it, and you still aren't given the right answer.

Barinder: 46:12

Yes. Yes. That's right. We need to, submit to our AI overload now Great. That That's right.

Barinder: 46:18

Yes. Yeah.

Justin: 46:20

Damn it, Brenda.

Bryan: 46:21

It it

Barinder: 46:22

it's such an exciting time. I know we've been talking a lot of, like, worry disaster scenarios and and cybersecurity risk, and that's rightfully the way to think about it. But it is an exciting time for our business. Like, if rarely do our we in our organizations have an ability to have, like, a stepwise function improvement in our productivity by using these tools, and it's like a new tool that we've never had before. Yeah.

Barinder: 46:46

We've never had visibility to do r and d in a typical, like, professional services business, for example. But now you can invest in your own business using these AI tools and have extreme leverage for productivity, whether in your marketing department, your sales department, you're writing technical reports, you could use these tools to all of a sudden even if it was, like, minimally 10% more efficient for that department, That's a ton of money. Right?

Mario: 47:13

Like Yeah.

Barinder: 47:13

We've been selling cyber security risk which is a cost center to the business only it's like insurance. Great. But nobody likes paying for insurance. This is like profit. It is if anybody doesn't invest in some, AI now, despite the risk and all the more reason to learn about it, we're as business owners, we're leaving money on the table.

Barinder: 47:32

And, yes, it's at the dawn of AI, but it's gonna go fast. So our learning has to be just at the same curve as the growth of AI.

Justin: 47:39

Yeah. I love the way you said that this is because it's not a new technology, and that's kinda what I was saying before. I've I was being told about this back in the nineties. But you said this is a new tool that we've never had before. So we've been able to harness this now and find ways to use it.

Bryan: 47:54

Right.

Justin: 47:55

If you look at, the difference between, like, a hand saw and a table saw. Right? Any any woodworkers in here, how fast you can cut a wooden beam or, you know, a whatever, even just a 2 by 4, 2 by 6 with a table saw or a chop saw versus, the old hand saws. Right?

Bryan: 48:13

Right.

Justin: 48:14

Amazing boost in productivity and output. Mhmm. But if you hit your finger on that table saw or that chop saw, I mean, that's ugly. And, you know, so we've got this tool, but damn it. We better learn how to use it.

Bryan: 48:29

That's right.

Justin: 48:30

And we better be super careful, and we better learn the rules around it. And there better be policies and procedures and right? So it's an amazing tool, but, goddamn it, be careful.

Barinder: 48:39

Just right? What kind of IT life do you live? It doesn't have a keyboard. I don't use it.

Justin: 48:44

Hey. Listen. I've gotta de stress sometimes, and I do that in the shop.

Barinder: 48:49

You're you're losing your, legit nerd cred here by using human power tools?

Justin: 48:55

Human power tools. I'm I'm I'm currently building out a a quite a pretty involved bar inside my house. So it it shows my my inner, I I You should have been automating that

Bryan: 49:10

with, like, a raspberry pi or

Justin: 49:11

something like that. Oh my god. There's so many things you can do with raspberry pi's. Let's not go there. Well, I'll you'll you'll turn me right back into a nerd.

Justin: 49:19

Let's so we the risks of using AI in your business, the risks of not using AI in your business, we've kind of already touched on that, but do we wanna take a quick, like, you know, just 30 seconds? Let's let's get the through this one quick because this is where you guys went when I had to cut you all off.

Bryan: 49:36

Yeah. Okay. I'll start. Don't be afraid of,

Justin: 49:39

like I knew you would, because this is your thing.

Bryan: 49:41

Yeah. Like, don't be afraid of AI because AI is not gonna replace anybody. But what will happen is if you don't engage with AI, if you don't learn how to use AI, your competition will. They will start using AI. They will become more productive than you by by the factor of 4 or more.

Bryan: 49:58

They will be able to reduce their prices. They will be able to become more efficient. They're gonna be able to blow you out of the water like there's no tomorrow. And so, yes, be afraid of making sure you do it right, but don't be so afraid that that you don't you don't test and iterate. So get get on the bandwagon.

Bryan: 50:13

Get working with AI. Start using the tools. Learn it learn it while we're in its infancy right now because you are able to implement it. You're able to learn on it slow and figure out where it's gonna make sense for your business and how you're gonna be able to empower your employees to be more productive and do it in a way that's safely. So start slow, figure it out, and, but don't be so afraid, that you're don't you you just stay away from it altogether because that's that's that's that's you're you're gonna end up just losing all all of it.

Bryan: 50:44

So

Justin: 50:46

So let's lay that groundwork now. Let's let's move into, and we're gonna wrap up with this. Okay. But and let's let's I'm I'm the I'm the example. I'll be your guinea pig.

Justin: 50:56

I know nothing about artificial intelligence because I even still call it artificial intelligence, which shows my ignorance. What do you tell me as a business owner? And we are not talking tech right now. Right? We're talking, what can I do in my business, and how do I do it in its simplest form to just get 1% better?

Justin: 51:14

Brian, I know how you love that term.

Bryan: 51:16

I love that term. Are you asking me or some of the other folks there?

Justin: 51:20

I mean, you you start, but I'm throwing this out to everybody. But I wanna like, let's let's keep this pretty high level simple. What what is a good takeaway that if, if, you know, one of our audience listens to this whole thing, they're like, okay. I'm ready. I wanna take that next step.

Justin: 51:35

What is a safe place to start that's gonna do, you know, the the least cost in both time, money, and effort, and the best output?

Bryan: 51:43

Okay. I'll I'll I'll start with this. I love using AI as a soup starter. Right? I I use that term a lot.

Bryan: 51:50

Those of you who are not familiar what it is, like, you're making a soup, you you and you start with a soup starter. It's like a head start. Right? When I'm trying to think of what I wanna talk about, what I wanna do, I often struggle with figuring out, you know, what what? Where do I start?

Bryan: 52:03

How do I get this thing? I need to do this project, but I don't know where to begin. You can use AI in that capacity. It's a very easy way to start using it, and it's a 100% risk free because you're not sending any information to AI. You're just asking it questions.

Bryan: 52:17

Hey. I got this project. I'm working on, you know, writing an article about, cybersecurity. What are the 10 bullet point things that I should talk about that are the most important right now? Boom.

Bryan: 52:27

It spits out 10 things. You're like, oh, jeez. You know what? I like 5 of those. I'm gonna talk about those.

Bryan: 52:31

Can you expand on those these 5 and and give me 5 more bullet points on each that I should cover when I'm talking about those Bam. Right? Now I've just done all of the planning work, and now I can go and talk about those things because I know those things, but I didn't know where to start. Right?

Justin: 52:45

So let me clarify. Sure. Because we're using AI as a, almost as a it's a generic term, but we're using it for something specific.

Barinder: 52:54

This this

Bryan: 52:54

in this case Take

Justin: 52:55

me back. Like, I don't even know what you're talking about. Okay.

Bryan: 52:57

In this

Justin: 52:58

case How do I do what you're talking about, dude?

Bryan: 53:00

Yeah. In this case, I'm using chat gpt or Gemini. I'm I'm using any of those those

Justin: 53:03

Where do I get to that? Is that a website? Are we just okay.

Bryan: 53:06

Yeah. So

Justin: 53:07

So go to Google, punch in chat gpt.

Bryan: 53:08

Right. And and OpenAI has a has a website. There's there's a bunch of them. Generative AI is, basically, the the language part of it. Right?

Bryan: 53:16

So you can use Bing has the built in Copilot. You can use chat 2 p t. I recommend Bing if you're paying for Office 365. You could use Bing, and and it'll and when you're logged in Office 365, it should say protected. Even if you're not paying for Copilot, it should still stay protected there because that information is still, allows you to connect and and do things, just not not to the extent of what you would get if you were paying for Copilot.

Bryan: 53:42

And I'll let Brendon get into that later. That's that's an easy that's an easy way that I've used AI, and it's a very basic risk free way. There's a whole whack of other places things that you can do, but probably more than I want to go into right now, but I'll I'll I'll pass the floor back off to my colleagues.

Barinder: 53:59

Okay. I will share it in 2 different ways. 1 is personal life. 1 is professional life. In my personal life, I'll share how I got into, AI first.

Barinder: 54:08

Right? And my favorite example is my favorite use case for AI. I have 2 little girls. Yep. My my girl is around

Bryan: 54:14

Love it.

Barinder: 54:15

3 years old. I I I needed to make up stories to put her to bed. I am an IT nerd. I'm not a creative person making up stories, and she wants to hear about unicorns and and fluffy snowflakes and and rainbows. Well or pirates or whatever it might be the flavor of that night.

Barinder: 54:29

So I would go on to chat g p t. So chat.openai.com. It's free. Try it out. So in a personal life, go play with that, and I would generate a new story every night.

Barinder: 54:38

Give it a prompt. Your prompt is your, your prompt is your, is, your, request to AI. It is basically, hey, tell me a story with my daughter named x and her favorite dog named Duke. They go on an adventure. They come across a unicorn.

Barinder: 54:56

It'll spit out whatever story you want. Great. I helped my daughter go to sleep. Like it was fantastic. That's how I got involved.

Barinder: 55:03

So you can use it to find recipes. You can use it for almost anything you can imagine. Just start playing with it, and and it's free. So that's a great place to start on your professional life, but don't totally different. Don't give member.

Barinder: 55:16

Don't give chat g p t anything confidential, intellectual property, private information because it is you have lost control over that. And so in your professional life, where do you start? Most of our clients are, use m 365, Microsoft 365 businesses. So if you pay for a Copilot license, the way Microsoft has brought and built out Copilot, it is using that same OpenAI chat GPT technology, but brought it inside your Microsoft tenant just like you have your SharePoint, Outlook, etcetera, all in your Microsoft tenant circle. Inside that circle is now your large language model, feature set.

Barinder: 55:54

So your data never leaves your, your Microsoft ecosystem. You control it. That's where the moat and the boundary is. There's no leakage of data. Great.

Barinder: 56:03

Right. You can still have intercompany leakage of data. So I don't encourage people to roll out, Copilot without talking to their IT professional first. Right. You have some experience with it, because you have to make sure you have the right permission sets.

Barinder: 56:15

Otherwise, somebody in your organization is gonna be like, hey, Copilot, can you tell me how much everybody's getting paid? And if they somehow magically have access to it, and you haven't done your internal walls properly, you're gonna have some leakage of data. So talk to somebody who's an IT professional and has some experience with this to do it the right way.

Bryan: 56:31

Right.

Barinder: 56:31

But that's how you do it. And if you just wanna get started, a basic team's premium license, which is the most, robust of the Microsoft copilot feature sets, it gets you started, has your intelligent recap on a meeting. I can go away for a 2 week vacation, come back, and ask Copilot or, or teams quest it it'll give me a summary of what I missed in the meeting. But if I have Copilot, I can even ask it. Hey.

Barinder: 56:58

While I was away for those 2 weeks, what did I miss? What was in my sales channel? Can you summarize all the conversations I missed that might be important to me? What were my takeaways? What were other people's takeaways?

Barinder: 57:09

You can ask it these questions. A very easy place to get started in the business side, but do talk to your IT professional first because if you just go roll it out, you might land yourself in some hot water.

Bryan: 57:20

Great.

Justin: 57:22

Alright. Mario, do you have any thoughts on this?

Mario: 57:24

Yeah. So I mean, the way I've been using, AI is well, because I when I'm writing a letter, you know, either to a client or to, you know, my employees, I I I tend to just, you know, not necessarily just misspell, but my grammar is probably at a, you know, 5th grade level. So I'll write my letter, and I will put it into chat CPT, and I'll say, please rewrite this. Or I'll say rewrite. Only Brian Brian's the only one I know that's very polite to his AI.

Mario: 57:58

He always says, please.

Bryan: 57:59

I do. Because when they become their own landlord, I don't want it to attack me.

Mario: 58:05

I'll tell it, like, rewrite this this the following letter and I'll paste it in there. It will rewrite it more professional, you know, and, then I'll copy it, you know, reread it, make sure that it highlighted the points correctly, and then I will send it. So I use that more just to kind of, like, fix what I'm doing. Yeah. You know, like, my weak points, I'm I'm using it to to fix what I'm supposed to, you know, do better.

Mario: 58:31

You know? So that that that is a very common easy way to use, you know, like chat gpt and and and and, you know, you know, anybody can start using it like that. But, again, be careful what you're having it put in. You know, you if you're writing a book, you know, you don't wanna upload the whole book and say, you know, do this because it will, it will share it with others and use that to gather more information to learn Mhmm. You know, from for for future use.

Justin: 59:03

I'm surprised nobody yet has mentioned Alexa.

Barinder: 59:06

But but before we jump into that, I gotta jump off, guys. My next meeting is here. So thank you for inviting me again. This is a fun podcast. I just a a quick plug, for what, our marketing folks are doing.

Barinder: 59:21

We got Copilot webinars coming up. If you add us on LinkedIn or follow us there, there's a guide to download or something for anybody who's interested in Copilot side. Go find me, and, and there'll be a whole bunch of valuable content coming up for small businesses that have this question of how to get started. But sorry. I gotta jump, guys.

Justin: 59:38

Thanks for sending my army of robots after you for getting off the podcast early, just so

Barinder: 59:41

I know.

Mario: 59:42

You you

Barinder: 59:42

you you

Justin: 59:42

You better be watching your back.

Barinder: 59:43

You know what? You need to send us a invite to your bar once you're done building it now that I know you're building it.

Justin: 59:49

I'll do that.

Barinder: 59:50

Take care, guys.

Mario: 59:51

Take care, Brenda.

Justin: 59:52

See you, Brenda.

Mario: 59:55

One more one more thing to, what we do for AI with AI is we use it to kind of help us write, like, quick little, snippets that we post to social media. You know, like, give us, like, you know, a 50 word, you know, document about, like, something about, you know, AI, you know, and we we'll put we'll help that with our marketing and stuff like that to kinda gather some information to put on on our social media as well. So we we we use that for, like, marketing. You know, you can help, but it kinda like what Brian, was mentioning before. You kinda wanna have that starting point, that spark, and then, you know, enter it and and then have it expand on that for you or help you expand on.

Justin: 01:00:43

Now what I started to say before, Brenda so rudely left in the middle of our transmission, I'm surprised nobody's talked about Amazon Alexa. Like, this is we've been using so and this is why I started with, how do you define AI? It's kinda like the cloud. Like, ask somebody what the cloud is, and they don't really know. And broadly speaking, the cloud is simply any kind of computing processing that happens away from your own device.

Justin: 01:01:11

Right? I mean, it it it's that simple. It's anything, and everything is considered the cloud. Right. So the whole movement of move to the cloud is like, well, what are you talking about?

Justin: 01:01:20

What what specifically do you mean? And and AI, you know, where when did it come onto the scene in 2020? Well, I mean, we've had Google for a long time, and Google is AI. Like, the Google search, that's all artificial intelligence, that that's running that. It's, you know, some iteration of it.

Justin: 01:01:38

But more importantly, I was, it was hard because Brian, I I consider you a close friend. We've known each other forever. In a lot of ways, I think we're alike. But in this way, we are we couldn't be more opposite. You're like, I'm always nice to my machines.

Justin: 01:01:55

I am a dick to mine. Like, I swear at Alexa all the time. I call her the worst names, and and she gets really offended, actually. So when

Bryan: 01:02:04

we're talking

Justin: 01:02:04

about what machines and

Bryan: 01:02:07

used to hear me talk to my Tesla. I I

Justin: 01:02:10

hate to think some time. Good. Good. Well, I'm glad that we can bond over this because I'm, like, really upset. I, Alexa gets okay.

Justin: 01:02:19

So back when we were talking about what what's good and what do computers do good, what do, people do good, do well. I I do know grammar better than

Bryan: 01:02:27

anyways,

Justin: 01:02:30

computers don't understand emotions.

Bryan: 01:02:32

Yeah. Yeah. Yeah.

Justin: 01:02:33

They could kind of fake it. Well, Alexa, they programmed they've done a good job of programming some fake emotions into Alexa because she gets really butt hurt if you call her names. Go home and try it. You'll thank me later. Also, I don't know if you know this.

Justin: 01:02:47

Just a fun fact of where you should start. If you wanna start with AI, go ask Alexa to sing you a song. Just say, Alexa, sing me a song.

Bryan: 01:02:55

Uh-oh.

Justin: 01:02:55

Have you ever done that? It's I

Mario: 01:02:58

think my kids have.

Justin: 01:02:59

She does. Or yeah. She'll she actually has these songs that I don't know. I don't think they're generated on the spot. I think this is a library.

Justin: 01:03:07

I have no idea, but it's kind of fun. I'm I'm way off track, guys. I I think I think if we're I wish I could plug this whole thing into AI right now and ask it to summarize so I didn't have to do a summary and a sign off, but, we haven't gotten there yet. Actually, our viewers don't really even know if we're real. This could all be deep fake.

Justin: 01:03:28

This could all we we may not even be We're

Bryan: 01:03:30

in the matrix.

Justin: 01:03:34

I I'm just gonna go back to that, analogy I came up with on the spot of the and I, I really should say chop saw, which isn't the right word for it, but that's what I call it versus a, a handsaw. Like, I still have a bunch of handsaws in my garage, and if I ever have to use one of them, I swear like a sailor, I hate those stupid things, but they're relatively safe. I just can't get anything done with them. I love to pull up my power tools, and it scares the shit out of me, but I can get so much more done. And so this, I think that really is a good analogy for, you know, this is a tool that if if we're if you're a carpenter and you're still out there using a hand saw, you're not making any money at all.

Justin: 01:04:14

And in our businesses, if we don't start moving into some of the technology that's coming up, you're gonna be left behind on that same level. But it is really important that we put parameters around it and we learn about safety. Just learn. Just be aware. Understand what's going on.

Justin: 01:04:28

Understand what you need to do, what you can do, what you should do, and you've gotta train your employees the same way. So Yeah. Exactly. We're we're gonna move to just a sign off, guys. Just, you know, recap final thoughts.

Justin: 01:04:41

What what is it that you want, like, the one key takeaway to be? And remember that while we sit here and talk as a bunch of tech junkies, we're talking to, you know, picture your best client, the CEO of your best client, and give them a summary of what we talked about and what should be their key takeaway. And, I'll I let's go, Mario. What do you wanna go first for us?

Mario: 01:05:04

So a big key takeaway is that you have to educate your employees. Make sure it's it's a great analogy. I actually love that, Justin. I'm I'm probably gonna steal it. It really is like using power tools.

Mario: 01:05:18

You know, it's you'll still be able to get the same thing done mostly, you know, but we're using power tools, but used correctly, safely will help you get there quicker. So, you know, you have to, you know, proceed with caution. And at the same time too, you gotta fight fire with fire. You know, if if hackers are using AI, you need to implement AI to Yeah. You know, at least stay at a level level playing playing ground with them.

Mario: 01:05:49

You know, use AI against AI, you know, because it will learn, you know, and to detect and learn from other, you know, instances to how to better protect you.

Barinder: 01:06:02

So Okay.

Mario: 01:06:03

That that's what I would say.

Justin: 01:06:05

I like it. Brian, what do you got for us?

Bryan: 01:06:06

Alright. So my summary very quickly is, and you probably know where I'm gonna go with this already because I pretty much

Justin: 01:06:12

1% better.

Bryan: 01:06:13

Yeah. Exactly. So first actually, first is

Justin: 01:06:16

I love it.

Bryan: 01:06:16

Get an AI policy. Figure out that piece first.

Justin: 01:06:18

Yeah.

Bryan: 01:06:19

Put a policy in place so that you can you can start putting the framework in. And then, yeah, just look at at using AI as a journey. Get 1% better every day. Just, you know, iterate. Get a little bit better.

Bryan: 01:06:30

And before you know it, you'll you'll be you'll be at a point where you're 300 times better than you are today. Right? So it you don't you don't want this huge project to implement AI all over the place. You just wanna start testing and and and implementing it in small little areas here and there and everywhere safely, effectively, and, yeah. That's it.

Justin: 01:06:53

And I'm gonna sign off with I I love and hate to talk about policies and procedures. They're boring. This doesn't get very many people excited, but I really think that this is kinda where the rubber hit the road where where security's concerned at large. And and if you want a place where you can use AI effectively in your business, get those policies in place and upload them to your your whatever platform you're using, which is probably gonna be SharePoint. And then and then index it, you know, and and and let your employees say, hey.

Justin: 01:07:24

What's our policy on this? Right. Or what do I do in this case or that case? Because here's I talk about policies and procedures all the time, but the problem with them is they usually get written. Somebody signs off on them when they're created.

Justin: 01:07:38

The employee will sign off on them when they're under duress, which means they've just been hired and they have to sign everything if they ever wanna see a paycheck, and then they go on a shelf or they go on a file, and they're never looked at or talked about again. So, this is a great way to keep those right at our fingertips. Yeah. And and they're just so critical, so important to our businesses. Yeah.

Justin: 01:07:57

And with So that's what I've got.

Mario: 01:08:00

With AI being so big now, your employees will be engaged with them. They're gonna wanna read that. Yeah. They're gonna wanna actually ask questions. They're gonna want to maybe suggest other things that they, you know, to that policy.

Mario: 01:08:14

That's gonna be the one policy that actually your employees, I feel like actually won't read as long as it's done now, you know, during this big hype. I think that it it will actually intrigue a lot of the employees.

Justin: 01:08:28

Yeah. Alright, guys. We are going to wrap up this episode. We always have the ongoing offer. Go on our website, unhacked.live, and there's a tab there for a free assessment.

Justin: 01:08:41

Book an assessment with any of us, and not only will we evaluate you from a cybersecurity standpoint, we can also give you advice and kind of get you started down this path of using AI, not just to protect yourself because we definitely want that, but let's start using it to make some money. Right. I love that we get to talk about something that we can use for productivity to help the bottom line instead so that's what we've got for this week. Next week, we are gonna talk about the, let me pull it up so I can read the exact title. Cybersecurity and productivity on a budget, unconventional tricks for maximizing tech ROI.

Justin: 01:09:20

So, while we're on this theme of productivity and making money and increasing profits, let's talk about how we really can keep these costs under control because I don't know about you guys, but I this is a conversation I have over and over with my clients is, you know, this stuff's expensive. So we gotta get the most value out of it possible. So that's what we'll hit on next week. And with that, guys, we're gonna sign off. Thank you for joining us, and we will see you next time.

Justin: 01:09:49

Take care.

Mario: 01:09:50

Bye, guys.