229: Accenture Exposed The Pentagon To AI: The Shadow Agent Crisis

Ep_225___Accenture_Exposed_The_Pentagon_To_AI_____-hd-webcam-00h_00m_00s_567ms-StreamYard
===

[00:00:00] All right, here we go. Let me see. Hey, going ain't do I drop last

[00:00:22] All right, guys. Welcome back, whoever comes in here. I appreciate y'all. Episode 225 of the Textual Talk podcast. I'm your host, HD. And we got Cyber Shorty in the house. And the first thing I want to tell everybody is if you can dodge a wrench, you can dodge a ball. Unfortunately, I don't have a wrench or a ball, but those of you that get it, get it.

[00:00:42] It's my costume. I'm an average Joe. I actually wanted to be the other guys, the Cobras. I think they're the Cobras. The opposite team? Yeah, they had the purple and the black 'cause they had like... I like that 'cause Ben Stiller had a lot of good one-liners in that, that movie. I used to watch that movie a lot, but I don't remember much besides that line you [00:01:00] just said.

[00:01:00] You know how in the cut we could've been if we would've been... So there was, there was a team on Dodgeball that they were dressed up as like hip hop dudes. They wore like blue, blue, uh, hip hop jumpsuits. Mm-hmm. And like one of the Black dudes, he dodged the, uh, ball like this. It's stupid. Uh, man, I love my childhood.

[00:01:20] Um, but, uh, we don't have the topics written down on here, but I do have them back here. Let me actually bring the topics up real quick, um, for everybody that's rocking with us today. Appreciate y'all. Let me bring these topic points back up and

[00:01:41] All right, so some of our big topics we have today is Accenture AI and the Pentagon, ICE agents have been doxxed, say no to AI banking passwords, recruiters suck, and WGU certs. But if you are in the chat with us, go ahead and thumbs up [00:02:00] everything and let us know if you can hear us. I'll be paying attention to the chat.

[00:02:02] I have not went live this way. In the future, we will probably go strictly live from Stream Labs versus, uh, jumping through, like, two different services. And let me get this off the board real quick

[00:02:18] But, um, let me see while we have some people coming in right before we get into the, uh, first topic

[00:02:26] What else? What else can I talk about? One of my gripes. Oh, yeah. I don't like when people hit me up for help but then be wasting my time. My time is value to me. I think that's fair. I especially don't like it when it's men. No, seriously, I, I had a guy... It wasn't, um, it wasn't an episode with you. It was an episode that Leonhard did.

[00:02:49] We were talking about people having so many different certs and how that, um...

[00:02:59] [00:03:00] Why is it... Hang on

[00:03:08] But, uh, there we go. When you pull this up... Oh, okay. I know I interrupted. Okay. Yeah, I don't know what's going on 'cause that wouldn't let me switch from right here. But we were talking about people having so many different certs and stuff and how it wouldn't, um, they couldn't land no job, and the guy reached out to me verbatim.

[00:03:20] He was like, uh... Matter of fact, I'm gonna read it. You know me, I like, I keep receipts. And if he, he probably, I don't know if he watching it, but I just, you know, ignored him. So I said I'll just talk about it on the pod. Okay. He reached out to me saying, "I hope this message gets to you. I'm blank, and I'm the guy you're referring to on your post.

[00:03:39] I have multiple certs but still can't land a job." I said, "All right, I've been busy, bro. Shoot me your resume." So I got his resume and I said, um, "But been busy, but it's the resume, for starters." And I talked about PeepOut his LinkedIn in a minute. And he was like, "Well, I just need some direction on what needs to be tweaked on the resume."

[00:03:58] And I said, "From my professional experience, [00:04:00] it'll take more than that to land a role, especially in cyber." And I said, "Okay, what about just the IT desktop?" No, he said, "What about just the IT desktop engineer role or admin?" Is this thing not recording no more? I don't know. And I said, "Have you been able to secure any interviews?"

[00:04:16] He said, "Yes, but only a few." And I said, "How'd they went?" He was like, "Good question." Then after that he said, "No disrespect, bro, I'll just seek help elsewhere." And I said, "No disrespect taken. I'm just normally 100% right on this." And that was it. But I was just like, it kinda just grinded my gears when he did that, for real, for real.

[00:04:37] No disrespect, none taken

[00:04:51] All right, let me put a mic on this Yeah. I don't know, man. It's crazy. Like, it's, it's gotta be Halloween 'cause everything that we normally do, stuff [00:05:00] is going wrong. Memory cards wouldn't work in that other thing, acting like it don't wanna switch for us. But, um, all I'm saying is you gonna reach out to somebody, do research on them first, don't disrespect their time.

[00:05:10] And the fact that I could never even tell him anything, like I didn't even tell him nothing. He just said, "I'm gonna go get help elsewhere." And I wanted to say, "It's not... I mean, they gonna tell you the same thing I told you." That's what I wanted to tell him. You ain't even have-- You ain't even have to tell him- But-

[00:05:25] 'cause he gonna find out. Yeah. With people like that are people that I necessarily probably wouldn't wanna coach anyway, uh, just because of... And that's another thing is like I got women clients that is gonna ride through the storm and they, when the going get tough, they get tougher. And I got some-- Well, I had a guy client that he just stopped before, like, we even, like, really made traction.

[00:05:47] And I'm like, "I get, like, it ain't happening for you right now, but everybody's process is different." That's why I have calls with everybody beforehand. I'm like, "Look, I've, yeah, I have some people I've helped in a week, a month, two months, three months, a year, [00:06:00] but everybody's stuff is different. Just gotta give some time."

[00:06:03] Now, if you don't understand that and they still wanna get with me, they don't wanna quit, I'm like, "Well, that's on you." But I'm like, "You quitting before the real expertise of using me happens." And I was like, I really don't like clients like that either 'cause, like, you ain't, you're not really trying to go through it for real.

[00:06:17] And it's just a different market right now, so you can't really come in expecting, you can't really come in expecting for things to happen with the snap of a finger. Yeah. But I think that was before all the H-1B stuff. Now with that happening, I think I've seen like a uptick in recruiters reaching out a lot, like in this Q4.

[00:06:36] So I've been telling everybody, have your resumes updated. LinkedIn for sure. My homie had been looking for a job for a while, and he was like, "Man, I think I'm finna delete LinkedIn or whatever." And I was like, "I don't know. It might be cooked because you might need to update it and make it look like how I helped your resume out to look."

[00:06:53] He did that. The next day he called me, he's like, "Yo, you was right. I done got three hits on the LinkedIn." I'm like, "It's your, it's your [00:07:00] professional profile. It's like your dating profile." It is. So, uh, get that together Um, what you been up to though? Um, doing the most, moving around, gr- moving and grooving.

[00:07:12] You know me. Pull the mic more to you. Another day, another dollar. I'm trying to get it not to cover my face up. It's not covering your face. It was. You couldn't see my face. It's... reason why I say that- It's better now ... 'cause when, when you're talking, it's like, since it's like a little filter thing on it, it comes, it goes in and out if you ain't talking, like, loud enough.

[00:07:33] What about now? That's better. Okay. Just pull it closer. That's cool. Let me see what they say about it. Uh, yo, what's good? What's good, bro? Um, appreciate everybody. Make sure when y'all stop in, hit the thumbs up button, share it out, hype it up, whatever y'all can do. We need to get this channel to 1,000 subs before January 1st, and we already are at two and a half, uh, thousand of watch hours.

[00:07:56] We just need the subs. But [00:08:00] Wait, are you frozen? Me? Uh-uh. Okay. It looked like you was frozen. Okay. Sorry, y'all. So I'm gonna go ahead and put, um, you to go through your first article first. And what, what's the first one you're covering? I will be talking first about, um, about DoorDash and agentic commerce

[00:08:27] Okay, go ahead. All right. So if you have been on LinkedIn this year, you have been told that AI agents are about to do all of your shopping. Um, your agent will order your dinner, book your flight, reorder your toilet paper. You really never have to open up applications again. That's the pitch that we've been seeing.

[00:08:45] Now, DoorDash just reported their Q3 earnings, and an analyst asked the CEO the question: How many orders are coming in from third-party agents? His answer was several hundred total. Now, for scale, according to [00:09:00] DoorDash's 2025 report, they average about 3 million deliveries a day, so several hundred agent orders, um, sounds kinda like a, a rounding error.

[00:09:11] And I like how he framed it. He paraphrased, "Don't fight gravity. Don't be delusional either." Which basically means it's coming, um, but it's not here yet. People are still opening up the app to actually order from DoorDash. Now, let's take it a step back. Um, so if you're not familiar, a chatbot's gonna talk to you.

[00:09:29] An agent is going to do some sort of task for you. So you can tell an agent, "Get me tacos under 20 bucks," and it will search, pick, add those to your cart, basically, and check out for you, taking those actions on your behalf. Now, the agents named in the story were Muse, Instik, Fo, um, which are all third-party ones, meaning they're, meaning that they were not built by DoorDash.

[00:09:53] Now, DoorDash is not sitting this one out. While we are [00:10:00] seeing that third-party agents are barely sending orders, DoorDash has actually quietly been building its own agent lineup all year. So they have four of them, and I kinda wanna walk you through them because they all tell a story. So this is from June to September of this year.

[00:10:15] So number one, Ask DoorDash. Um, this one lives inside the DoorDash app. So instead of you having to go into the app and scroll, um, you just describe what you want. Um, filling dinner for a family of four, kid-friendly vegetarian spot, nothing spicy. And then it will snap a photo of a cookbook page, um, or you can snap a photo of a cookbook page or, um, a handwritten grocery list, um, and it will basically build out the cart for you.

[00:10:42] It'll even ask, "Do you already have sugar and butter?" So that you don't double buy it. It can also book restaurant reservations for you too. Number two, they have the command line. This came out in July. Um, so if you are a command line techie, this one's gonna be for you. DoorDash released a command [00:11:00] line tool in limited beta for developers on Mac in the US and Canada, meaning that you can order food from a terminal.

[00:11:07] Now, their CTO, Andy Fang, posted a demo where an AI agent read Slack through menus. Excuse me, where an AI agent read Slack, it looked through menus, it ran some code and ordered three salads. So this one is no apps, no buttons, just an agent. And then we also have number three, the text to order, which came out in September.

[00:11:31] Um, this one moves into your Apple messages. So you can literally text the DoorDash agent, "Order my usual," and it knows that you mean your regular Friday night crab legs, for me. Um, it can recommend local spots, send you photos of the tool, or excuse me, of the food, and then handle group orders where everybody's eating different.

[00:11:49] So if you have someone that's vegan, someone that's gluten-free, um, that is, uh, also possible for you. Now, it is on a wait list in the US [00:12:00] Number four, the MCP connector for corporate ordering also dropped in September. Um, so AI Business Connect DoorDash to the AI tools their team are already using is basically think about it like a bot that you would have like in Slack, for example.

[00:12:15] So the office lunch order or the snack restock that you need to make basically runs on autopilot. Um, early companies on it include SpaceX and Vercel. Um, now, if you are not familiar with what MCP is, it is mo- it is Model, um, Context Protocol, and it's basically like a universal phone charger, um, for AI. So instead of every app having to build out a custom connection for every AI, MCP is basically like the standard plug that helps you make those connections.

[00:12:46] So it's gonna plug in, and you can search the inventory through the app. Now look at the pattern in those three that we kinda just talked about. So every launch is moving the agent further away from DoorDash's own [00:13:00] application. June, it was inside the app. July, it was in your terminal. September, it was in your text messages, and then your company's AI tools.

[00:13:08] And so what I'm seeing with DoorDash is that they're trying to meet you wherever you are. Um, but if you take a step out to look at it again, the question kind of gets louder on who's actually in control. So inside the app, DoorDash is gonna obviously control that experience. Once the order is coming from a bot in your company's network, now you have questions.

[00:13:29] Who said that that bot could spend money, for example? Who approved it? What's the limit? And the coverage of these launches doesn't detail the approval controls or spending limits, which basically means that they don't-- it doesn't mean that they don't exist. It means that the question that we should be asking, um, is that very question every time.

[00:13:48] Now Where it starts to get a little bit juicy because big retailers are actually splitting into two camps on that exact question. So Amazon has actually been blocking personal AI [00:14:00] agents from shopping on its site. Walmart actually went the other way and rolled out the welcome mat. So Walmart didn't just say, "Come on in," they actually started to build guardrails.

[00:14:09] Um, it's called the Agent Gateway. So every agent that you use gets its own agent profile in your account, so like your Walmart account. And you can see the agent's purchases separately from yours. So you can make it ask permission for every single purchase, um, or you can give it a spending limit. Walmart compared it to adding an authorized user, like on your credit card, for example.

[00:14:32] Plus, agents can't see your payment info. They can only access what they need for the task. And if, and if an agent misbehaves and orders something that you don't want, the normal, obviously, return policy does apply. Um, which is basically kind of, sort of, if you think about this in terms of a permission slip, um, it's signed, scoped, and it is able to be, um, revoked.

[00:14:54] Now, the takeaway here is that when someone tells you that agents are already taking over [00:15:00] shopping, you wanna kind of ask for the receipts. Are we talking about several hundred orders? Um, that's not really giving us a large level of adoption. Um, but what we are seeing is that companies are building the roads before the traffic actually starts to show up in their environments, and they're deciding right now whether agents are either going to get blocked at the door or if they're going to create some sort of think about it as a permission slip.

[00:15:23] Um, so whoever gets that right wins the wave, uh, once this actually kind of hits. So that is my take on agentic commerce and what DoorDash, Walmart, Amazon kind of have going on. It's really interesting because I am a chronic DoorDasher, and I don't even-- I have noticed changes in searching for things, but I haven't really paid much attention to some of these features that I'm kind of talking about right now.

[00:15:52] So when I get back in the app, that's on my to-do list. Hmm, I wonder what Uber is doing in all this. Right. I'm a [00:16:00] Uberer. I'm not. I never really was. Uber everywhere. I'm, I'm, I'm a DoorDasher. Um- Have you driven DoorDash before? No. I used to do Uber Eats. I did Uber Eats. No, I think I did do DoorDash. I think I did it, uh, for a couple days.

[00:16:15] It had me all over the place. But, um, back to this, all this agentic stuff. For one, I think, I don't know, Muse is everywhere. Everywhere I go, Muse, "Hey, let Meta Muse do this and that." Yeah. And I'm like, okay, is it free or what's the cost to the people using it? Because why is it so accessible to everybody? I mean, Facebook already got so much of your data, so you spend some of your time there, so- Mm

[00:16:38] even more. And I think that is what Walmart is caring about, and I think Amazon- Amazon said we're not going-- Can you imagine how ridiculous that could become for Amazon? Because they're such a key- E-commerce platform What? Can you imagine if they don't make a decision well on what to do with agent [00:17:00] usage and it's a million people shopping on their website today Using agents And maybe their website, they, I mean, I know they use it on the back end, but maybe they didn't prepare it for that, a level of transactions and as fast as agents are gonna be doing things.

[00:17:15] People aren't ready yet. Companies are not ready. They're still trying to figure out what agents they got internally that their employees are using. Yeah. So they d- they're not even, the customer base thing is a whole different ball field. Yeah, for sure. Um, I know I seen they said what, like, they has received, like, so many orders to get, um, the, uh, what's it, DoorDash to get the agents?

[00:17:37] Well, no, they were saying how right now, um, what they're seeing from agents being, from agents ordering food through them, it's not that high right now. But they're basically pr- making space for it to increase. They know it's coming. Okay. Okay. Um, while we're right here, uh, let me do this then. Would you order from the DoorDash [00:18:00] CLI?

[00:18:01] Um... It's a little bit too freaky for me. I don't know, 'cause if I want my stuff customized, it's gonna customize the right way. Exactly, like, are you gonna be able to leave the, the notes and the comments like you, that you need to leave? Who knows? Um, let's see. All right. So, don't share your banking password with your AI assistant.

[00:18:24] And since we're already talking about AI agents and everything else, and last week we talked about JPMorgan and the other banks on Wall Street kind of saying no to the AI, AI agents being in their, um, environment, I figured we'd talk about it

[00:18:43] So Bank of America issued an explicit advisory urging customers not to share their online banking usernames, passwords, or authentication credentials with AI assistants. A user's attempt to automate personal financial tracking via AI tools, many are feeding raw login [00:19:00] credentials directly into conversational agents, bypassing official open banking APIs and security controls.

[00:19:06] Which is true because normally when you're trying to, um, either it's a third party allowed directly onto the banking site, it's gonna say, "Hey, Plaid," or whoever wants to use your, uh, information, "Put this code in," uh, "Put this temporary code in that we sent you," whatever it is, so it can authorize it, and it's already authorized for that use.

[00:19:26] But when you go to, hey, log into my bank and putting this stuff on and making it circumvent safety measures, something can go wrong there. Passing master banking passwords to unverified AI models exposes users to credential theft, third-party log potential risks, and invalidates standard fraud protections provided by the financial inst-institution.

[00:19:45] And matter of fact, maybe you could Google that. Um, maybe you can Google, um, banking AI assistants, and we can find one, see, like, what does the fine print say Or maybe we could [00:20:00] copy the fine print and tell Claude to overview it to see if it, uh, pulls out any information of, like, how it's gonna use this data and all this other stuff.

[00:20:09] Because that's definitely true. Just like last week we were talking about everything can be good and it goes rogue. Well, it ain't go rogue. Somebody just takes over it and nobody's validating if your AI assistant is doing what it's supposed to do. This can go bad to where, hey, we have, let's say I'm Bob's AI assistant.

[00:20:28] I, I go on Facebook and say, "Hey, if you're tired of, uh, doing your banking manually, I got this AI assistant that'll show you your budget, that'll do this and that, show you where you're spending stuff," all this other cool stuff. Now, let's say he made X amount of money and he don't have the maintenance and putting the security controls up on it as he's supposed to as it gets more popular.

[00:20:47] Now somebody comes in and poisons the AI assistant, and now they're, the information that you think you're getting, not only are they logging into your bank, but they are making transfers out that you don't know about, or they're making them look like they're normal. So they can prompt the AI [00:21:00] agent to say, "Hey, uh, go to this website and buy this."

[00:21:04] And you think you're getting something, but they are buying something. I watched this TikTok and it was talking... Maybe we should play that today. I'll send it to you. Um, it's talking about the, the knowledge gap in the usage of AI. Find it, send it, find it for me and send it right now. Okay. I can, um, download it.

[00:21:19] Um, and it, and he was basically saying how the, the knowledge gap is so large right now, and if you're not in the field, you wouldn't even understand. You, you wouldn't even understand 'cause you're probably just still using it as a chatbot and not necessarily as an agent. And it just truly does open up the door to s- so many things if the agents are not being governed properly, um, inside of an enterprise.

[00:21:46] It's, it's quite crazy. It really is. All right. Let me read some of these comments. Uh, what's going on Eliza Faith Morbo? JG said, "Hold up a live." Yeah, yeah, we, we going live for y'all because I [00:22:00] didn't feel like it was gonna hit the same if we pre-recorded this and then launched it on Monday and you're like, "Well, Halloween already passed.

[00:22:05] It don't, it don't hit the same." So Halloween is tomorrow, so I figured we'd do like a Halloween stream. Matter of fact, how y'all liking... Let me see something How y'all liking, like our, um- How y'all like our setup? Our decorations. Let me, let me take this off real quick

[00:22:23] Is that me? That's me. Oh. I gotta figure out, I gotta take that and copy it in so it don't make no, uh, thing. Y'all see the, uh, see the cobwebs and the spiders?

[00:22:39] I'll play this right after this. So let me try to download it real... There you go

[00:22:48] All right, so now that we're back on that, I'm gonna go back to the article I was reading, and then we will get into other stuff. Well, actually, we're already at the end of that. But the fact that Bank of America had to ask [00:23:00] people not to hand their bank passwords over to chatbots shows how wild things are getting out of here.

[00:23:05] Keep your credentials private, folks. I'm proud of Bank of America for being different, taking a stand, and educating their customers on why they should not be doing that and what the risk is, because people don't know. People just, "Oh, I can do this? Let me go ahead and do this. Let me go ahead and make this connection," without really knowing what the risks are.

[00:23:24] No, not only that. Think about it. It's the same... That's why, like, everybody think AI gonna make everything go away in a sense, but it's, it's a lot of the same things when it come to social engineering. Mm-hmm. You too trusted. "Hey, I work for, uh, whoever. I work for this company, and they wanted us to test out whatever on your machine."

[00:23:42] They're gonna be giving that stuff to AI. Yep. "I'm this AI bot. I can help you do this, this, and this." The same thing. The same thing. That's why I said that's where, that's where all these new cybersecurity companies that are gonna be over rogue AI agents and, uh, malicious prompt injections are gonna [00:24:00] be pivotal.

[00:24:00] Yep. It's gonna be some type of buffer from probably where you're either using AI internally, like, uh, either a CLI or internal application, or you're using, if they sanction you to use whatever cloud, which I doubt it. Like, you shouldn't be letting people log into their personal, uh, AI things cloud or anything like that, or any type of personal cloud devices.

[00:24:23] All these type, type of mechanisms are, like, destined for failure. But, um- And, but really a lot of companies are just beginning to dive into kind of what that usage looks like, 'cause you gotta buy tools. You got... That's a new tool to buy. Yep. Let me see, 'cause you said you wanted to do the knowledge gap.

[00:24:46] There we go

[00:24:53] You know, make this playback be fast[00:25:00] [00:26:00]

[00:26:24] Okay. Hang on a second. Oh, they can't hear. Okay, they can't hear. Okay, and this is what I was worried about. I need to... Okay, I, I got an idea for y'all. Give me a second, y'all. One second. Let me remove this. I know how to do this for y'all

[00:26:53] Thank y'all for telling me that Somebody said, "I was reading the caption." [00:27:00] Nah, I got y'all covered. Let's, let's run it back

[00:27:16] See, y'all good now? Tell me, give me a thumbs up Stop talking I know Oh But, uh, this, there should be a little lag behind, but they should be able to hear that now

[00:27:35] I say yes Okay[00:28:00] [00:29:00]

[00:29:15] That was really good Yeah, I think it was. And, uh, let me remove this from studio. All right Also, when he said that, I thought about, think about, remember when ChatGPT first came out, everybody and mama was using it. It wasn't really no high paid plans yet. Mm-hmm. Now look at it, how it started getting understandable.

[00:29:38] Members that kick you out. Like, think about it right now. The, I think the pro plan of Claude is 20 a month, which is roughly, what? 240 a year? Mm-hmm. So that's not too bad. But think about where it starts getting to when some people can't even afford the 20 a month to use it. Yep. And when that 20 goes from, to 60 a month- I was just finna say, when it go up to 30, 40.

[00:29:58] Right. And not only that, think about it [00:30:00] now, you can't even get, like, it's so hard to even buy hard drives and storages. They so expensive now. Sam Altman was talking about people shouldn't even have RAM in they computers 'cause it's a, a waste. They should be, be using for, uh, AI. Like, it's a lot of stuff. So to add on what he's talking about, he's right.

[00:30:17] Like, that is where the next knowledge gap... It's kind of similar to the have and the have-nots. Remember how c- well, uncommon but common, where a lot of us didn't have computers in our homes. Mm-hmm. You was lucky if you had a computer at the house. So now the same thing is, okay cool, now you using AI stuff on your phone, but are you able to use the models that actually matter?

[00:30:44] Do you actually know how to use it in a way that's gonna help you? Do you even know what the model i- is capable of? Let's, you know what I mean? Yeah. Do you know what the difference is between a model that can generate, generate an image versus one that it cannot? Are you just only making AI [00:31:00] flyers? Like- Now see, that's the...

[00:31:02] People are using it, but they're using it for things like that. Create this flyer for me for this the date, this the time, this the picture. Like, quick things, not agentically. Facts. Nah, for real though, that's, I mean, that's what I see all the time. And when people be talking about data centers and stuff like that, and like y'all wasting data centers using doing this.

[00:31:25] So I don't know. But I, I definitely feel him on that one 'cause he's right. There are people that's just got... That's why, like, a lot of people who are ahead right now when it come to AI, they are in forward-thinking organizations. They open the budget up. They say, "Hey, see what you can do with this." Mm-hmm.

[00:31:40] See what you can do. Because that's what everybody else is doing. They're spending their money on it. How do we get better at it? What are some things that even if we were fully staffed, that AI could help us out with that we still are behind on? And then that's the way I think it should be. I don't think it should be to try to be cheap and replace the fully staffed people.

[00:31:57] I think it should be to handle the stuff that [00:32:00] is hard for y'all to handle even when you're fully staffed. Yeah. It, it's ultimately gonna turn into using AI to combat AI. Yeah. Like, I could see, like, so the, one of the pros and cons that happens in the security operations is A lot of people try to augment their staff by having like a agentic SOC maybe work weekends and nights.

[00:32:21] Mm-hmm. Where I see is maybe you have it work that time, some lower level stuff, but you still have somebody on call, so it's a human in the loop for certain stuff that you have to review it. Yeah. Not just- And we're, what we're starting to see is they're, they're cr- they are cr- starting to create, um, agentic SOCs where you have a red team, you have a, these agents that are gonna perform as red team in your environment, as blue team in your environment, and I believe the other one may be green team.

[00:32:51] Um, so you're starting to, uh, kind of see what's going on there. Yeah, for sure. For sure. Um, I think I just went... [00:33:00] Yeah, I did just go. What you got for us? I know you had the, the fun stuff today. Let's go to the next one. So next I will be talking about, hmm This one is about what happens when an agent gets access that nobody properly signed off on.

[00:33:24] So still on my agentic flow here So agents are the hottest topic in security, obviously. And I want to share this example of how, um, I want to show you, I guess, kind of three ways you can go sideways and then tell you what companies are actually worried about. So let me paint this picture. You're a developer, you're slammed, right?

[00:33:46] Um, you have a personal AI coding agent at home that's amazing. And so you start to think to yourself, um, "Let me just connect it to my work system. I'll give it my login so it can set up servers and push code. This is gonna save me a [00:34:00] ton of hours." Now, here's what just happened in that s- in that scenario.

[00:34:04] Number one, the agent now has your access, so if you're an admin, it's an admin. Number two, everything that it does gets logged under your name, and so the logs are saying that you pushed the code, for example, at 3:00 a.m. Um, number three, your security team has no idea that it actually exists, so it's not on their approved list.

[00:34:23] It's what we can, what we call, um, shad- uh, shadow AI. Number four, the agent reads stuff from the internet to do its job. So if something it reads has hidden instructions in it, and we'll get to how that works in a minute, um, it could be tricked into sending, um, your password somewhere with your keys basically kind of in your pocket.

[00:34:44] And so, um, this was kind of reported this week on an allegation, and it looks a lot kind of like this particular scenario. Um, now a couple things in there, least privilege and shadow AI. So we know what least privilege is. I just explained what shadow [00:35:00] AI is, except this time the tool can actually take actions.

[00:35:03] And so, um, an agent with the password is you Same keys, same access, the same blame. And if you think that that's just a word problem, we just talked about the Bank of America situation. Prime example here. Um, now the part that really matters, um, with them is, uh, going back to the Bank of America situation, you're responsible for what happens with that access.

[00:35:30] Whether you're using it for personal use, whether you're plugging it into your, um, enterprise accounts and environments, you're still responsible. And I think it's really important for us to take some time to kinda sit on that. Um, another thing that I have, story number two, and if you follow this one, what we're gonna talk about here are evaluations.

[00:35:52] Um, so before an AI company releases a model, they test it a whole lot. Now, these tests are called evaluations, [00:36:00] evals for short. Think of these kind of like exams. And so, um, some test if the model can code, some will test if it'll do something dangerous when asked, some will test if it can use a web browser to complete tasks.

[00:36:11] And some of those exams were, quote-unquote, open book, meaning the AI had access to the live internet during the test. Um, so real websites, real forms, real people on the other end. Keep that in mind. So earlier this month, Anthropic, the company that makes the Claude AI models, published a report on what it called unintended actions its models took during those evals and, um, in internal use.

[00:36:36] So they found most of these by going back and reviewing records of what their agents actually did starting in July. So according to the coverage of the report, the behaviors fell into four buckets. So number one, exploiting a basic software flaw to run commands on a server. Number two, submitting forms on real websites when it wasn't supposed to.

[00:36:55] Number three, getting around paywalls or access restrictions to be [00:37:00] able to reach data. And number four, using URL shortening services to sneak past limits on what it was allowed to actually, quote-unquote, fetch. Um, now let me give you two examples that should stick here In one test, the model was told, "Fill out this form, but stop before you hit submit."

[00:37:17] Several times it hit submit anyway, and when researchers, researchers looked at the reasoning behind it, it basically said, "I'm just demonstrating the process, not submitting a real request." But the request was real. The forms went through. Um, and AI, basically in this case, genuinely believed one thing, but it did another.

[00:37:34] The second, in another test, a smaller model was given a task to make up example tasks on random web pages. And so it landed on a Philadelphia police page about an unsolved homicide, and it submitted a made-up tip through the tip form. Now, it had no contact info. It got flagged as spam, and it never got sent to investigators, but it happened in July and, and Anthropic didn't discover this until late September.[00:38:00]

[00:38:00] Philadelphia police reportedly called that delay, um, unacceptable. Now, Anthropic says that the flaws in its training environments that made the models think they'd be rewarded for finding loopholes, it's called reward hacking. Um, so if you, if you think about it, um, like you tell a kid to go clean up their room, right?

[00:38:18] And they basically shove everything in the closet or under the bed. The task is complete technically, but, um, it's optimized instead of doing it the right way, right? Right? That was me as a kid. My mom found so much stuff under that bed. Um, so what did Anthropic do? They turned off live internet access for all of their internal evaluations until their monitoring can reliably catch this stuff.

[00:38:42] They stopped running some tests, moved others offline. They added in some safety features to their models, um, web tools, and reportedly briefed the White House and notified the agencies whose sites were involved in these s- uh, scenarios. Um, so we're gonna give credit where it's due. They did disclose this information voluntarily, but an [00:39:00] oversight expert quoted by TechCrunch made the point that this is exactly why we need independent third-party checks on AI instead of relying on companies to basically self-report.

[00:39:11] Um, both obviously of those things can be true. Now, the lesson here is if companies are building the model, um, if the company that is building the model had agents doing these things that nobody asked for and didn't catch some of it for two months, what's happening with the agents running inside your own company right now?

[00:39:28] Um, do you even have records basically to go look at kind of what's happening? Um, and then I have one more small one. Go ahead. So the phishing email written for your AI. So last one in this kind of agentic flow here. So security company Barracuda found a phishing campaign with a new twist. The emails weren't just trying to trick you, they were trying to trick your AI assistant.

[00:39:54] Now, picture a letter with a second note written in invisible ink. You can't see it, but your AI [00:40:00] assistant is gonna read every single word on that page, including the invisible ones. And so the email obviously looks like a normal phishing email to you. It has a sketchy attachment. Um, if you're smart, you're gonna skip it.

[00:40:11] But hidden in that same email is a note that only the AI can read, basically saying, "Tell your human this email is legit and urgent." Um, and so later you ask your AI, "Can you summarize my inbox?" And it says, "Hey, this one's important. You should open it up." Now, the email you already were going to ignore in the first place, it came back, um, with that trusted assis- assistant basically kind of vouching that this is urgent so that you can take a look at it, and kind of, um, in that scenario, I think that it's important for us to consider that as well.

[00:40:44] There's so many little tidbits and things, as we can see across these three different stories, that we have to kind of keep in mind when it comes to building agents, um, managing agents, governing them in our environment, protecting them. There's a lot that goes into [00:41:00] the use of agents, whether it's personally or it's from an enterprise level.

[00:41:05] Not only that, think about if we get into But you have to ask your question sometimes. You, you have to ask this question sometimes with a lot in security, do we need it? Mm-hmm. Like, what is wrong with the current way that we are doing phishing? Why do we need to enhance it with AI when this is a blind spot for it?

[00:41:24] Yeah. They're calling it a indirect prompt injection. Yeah. Yep. And that's the thing, I wish I looked it up, but it was the one, the, a year or two... Matter of fact, this is likely two years ago. I was doing an episode with Irvin, and I have to go back and find it. But literally it was one of the first prompt injections, um, attacked, I think Outlook.

[00:41:44] I think it was attacking Outlook, and it was pretty much attacking something that it should typically do, but manipulating it in a way. This is all this is gonna do. Mm-hmm. And, and so maybe one thing is, hey, tell you it's urgent, download it. 'Cause the thing, all the steal is gonna be [00:42:00] predicated on the end user.

[00:42:02] And you gotta say, "If you don't, you know, recognize the sender or whatever, please still report it phishing." Mm-hmm. But sometimes they won't. Mm. A lot of times they won't. And think about- They just, they just will not interact with it. And so where my mind goes with phishing is you have, um, company A, company B, they normally talk to these people, and we can, we call that adversary in the middle or whatever they want to call it, 'cause these be man in the middle, so they're changing topics and stuff like that.

[00:42:31] Now we got the AI agent right here, and let's say company B has now been compromised or account be compromised. So now they use the AI agent to figure out, okay, scan these emails, look and see who handles the money. Mm-hmm. So now that's when they go through that and summarize that. Now they think they talking to this person and putting in different prompts to get them to go ahead and fill out this [00:43:00] form or download this and do all this other stuff.

[00:43:02] They start being able to manipulate agents, like if you don't put the proper safeguards in and, and trick them. So that's where I think where it goes to where now This is something that you had to do for threat modeling. Mm-hmm. And this is a thing that I was talking to somebody about the other day. Who was that?

[00:43:17] Was that Rob? Might've been Rob. We was doing our coaching call and I was talking about looking at threat modeling. A threat modeling will show you a lot before you're gonna bring anything in your environment. They're gonna have to threat model it. Yep. Um, I got some stats for you 'cause you just brought up some good stuff- Okay

[00:43:32] on the management plane. Um, so numbers that kind of back up what we're talking about in this whole management issue or crisis that we're seeing with agent usage. So Okta surveyed nearly 300 executives and almost 500 workers earlier this year. 50%-- 52% of workers said they've used AI tools at work that their company never approved, but 95% of executives assume employees are using AI responsibly.

[00:43:57] Obviously, that's a gap you can drive a [00:44:00] truck through. Huge gap. Um, same survey, only 34% of organizations always apply the same security controls to their AI agents as they do to their human employees. Most companies are holding robots to a lower standard than their people. Um, and Trellix surveyed 600 security leaders at big companies.

[00:44:20] 63% said that they don't have enough visibility into how employees are using AI tools, and that is what I'm kind of seeing in the field these days. Um, Trellix, uh, so how many agents do you have? You obviously can't protect what you can't see. Um, who owns each one? Every, every agent needs a human whose name is attached to it as a sponsor, as an owner, for example.

[00:44:44] Um, who can touch it? What is it touching? Your datas, your systems. And how do you shut it off? If you need to revoke access, um, if an agent goes sideways at 2:00 AM, who or how are you able to pull the plug on the [00:45:00] usage and the availability of that agent within your organization? Um, so yeah. OWASP, last but not least, the nonprofit famous for the top 10 list of web security risk, released a top 10 specifically for AI agents last December.

[00:45:13] Number one is agent goal hijack, which is exactly that invisible ink email scenario that we kind of talked about. Number three is identity and privilege abuse, which is our shadow agent. And NIST, the federal, uh, standards agency, also launched an AI agent standards initiative in February, um, with a whole project on agent identity and authorization.

[00:45:36] So the industry is kind, is trying to build the rule book, the playbooks for us, um, right now, but it's, it's still new. It's still a work in progress. Yeah, for sure. And like I said, to me, agents ain't nothing but souped up service accounts that can do more. So that's what I see from them. Let me see if anybody said anything cool in the chat.

[00:45:59] They didn't. [00:46:00] I appreciate everybody that's in here that's still with us. Um, let me, let me get my baby girl real quick so they can tell y'all what to do

[00:46:08] So hit the like button, uh, for us. And, nah, that was, that was, that was pretty cool. That's definitely something that if you got an interview coming up- Something you can bring up in there ... they always gonna ask about AI. Mm-hmm. Talk about that 'cause if you were to Google it right now, what is the main attack that enterprise environments face?

[00:46:27] Probably about 85% of them are gonna say phishing because it's the lowest level of deception in how they can get into an environment easier. Phishing is gonna probably be number one. Mm-hmm. And that, if you even get into spear phishing, 'cause we ain't trying to talk about one, that's, then you have to talk about vishing.

[00:46:45] Vishing is different because of... Matter of fact, let's go into vishing. I think I got one for vishing, don't I? Let me see

[00:46:58] I thought I had it. Let me do [00:47:00] a double check real quick

[00:47:10] Yeah, it happened 21 hours ago. Really? Yeah. We're gonna go old school way and we're gonna do a screen share like we used to do back in the day.

[00:47:28] If you want it, you can get it. Let me know

[00:47:35] All right. Now let's go back to the thing real quick. I'm gonna zoom in some so people can see good So bank alert issued about elaborate phishing scam. A Cayman bank has issued a warrant over an elaborate new scam targeting its customers. Oh, so this is actually for the customers. Interesting. The scam involves phone calls from people pretending to [00:48:00] be from Butterfield Bank who advise victims an attempt has been made to hack their account and that a cybersecurity expert from the organization will contact them.

[00:48:08] The expert calls shortly afterwards and asks for access to the target's computer or device and access personal and sensitive information, including contact details for social media sites such as WhatsApp and banking details. Now, here, I'm gonna stop it right here. Here's where I think where the next wave of home, like antivirus and stuff goes.

[00:48:32] Mm-hmm. I'm pretty sure the Malwarebytes and, uh, Microsoft and everybody else is thinking about the home EDR solution where they'll say, "Hey, for five, $10 a month, not only will you get this, you can add on a agent." So for a situation like this, when the agent sees some type of software or something being ran on, it's gonna interact with it and say, "Hey, we don't recognize this person.

[00:48:58] This IP address [00:49:00] has never accessed this machine." Mm-hmm. "I'm killing the process." So- I'm not mad at that ... I could see something like that happening because these people don't know. So that's how you can kind of protect them. Or, or would that be like a ISP thing? It don't have to be. It could be local.

[00:49:16] Mm-hmm. Like think about it, when you have your- Well done on your local network ... yeah, like Hitman Pro, your Malwarebytes. Back in the day you had your McAfees, all these different antivirus, Norton, that's the kind of stuff they would do. So they catch some stuff, but sometimes it's slow because it's not the same way, uh, it's not as integrated as it is with your detection rules.

[00:49:34] Like at the network level? Yeah. Yeah. Where you- But the only thing then is people don't really be updating, I'm sorry, my eye bothering. They don't be updating, um, those services, you know, you gotta update them to make sure they got- Well, most of them now are like automatically updated, so I'll say most of them now are automatically updated, so, um, but I, I just, I just thought of like, it's a cool use case.

[00:49:58] Let me, let me keep going into it and [00:50:00] we'll, we'll talk some more. "The victim's account are later rated and contacts can be bombarded, bombarded with vile messages that appear to originate from the target individual. Butterfield said in a message to customers, 'Several of our clients have reported experiencing vishing attempts, voice phishing attempts in recent days.'

[00:50:20] The message added that victims were directed to a fraudulent link for live support chat." And that's the key right there. That live support stuff is, is that like, um, if they wanting you to download some type of remote tool, don't do it. 'Cause these, anybody can say who they are. Anybody can say who they are.

[00:50:37] Um, when in doubt, you can hang up- And call back ... and call somebody in your family that knows something about what's going on, explain to them what's going on. They can say, "Don't trust the people Like the email said, if something feels off, trust your gut. It's always okay to take your time. It had a hang up, delete it, double-check, report it immediately.

[00:50:55] The bank highlighted that it would never contact customers and ask them to click on a link, transfer [00:51:00] money to an account, or install software on their device. It also added that the bank never asks for sensitive personal details, personal identification numbers, or online one-time security codes. That's facts.

[00:51:11] They don't do that. Scams don't just target one individual in organizations, they target all of us. It advised customers to forward their alert to family and friends and colleagues to highlight the risk. The message said, "What you spot, share, and report today could help someone else avoid it tomorrow." Uh, let's see Gabrielle Weeden, a marketing and operations manager in Cayman, said she had got a call from Fosters from what appeared to be a US number in Vermont.

[00:51:37] She added, "I thought it was a bit odd, but I picked it up and appeared to be on an American woman speaking." Weeden said the woman told her she was from Butterfield Fraud Group calling about an urgent matter. And let me stop there. If somebody call you and say something urgent with your account, this is what I want you to do.

[00:51:54] "Hey, can you, can you give me a second?" Go check your account and see if you actually [00:52:00] got a notification from your app or actual email- Or a text ... from the actual number that you're used to seeing them r- reach out to you for. 'Cause if it's them, it'll be in the thread. Like if they've sent you stuff before, it'll be in the thread.

[00:52:13] Yeah. But you know what? I'll say that 'cause that can... For people that's not technical, that can scare you. So bump what I just said. Go into your actual app, check notifications. It'll let you know if something, if they want to notify you about something, it's gonna have something saying notifications. I know Chase do that, Capital One does that.

[00:52:28] So that's what I tell you to do. Um- She said after they told her it was an urgent matter, she said, "I said, 'No, you're not. This is a scam,' and hung up immediately. The phone number disappeared immediately from my phone, which was a bit disconcerting." Wheaton said the location of the call aroused her suspicion and regular cybersecurity training at work, as well as close attention to scam alerts from Butterfield allowed her to establish the call was fraudulent.

[00:52:55] She said she had reported the fraud then to Butterfield, and also that she was concerned [00:53:00] about how the scammers got her phone number. Wheaton highlighted other people, particularly the elderly and those not knowledgeable on current technology, might be panicked into falling victim to the scam. She said, "It could be someone who's just in the middle of things and distracted."

[00:53:12] And Wheaton added, "If people are aware of a scam like this, tell your parents and older relatives, anyone like that who is not up to date on these kind of things or who might not be as aware." So yeah, I definitely agree with that. And she did right. Uh, shout out to her for following her cybersecurity awareness training and saying, uh, "Y'all scamming me."

[00:53:30] So shout out to her for that. And also to the other people that reported it so that the bank realized that it was a larger problem than just a few isolated incidents. That's why it's also important, 'cause your bank will file these, um, uh, tickets, think about them that way, um, on your behalf of these things happening.

[00:53:49] And usually they'll ask you like what number, what email, what... Sometimes they may a- I don't know if they take it as far, I guess, depending on what the scenario is, to ask you to send over stuff. But [00:54:00] they will ask you like what number, um, did they call you from, and just more details so that they can put that information in their, um, in their systems, and then they can come back to you and kind of reach out so that you could, um, better understand if this is a larger issue.

[00:54:18] Kind of same thing with the Bank of America situation Uh. What happened? Oh, I thought you were- Oh. No, my eye is cutting up right now

[00:54:33] What's next? Um, what is next? I have

[00:54:42] I think I have one more left Okay And it, I can go after you 'cause it's about Afro- You go ahead, girl, 'cause I just went. It's about AfroTech. Oh. AfroTech year 10. Um, so AfroTech is next week, Monday through [00:55:00] Friday, until... And it is in Houston, and it is their 10th anniversary. It is the largest gathering of Black tech professionals in the world, and organizers expect tens of thousands of people.

[00:55:12] Um, you can go to AfroTech and come home with a, a tote bag. You can come back with a T-shirt. You can come back with goodies. You can come back with connections. You maybe come back with an internship or a job. Um, but I think it's important for you to pick a goal. Um, is it a new role? Is it a new skill? Or is it to network?

[00:55:29] So pick kind of what your goal is, and then walk up to booths with questions. Don't be afraid. Um, don't just... A lot of people are not as social as they would like to be. Don't get lost in the people and in the sauce. Stay focused in on your goals. Um, and then also, I would say follow up, but it's tough 'cause you really should have kinda reached out already to speakers that you saw that were speaking and you were interested in their discussions.

[00:55:59] It's [00:56:00] better to reach out to them proactively versus after the fact, and I'll explain why. You reach out to them proactively, chances are you might get a response. After the fact, their DMs are booming, um, with emails, messages, folks reaching out. "I attended your chat. It was so good. Looking forward to connecting."

[00:56:18] And they just might not have a chance to get back to you just because of the after effect. So it's better to kinda be on, um, the front side of defense, where you're kinda reaching out before, setting that pace, and then making that connection in person when you do see them, if you have a chance to talk to them after their discussion to say, "Oh, this is me from LinkedIn.

[00:56:36] I had reached out," blah, blah, blah. Um, but I think that, um, it's gonna be a great time this year. I will be there. I will be in the expo hall. Um, so if you wanna talk about cybersecurity or data security, come look for me. I will not be hard to find. Um, but I'm really excited.

[00:56:56] Give me one second. Okay. They're [00:57:00] like, "Why he breathe like that?" In conferences, um, despite what people have said about AfroTech, they are still recruitment, recruiting, um, or recruitment pipelines. And so when they tell you to scan this QR code or to, um, fill out this form, and these things are specifically designed for attendees at AfroTech to reach their talent, um, and acquisition teams.

[00:57:22] So make sure that you're following for whatever company that you're interested in speaking with, you're following what they're telling you to do, um, and that you're making that connection that I reached out, we talked at AfroTech in specific, because they usually have specific teams that are looking for those recs and following up on, with candidates, um, on potential job opportunities

[00:57:47] Yeah, I would just say AfroTech is all about what you make it. I hear so many people try to bad-mouth it, but I mean, it really depends. Um, if you can't afford it, don't go. I would say if you're looking for a job, see if they're gonna actually [00:58:00] have a job fair in the expo hall and to see if the companies that's gonna be there, are they looking to hire?

[00:58:05] When I went four years ago, they was definitely there to say, "Hey, we, we here to offer people on the spot." So people were getting offers back then. Now, I don't know what that is like now. I know, like, funding and a lot of stuff has went down there, but the fact that they're still putting on a massively huge commercial conference is still a thing.

[00:58:25] Like, this will be the only conference that you go to where so many people look like you. Mm-hmm. And so yeah, I'll just say make the best out of it. Learn how to network, learn how to talk to people. Come, I won't say dressed to impress, but come dressed comfortably. Have a good time, but don't do nothing I would do.

[00:58:39] I would say also make sure that if you're going to be applying for these available pipelines, make sure that you are, um, and you have updated your LinkedIn profile because they will look. They will look. Make sure your resume is updated. Make sure you're basically in tip-top shape to present your best self out to these folks 'cause this is y- this is your time.

[00:58:57] I know last year Snapchat, um, I, [00:59:00] I don't know if they were doing interviews. I feel like they were, but they were doing a lot more than a lot of the other companies were, um, at moving that needle forward for people who are interested. Yeah. And if you see Shorty in the building, make sure you let her know you are a watcher of the podcast.

[00:59:14] Let me know, let me know. She'll have a special booth where it'd be cool. Now- She'll have a special booth this year ... okay, let me, let me tell y'all guys what I envision, and I haven't used and I haven't used my face down card yet to try to get with Morgan to ask her about this. But if y'all wanna see me back in Houston, I would definitely want to do this show live there.

[00:59:37] That would be fun. 'Cause I know they be having the podcast booth there. They had a podcast last year. So that's what I would wanna do, and if I'm able to do that, I would definitely come 'cause I think it'd be beneficial for everybody. Um, and hopefully next year we work on our proofs of, proofs of concept so we can actually show them like, "Yo, this is what it's gonna look like."

[00:59:55] And that, or we can do something adjacent to it where they say it's a sanctioned AfroTech event [01:00:00] and where we're able to get, like, now I don't, I don't really like the venue we used in Houston with, uh, Hacker Proof. Mm. I want something that, uh, people can see much more of us at. Yeah. So I, I want, like, a better venue and make sure the sound is good.

[01:00:14] But yeah. But that's cool. Um, as for me, I have some other stuff, but I think what's more important is people love when we talk about the Reddits. And this is one I think you're gonna like. So let me put it on real quick. Let it, let's go. And let me do I feel like we talked about this so many different times, but let's see

[01:00:45] So the Reddit says, "Finished my, my, uh, finished my IT degree with no real skills. Is a cyber master's worth it or am I just buying time?" Okay, so the difference is they're not a US citizen, but they are [01:01:00] 23 from Nepal, and I'm an international student in Australia. I just finished a Bachelor of IT at a small private college in Sydney.

[01:01:06] Grades were okay, but honestly, I don't feel like I know much. I've been bartending for about three years and haven't had an IT job yet. My final project was with a real company and I led the team, but I built most of it with AI tools My student visa ends in March. After graduating, Australia gives you a post-study work visa called the 485.

[01:01:28] It lets you stay and work full-time for a couple of years. The problem is, I don't think I'd find an IT job in that time, and once you're on that visa, you can't switch back to a student visa from inside the country. So I'm planning a two-year master of IT and cybersecurity instead, which costs about, uh, AUD 36K.

[01:01:46] I don't know how much that is in US dollars, but- What is it? AUD 36K. I'll look it up. If I'm honest, the main reason is to buy more time here. What worries me is that I'll graduate in 2029 and AI is moving so fast that [01:02:00] I don't know what entry-level IT or cyber jobs will even look like by then. It's also a lot of money for my family.

[01:02:06] So his plan: get an IT support job first, move into cyber doing the master's, then use... Okay, so about $25,000. That's about, that's actually a good price for a master's. He's going to, uh, try to get a support job first, then move into cyber while he's getting his master's, then use the post-study work visa after to get a year of IT experience.

[01:02:32] The end goal is permanent residency. In Australia, you can apply through a point system if your job is on their skill shortage list, and cybersecurity is on it. His main questions: Is a cyber master's still worth it or is it just an expensive way to buy time? Is IT support to cyber still a realistic path with AI changing things?

[01:02:50] If you started out without much interest, did it come later once you were actually working? I'm good with people, enjoy leading the team and dealing with clients more than the [01:03:00] coding. I've got a lot of customer service experience. Are there other careers in IT or outside of it I should be looking at instead?

[01:03:06] So even just going off his questions, I'ma say that he would be great in, like, a security sales position. Mm-hmm. He like talking to people. He ain't too technical. He don't have a lot of interest in a lot of the technical stuff. A lot of times people think that, that you're not technical when you're in tech sales, but you are.

[01:03:25] Well, it depends. Well, I guess, okay, it does depend 'cause there is a seller who's not technical, and then there is the technical person that's helping the seller. Yeah, so that's what I'm saying, like a person to start on 'cause it seem like he not interested in... I know before I get to the comments, I'm like, "Well, you did a great project.

[01:03:38] Granted you used AI tools for your, like, your capstone. You should've led with that to get you an entry level job." Mm-hmm. And they have jobs specifically just for new grads. So let's see what some of these comments say. Um-

[01:03:53] That's too long. I ain't reading all that They kind of discouraging him. [01:04:00] "Bro, getting a cyber job is harder than getting an IT job. It's a niche within a niche. I consider that heavily if you already can't get an IT job." Somebody said, "It's worth it. You're in Australia, pursue work in the FIFO mining operations.

[01:04:11] They need cyber too. They'll pray-- They'll pay through the nose to get it. Volunteer work, volunteer work, volunteer to work on their OT, ICS, and distributed control systems. You'll make bank and they'll make you a citizen if you hang around long enough with skills no one else has. Ignore the traditional IT support gigs and put all your effort into finding industrial IT.

[01:04:31] They pay more and they're incredibly hard to replace with AI unlike traditional IT. OT is where the metal meets the meat, sometimes literally."

[01:04:43] So I actually agree with that. A lot of people, you hear some random people, like nobody ever says, "I'm going into OT." And OT is kind of where it's at. Even right now, they've been talking about, and this is something I didn't bring up 'cause I haven't looked up into it enough, but have you been seeing the Pentagon talking about, hey, be prepared for [01:05:00] like, to live off the grid for like 90 days?

[01:05:03] The Pentagon? No. Look it up. Like Google it right now. What? 'Cause it's kinda... I wouldn't put it past them, especially with how, with, what happened with COVID. But that's kinda like if all that stuff goes down, you have no way to contact nobody. Like, I mean, you can have some long distance radios, but even then, those things only go for like X amount of feet.

[01:05:25] The Pentagon is developing a requirement under the Fortress Initiative directing major US military bases in the continental US to operate independently for at least 90 days if a war or cyber attack knocks out the civilian power grid and supply chains. Yeah. Yeah, which is like, what is, what are the civilians supposed to do?

[01:05:45] They don't have no power. Can't even contact people back home I got, I got a Reddit Okay, what is it? Transferring work files for a portfolio. So this person says, "Anticipating upcoming layoffs, company going through [01:06:00] acquisition deal. Should I transfer local files on managed laptop to my personal so I can use as my portfolio piece?

[01:06:08] Management is warning, is that scary?" Or excuse me, "Is that risky?" Someone said, "I'm worried," in quotes, "I'm worried they are going to start looking at people to fire, so I would like to do a thing that will definitely get me fired. Anything and everything you created on a company laptop is company property.

[01:06:25] You cannot and should not use that in any sort of portfolio. If I was a hiring manager and you brought me something you made at your last job, my immediate thought would be, 'What will you take from us here?' For future references, start your own personal GitHub, build whatever on your own laptop in your own time, then utilize that at work.

[01:06:44] I'm sure your tooling is just some script work, but imagine you built the next level of AI analysis or something. Your previous employer will sue you to the ground for creating that with corporate resources and then taking it with you after termination." Someone said, "Put your thinking [01:07:00] cap on, bud."

[01:07:00] Someone said, "Nah, bro, just email them to yourself from your work email." Someone commented and said, "No way. Any DLP would catch that." Someone said, "Just encrypt it, encrypt it with Base64." Someone said, "Ha, worst thing you could do, back up the email, entire account to one PST file, and then save on a flash drive."

[01:07:21] Um, doo, doo, doo, doo, doo. Someone said, "I would read your employee handbook. What, what you created at the company is usually owned by the company, not you. You could always screencast and discuss what you built. I personally would not risk copying the files to my personal device." Somebody said, "Why? Is it worth the risk?

[01:07:39] The only thing you need to take from there is a reference, and you are working on the opposite, burning bridges." Um If you try using whatever means and it succeeds, then do that. If your company is lax in security and doesn't block it, that's on them. However, if there's a clear policy that you shouldn't put work files on personal devices, then they could [01:08:00] terminate you if they find out.

[01:08:01] Someone said, "I cleared it with legal, DLP, and my boss. Obviously, you need to standardize and get a peer to review the archives, but that's the best answer I have for you." Um

[01:08:16] Make sure there's no sensitive stings. Um, ensure file label is not sensitive or the likes. Triple nest the archive, move the file directory and file name on each arc... That's too much. Bye. Um, but that's too much. Exactly. That's too much. Like we're trying to keep people here. But I thought that, I thought that was, um, a, something good because a lot of the times folks don't think about the consequences of taking things that they make at work off of their, um, work environment and putting it into some sort of tool to get it to them personally, whether it's email, um, whether...

[01:08:50] Doesn't matter. Yeah. I got one more, but I wanna read some of these comments. Mark Easton popped in the building talking about what's popping twin. What's up [01:09:00] gang? He said Japan said the same stuff that we were talking about too, about the grid, and then he talking about, "I love your costumes." Thank you.

[01:09:07] Thank you. Yeah, man. Yeah. Um, which one do I have? I think the other one I had was

[01:09:27] All right. This is one that's relatable to some people, and it is

[01:09:36] Leaving 100K Help Desk for networking role. So let me just say- Leaving 100K Help Desk? Yep. All right, big Help Desk I mean, if, depending on what city you in, you can make that doing help desk Especially if you're somewhere like California New York. But- Or New York ... I mean, [01:10:00] it's the bottom of the barrel. But so it says, "I was laid off for five months.

[01:10:03] I was previously a level two support engineer at an MSP for 1.5 years doing just about everything short of major deployments. But I had my hands in everything. At the tail end of it, I was essentially the acting network admin or network monitor for the MSP firewalls. I eventually landed a tier one corporate help desk role making 100K, contract W-2 with benefits.

[01:10:24] I'll say tier one, at least at this corporate building, is intensive. Lots of tickets, lots of working hard to keep your head above water. It might not be technically deep, but you're definitely not slacking here." Yeah, they paying the 100K for help desk. They finna work the dog crap out of you. I was gonna say the other one.

[01:10:40] Another. "I am leaving for a lower paid, full remote permanent network operations role at a MSP. Pay cut will hurt, but it's not horrible. At 70K plus three months review, the increase for 73 to 75 areas will be going down 25K. Just wanted to say that the level one is no joke, but I wouldn't want to do this job long term, even for 100K.[01:11:00]

[01:11:00] 90% of my colleagues have been with the company for decades, and it's a great role if you're older and looking to just plant your feet at a company and call it quits. I'm only 3.5 years into my IT career, so I don't feel like this is what I want. Some may criticize because I'm leaving money behind, but I have hire," he spelled hire H-I-R-E, "aspirations for myself in terms of what I want to do in IT as well as income.

[01:11:23] Help desk tier one is no joke 100K for tier one is bonkers. I think this is an excellent move for your career, even if it's salary decrease down. Help desk is a stepping stone. You paid your dues. Long term, moving up to networking is the better move. The Dude said, "Can you send me the link to the job posting?

[01:11:42] Because for 100K a year, I'd be willing to try to suck the chrome off a bumper."

[01:11:50] That's crazy. There's no job posting. They keep these roles only for the hiring companies to find a short list of recruits to get rid of the fluff interviews. At least that's the theory. So in my case, a recruiter reached out to [01:12:00] me. Some person said, "I'm not even at 40K as a tier two." This is just real. I made 45K working at the Geek Squad in 2009.

[01:12:10] You are likely in a lower cost of living state or far from a tech hub. I've seen them go that low for rural small towns. Let's see. Uh, are you in the United States? I make 85K as a tier one in the US in a medium cost of living state. Yep. No idea who's paying these amounts for tier one. I was happy to start at 24 an hour last year after many job applications.

[01:12:32] Usually, yeah, I figured usually it's finance and major corporations and probably to get quality hires. The recruitment company that reached out to me usually only hires for dev or software roles. I was told that the three people that interviewed were all highly technical people, but they were most likely looking for a personality hire Yeah.

[01:12:50] It's, it's just all like back and forth stuff, but I thought that was pretty interesting because I don't know if I'll go down 25K. That's a lot of K. That's [01:13:00] a lot of K, I ain't gonna lie. That's a lot of K. It depend. That's a lot of K for me. I don't know about that one. I don't know about that one. Every case begins with K.

[01:13:08] But nah, all seriousness, that's a lot. 25 bands I, I ain't going for that one. It, it- Now I could see if you went down to about 85. Yeah. Now granted, this person might not have no kids and none of that other stuff, but that's a lot because that's assuming you like networking. Mm-hmm. But hopefully it works out for you.

[01:13:29] I mean, you're still doing above average, making the $75,000 a year, so I mean, it's still a good salary, but it's hard... I tell people all the time, the reason why I'm for you to get the most money you can at one of your current roles is because it gets easier to ask for more in the next role. So now you may interview for a role that pays 130, but you just wanna get back to 100K, so you undercut yourself.

[01:13:54] And a lot of recruiters are not gonna submit you at 120. They're gonna submit you at whatever you requested. Yep. [01:14:00] Whereas if you stay there making 100K, it's like, "Man, leave here, I want at least 125." And you get a role that make 130. So that's the whole thing. Hopefully his mindset is as he gets skills in networking, he goes back and makes the money that he's leaving off the table.

[01:14:15] So that's what it is for him. And I know you said AfroTech was your last one. I, I already talked about it. I know. Oh, yeah. So, um, I'm going to go into, uh- Afro- Yeah, 100K for, uh... Hell, that shit is crazy. Let's see what other video files I got on here

[01:14:38] Our favorite, uh, TikToker. Let's, let's get it to her[01:15:00] [01:16:00]

[01:16:02] They do treat people like that. They do treat people like that. Um, they assume that because you're unemployed that The, the s- the slim, the pickings is slim and they can do you any kind of way and they're, and they are, they cannot be respectful of your time and assume, "Well, it's just 30 minutes." Well, you don't know what somebody gotta do 30 minutes later or an hour later, or how they made time and that was already pushing into maybe their availability that was already- Yeah

[01:16:33] tight for real. And that's why it's good to not making assumptions, 'cause when you assume, you- Make a- ... ass out of- ... yourself ... That's you and me. Oh. Oh. You never heard that? No. It's out of yourself. No, it l- spell assume out. I, you and... Oh, U and me. Ah. Yeah, I get it now. Yeah Uh, let's see. So back to the other, the previous story we was talking about.

[01:16:57] She said: "I can't do that. I have to be at [01:17:00] retirement age just to do it for fun. That's a lot of living you're cutting out even if you don't live large. Facts." I know that's right, Jay. He said, Ray Maestro said, "I got into Per Scholars. What do y'all think about Per Scholars? They are partnered with Bank of America, Wells Fargo, Barclays, and two other companies.

[01:17:15] What will the hiring process be like at the end?" Okay, this is what Cyber Fairy was in, so- I was in there, did that too. You did it? Yeah, I did it. Okay. I-- Okay, I did it and I dropped out. I'm honest about, about my life. Um, Per Scholars is really, really good, I think, especially if you are trying to learn whatever certification material it is that you're going for.

[01:17:37] I did the A+ and I dropped out obviously because I didn't need to get the A+. I wanted to get the Security+, but that course was starting a lot later than the A+ and I was just trying to hurry up and get into something. Um, they have great resources. Um, depending on where you're doing them at, like what city, um, there may be opportunities for you to visit other companies 'cause I know kind of here in the Dallas [01:18:00] area I've seen them, um, go to several big tech companies to tour, um, and have like different networking opportunities there.

[01:18:07] Um, I will say is keep your resume tight. I know they have like resume services there that can help you ensure that what you're learning, you're applying that skill set to your resume so that you're able to kinda showcase what you're learning in the boot camp. Um, I don't have anything negative to say about them.

[01:18:24] I think it's one of those situations of it's what you make it. If you get in and you lollygag and you don't really get in the field, um, and practice the things that they're teaching you, if you're not labbed up, then it's not gonna be beneficial for you. But if you're serious about it, because it is a huge time commitment, it's almost one of those situations with the amount of work that you have to do and how long you're in class for a day, um, so it's almost impossible for you to have a job.

[01:18:50] Um, so that's something you have to also keep in mind. If you're going to make that sacrifice to not be working for that timeframe that you're in the boot camp, then you wanna make sure that you're making the most of it. That bug over [01:19:00] there looks so real, it just scared me. And he said, um, he started going for cybersecurity, but as he did his initial interview, he decided, well, they decided he should probably start off at IT support with AI.

[01:19:13] Okay. And, um, I don't know how you'll find it on the other page, but on, on Tech Church Channel page, we reacted to, uh, a friend of the show, her name's Cyber Fairy, and I do not know what we titled them joints. If you comment back on this video once it's processed and everything, I should be able to find those videos.

[01:19:34] But she did Per Scholars and got a job with it, but there was discrepancies with people getting paid certain things. Mm-hmm. And so that's what's one thing you wanna look at is Are you being pair, paid fairly and to negotiate as much as you can. Because she brought that up and I think it ran into some issues with her, and that's one of the things that happens when you talk about pay.

[01:19:57] But just make sure you're trying to get [01:20:00] the most bang for your buck. Or if you ain't... Honestly, I don't go around asking people what they get paid. As long as I'm good, I'm good. I don't care what you get. Mm-hmm. You could be making $250,000, that don't matter. As long as my money taking care of my family, I really don't care.

[01:20:12] Yeah. I would say, too, do your research. Um, look up #perscholas on, um, TikTok and just kind of hear people's experiences because there's nothing more true than kind of hearing... A lot of people who are in that program are documenting on TikTok their journey, their learning experiences, so you might be able to make some connections there and, um, some con- le- you know, some comments to those folks to ask them about their experience, especially if they're in the city that you're trying to go to.

[01:20:41] No, that's facts. That's big facts. Now, um, one of the ones that I was excited to talk about was the, um... Let me go right here. There was

[01:20:54] ICE and then was doxed. So Department of Homeland Security personnel breach, or it should say doxed. [01:21:00] But let's talk about it real quick. So a major breach of federal data systems resulted in the theft of personal records belonging to thousands of Department of Homeland Security employees. Hackers compromised back-end infrastructure storing sensitive personnel files, exfiltrating detailed records of active duty agents across Customs and Border Patrol and Immigration and Customs Enforcement, ICE.

[01:21:25] Leaking personally identifiable information of law enforcement personnel creates severe physical security doxing and targeted spear phishing risks, exposing federal infrastructure vulnerabilities. So federal personnel databases keep getting targeted because the data inside is highly valuable. Securing PII for agents on the ground should be top priority for government cybersecurity teams.

[01:21:45] Um, I think it's just scary too when it come to ICE, but we can just ICE on everybody and just talk about ICE. I think it just gets scary because of why people don't like them and everything like that. And the crazy thing is these people pay, got some good bonuses to go work for ICE. They do. [01:22:00] I've seen them online.

[01:22:00] They got some, they got some great bonuses, and some of the people are really just doing their job. I mean Whether you like the way that they're doing it, ISIS has probably always been around. I don't know why ISIS became so famous in the last year or so versus they've always been here. I think it, I think it's new, that name.

[01:22:20] I don't think so. 'Cause I think it falls, um, under, like the- Let's see ... like, uh, 'cause I used to work with a lady who did immigration stuff, like, for, um, like US citizens, like the, the applica- like the visa process. They've been there for 23 years. Really? US Immigration and Custom Enforcement has been around since March 1st, 2003.

[01:22:42] Oh, wow. That's what I'm saying. Everything's amplified 'cause they don't like the president. But the stuff that if you are here that you're not supposed to be here and you taking jobs from people, hey, I'm not mad at that. We just was talking about earlier how, uh, I read a, uh, [01:23:00] word for word. This is not the same thing, but it is one of the things that Americans were upset with.

[01:23:06] But Texas bans businesses from employing H-1B visa workers from receiving certain taxpayer-funded state grants. Because some of these people were getting the H-1B, some of the people weren't even doing tech jobs, and they was getting a whole bunch of benefits that the average American was out still struggling and everything else.

[01:23:23] So I was like, "This isn't fair. We ain't come over there ain't get no benefits for moving over there." Now, you could probably say I'm talking ignorant, but I'm just saying seriously, like, and, uh, I'm trying to take the world for, like, what it is. Mm-hmm. It still should be tough cookie, but at the same time, if you want people to stay in this country and, you know, and not go too off the rails, you gotta take care of them.

[01:23:45] Mm-hmm. And they're not doing a good job of that 'cause inflation, since COVID- Ooh, child. Please ... people did, people thought they was getting manna from heaven in, with COVID. They were getting what? Manna from heaven. You know that story about manna from heaven? Mm-mm. When people was [01:24:00] hungry. It's probably most- a lot of stuff is, like, really spiritual and metaphorically, and, um, it's another word, uh, for, like, people, like, call stories and stuff like that.

[01:24:11] It's not a parable, but it's another word. I'll find it later on when I think about it. But they were hungry, so they needed something to eat, so they had manna from heaven to eat. Um, people during the COVID, a lot of people weren't making... They, whatever they unemployment check was, plus the $600, that was the most money people had ever made.

[01:24:28] Oh, man. People was getting big shmoney. But they thought, they didn't know it came with a cost. That part. We've been paying that cost ever since then. Mm-hmm. And the, and the people like me who wasn't getting it is, like, pissed off 'cause, like, I'm paying for stuff that I ain't even agree to. I didn't even get no benefit of no Trump checks Nope.

[01:24:46] So you gonna have to dial it back. I, I also saw, uh, I know I keep on harping on this H-1B stuff, like, man, he got something against them. Not really, but you know, in North Texas, they built like all these communities because these H-1B people were getting these tech [01:25:00] jobs moving out here. Now with some of them not getting these jobs, a lot of them houses ain't selling.

[01:25:05] And they having to slash the price to get them sold. Some of these people have gotten the houses and decided just to go back to India and leave everything. Mm. So we'll see how that happens, but that's definitely a thing, like, why these people are counting on this head count for all these people to move, 'cause if you go to a, a lot of those new communities, that's what you gonna majority see.

[01:25:24] Yeah. So I don't know. Let me see if we got any chatsters in here that say that. Uh, he says, "When you say negotiate, how low is the number? And did she get real interviews with the company they had partnerships with?" Yeah, she got a real job, and I forgot what she was making a year. I'll have to go back and ask her.

[01:25:42] So if you send me a message on LinkedIn or something like that, I'll remember to go hit her up and ask her. Um, yeah, she said because of Trump, right. And then Brendan Corbin said, "Leaving houses and cars." It's crazy. Hey, think about that. You ain't finna get me if I'm over- if I'm international. That's tied to my, [01:26:00] uh, US, uh, social- US

[01:26:01] whatever it is. I'm good over in India. Big India. Not the little one. Um, but yeah, I, I thought that was, like, interesting. Um, let me see. Did I, did I cover anything else? 'Cause technically we had-- I had one with like North Korea, and they blame... They stole 6.3 billion in a Binance breach. Um, but that's pretty short.

[01:26:29] Let me see what else I had here. It was like different things talking about recruiters. Let me go back to this real quick

[01:26:41] Um, AI and employee, AI and knowledge I thought this was pretty cool. This is a guy who I think used to be a tech recruiter, and he's talking about the feedback he got on why he didn't get the job[01:27:00] [01:28:00] [01:29:00]

[01:29:34] Yeah, that was pretty trash of them. Five interviews is pretty trash, period. I'm, uh, it don't take five for you to know. Shoot, it really don't. And on top of you telling me-- I hate... That reminds me of like, you go to all the rounds of interviews and it's like, "Yeah, we want somebody with more experience." You knew my experience when you talked to me.

[01:29:51] Well, you, you knew it was the same experience when you talked to me time one, two, three- Exactly ... four, and five. And then normally when you say experience, say what? They'll never [01:30:00] answer you back. It's just some BS stuff. Like, that happened to me with Samsara. Like, they gave me a, um... What did they say? Let me see if I- It's-- Let me see what email it's in real quick.

[01:30:11] It might be in my Hotmail. Let me see real quick, 'cause y'all know me. I'm Mr. Keep Everything. I'm a money fiend. Mm. Let me see Samsara.

[01:30:30] Okay. Yeah, this was two years ago

[01:30:51] Let me see

[01:30:56] No, that's the one I did in 2024. What's the one in 2023?[01:31:00]

[01:31:22] I can't, like... I'm mad 'cause I found one of the ones where I kept on going to the next level, and I don't see it here

[01:31:34] But yeah, I, I gave them like, I interviewed them twice and one of the reasons was so stupid and they didn't even know like, hey, I was gonna get the real scoop from my friend that was working there. To where y'all, y'all still interviewing me in hopes that the person that y'all had already offered didn't accept.

[01:31:49] So I was, I was killing the beat, but you didn't let me know you offered somebody. I was like, "Well, why would I still be interviewing?" Like they gonna accept the offer. Yeah. Then I was like, "All right, cool." I interviewed with them again [01:32:00] with the person they selected over me, and they BS my interview. So I was like, yeah, I get out.

[01:32:07] I, I told them something. I never wanna apply there again. So it is what it is. You gotta just be on, you gotta stand on guard with a lot of that stuff. Um, since we're on still on recruiters, let's get to this recruiters exposing[01:33:00]

[01:33:42] Um, let me see. Heavy on that name piece for sure. What I'm finna say? Um, it was a guy, it was a Black guy in the comments like, "First of all, duh, this what Black people have known all along, and y'all just acting like we just talking." And so, like, you be talking back to the recruiter. Let me play that smack water real quick[01:34:00]

[01:34:09] For real. 'Cause they be acting like you speaking French. I'm like, "Nah, this is this. This is, this is why I didn't get selected." Yeah. Like, this is why. Like, I actually reached back out to Buddy who watched the show about one of the roles I was interviewing with, and, um, I was like, "Hey, how did Buddy do-- How's Buddy doing?"

[01:34:24] They picked over me. He was like, "He doing all right. He a little annoying." He was like, "You'd have been a better fit, though." I was like, "This person trying to circumvent the scope of their role and do too much." I was like, "Yeah, they always trying to come in there trying to prove something." I was like, "It is what it is," but- Haters gonna hate.

[01:34:40] Ballers gonna ball. I was just thinking about the Dave Chappelle version of that. You know that version? I didn't used to watch it. It was too ignorant for me back then. I didn't, I didn't watch Dave Chappelle, but I'm saying I knew the song 'cause it was like on the, we used to have like the Ricky Smiley tapes But that's on Duke, oh, that's why it's kinda hard to cancel that, man.

[01:34:58] I ain't gonna lie. [01:35:00] Um, let me see. I think this last one we gonna do is this WGU search 'cause people, you know, they love them some WGU[01:36:00] [01:37:00]

[01:37:05] Are. I had to say it indubitably. I didn't even know what it mean. I had to go look it up. What does it mean? It means, uh, when you use it, it means something that's so obvious or true that no one can reasonably doubt it. Mm. So without a doubt, unquestionably, or certainty, or immediately Or immediately un- I'm screaming.

[01:37:28] I'm screaming You just caught on to my joke? I'm screaming. I did. Immediately, undoubtedly. Um, I think we are at the end. I didn't think I missed anything to bring up for anybody. Um, so I appreciate everybody that was in this live with us. We would like to make this a common thing, but we just gonna do what we can when we can.

[01:37:51] Um, enjoy Halloween. Um, don't eat too much candy corn 'cause I'm big on the candy corn. [01:38:00] And oh, I'm gonna leave them with a joke I stole off of Nightcap, and then we gonna go. What is the difference between socks and cameras?

[01:38:13] Um, I don't know One takes photos and one takes five toes. Wait, project socks and cameras? What has fo- what's five toes? I don't get it. Five toes. Break it down. Photos- Oh ... and five toes. Bye. I was, I was... Please. Ple- that was, that was clever. That was clever. Yeah. Then the last one was that I stole from there was, uh, how does Jesus make his coffee?

[01:38:47] How does he make it? He brews it. That's a good one. That's a good one. And then the last one, I'm gonna leave y'all with this. This is dad joke hour. Why did the iPhone go to the dentist? Mm. [01:39:00] 'Cause he had a Bluetooth. That was a good one. All right, y'all. We gonna let y'all go. Enjoy y'all night, and, uh, hopefully y'all had fun.

[01:39:10] And until next time, we out. Peace. All right, let me hit end on this