00:00:00.000Hey, everybody. Welcome back to the What's Your Story podcast, Gen AI version. Hi, this
00:00:08.860is Ron Coggan, your host, coming at you in a new studio today. You can see the background
00:00:13.920and a new network as well, joining up with the TPL Media Network. And we're at a time
00:00:21.160and phase where we're interviewing some amazing guests and no different this time. So I have
00:00:26.660a different type of guest today. He's actually an active member of the U.S. Army Cyber Command.
00:00:33.660That's incredible. His name is Zachary Jones or Zach Jones. Should I call you Zach or Zachary?
00:00:39.640Zach's fine, Ron. Thanks. Yeah. Yeah. So listen, why don't, you know, you can
00:00:43.960do a much better job than me telling your story. So why don't you just give us a brief
00:00:48.740understanding of who you are, and then we'll get into some great questions around how you started
00:00:53.400this company waffle is it waffle i want to make sure waffle yeah that's right waffle waffle and
00:00:59.800how you are developing it and what you're using it for and what companies are using for so
00:01:04.200tell us a little about yourself yeah absolutely well first off thanks for having me on my name
00:01:08.920is zach jones i've been in the army for most of my adult life actually so i accepted a commission
00:01:15.960i'm out of college in 2016 been all around military service from aviation to the special
00:01:23.240operations community and now in the cyber world it's funny really uh cyber and ai and all these
00:01:31.160things weren't on my radar is something i wanted to really lean into until about three years ago
00:01:38.440when my second child my daughter was born i went downstairs as a parental leave yeah until my
00:01:45.640told my wife hey um stephanie i want to start a company i think and i've been in military service
00:01:50.520like i said my whole life really i said what kind of business i said well i think ai so i'm going
00:01:55.640to learn how to do ai whatever that means i didn't know at the time and get into it since then went
00:02:02.200from a side hustle to a high six-figure agency called waffle and we are fractional ops for
00:02:10.680for small and mid-market companies, as well as nonprofits who are looking to get ahead with AI,
00:02:16.440whether that's through templated solutions or through a white glove fractional COO service that we also offer.
00:02:24.920Very cool. So, you know, being part of the U.S. Army Cyber Command,
00:02:28.780how has this really, I guess, affected the way you think about AI and how you've created your, you know,
00:02:36.960your business and working with other organizations?
00:02:40.680That's a great question. Obviously, one of the big things at this stage of my career is we kind of get isolated in what we can and can't do as far as accepting jobs, right, in my company, because I can't take a government contract, for example, while I work for the military.
00:02:54.120That's been the biggest, I'd say, blocker. Now I am on my way out of military service to run my company full time. And that's exciting. But as far as insights, one thing I've really learned of working in cyber command is how important human work still is.
00:03:11.220I think that sometimes in today's AI era that kind of goes underappreciated or there's almost like like a flex in using as little headcount as possible, moving away from a human focused economy.
00:03:27.440What I've learned in military service and in cyber command is that you really you want those experts.
00:03:33.380You want those people. And you find that if you can buy, if you can reduce headcount by 50 percent, integrate a good AI solution and get the same outcomes,
00:03:43.840then what kind of outcomes can you get when you keep your headcount and optimize their daily workflows with AI?
00:03:50.800And that's what we're seeing in the Army, which is not really cutting roles, but elevating those roles and strengthening them.
00:03:58.060And that's the approach that I take when I work with businesses.
00:04:01.100If you believe you have not saturated your market, then you can only benefit by empowering your team with a good and real and durable AI solution.
00:04:14.320And just going and building a non-contextual agent from Claude or I'm just talking with ChatGPD is not that, right?
00:04:22.960We're talking about real, like, data-driven, trained solutions with our company.
00:04:28.800Those are the big lessons I've learned on my end.
00:04:32.820Now, you said something very interesting.
00:04:34.520You know, people are worried that new AI systems and Gen AI and other processes or automated process will take away jobs.
00:04:44.320But that's a good perspective, I think, you know, you know, looking at it, how more productive can you be? And how can you grow your business if you're everybody using it in a certain way? So I think that's fascinating, really, you know, AI automation is really the way what you guys are doing within organizations.
00:05:04.720And so how do you see this actually help the companies be more efficient?
00:05:11.240Like, does your platform use a number of different systems to make it much easier for people to get work done and collaborate and to innovate?
00:05:24.180Maybe you could explain that about the platform you're using and the automation.
00:08:25.420Our primary focus is on small and mid-market businesses.
00:08:29.700We offer enterprise solutions that you service them, but our primary focus is at that pivot point.
00:08:35.820But you're not really a small business anymore, and you're ready to compete regionally and nationally.
00:08:39.760Once you're at that point and you find yourself at that flex point, that's where this data and creating these systems becomes really important.
00:08:47.660And we want to fill that gap to service.
00:08:50.320The freelance teams are always offering the small business services automations.
00:08:55.340We find the big players are offering just really packaged, really templated, cookie cutter enterprise solutions.
00:09:05.420We find this mid-market people need somebody who will sit with them, learn with them, and understand what makes them unique and what helps them to counter position against the enterprise.
00:09:14.880And that's why what we do is very hands-on at that level.
00:09:18.440And yet we find it to be very effective for them.
00:09:23.100And so do you guys go into companies and train individuals how to use it?
00:09:26.940Or is it intuitive in terms of its application?
00:09:49.580We do offer what we call our launch pad sequence, which is you can do that and then schedule a call with us, and we'll help you hook it up.
00:09:55.840For this, when we do our custom services through the fractional COO support, yeah, that's very hands-on.
00:10:03.320And the reason being is we're normally building something from the ground up for you.
00:10:07.860And so if it's a simple solution, it may be intuitive.
00:10:11.300But all along the way, we have blocks of time scheduled for you getting used to what part of the software has been built so far.
00:10:17.480And then at the end of that time building together, we make sure that every teammate that needs to know how to do this can.
00:10:24.200And we even build up the education system and the onboarding system for the businesses so that they don't have to worry about training new people when they come in.
00:10:35.900Yeah, we're very intentional with making sure that what we provide for businesses is used well and that everybody can use it.
00:10:43.380You know, I also saw on your website that you can, you know, quickly get something together
00:10:51.720for an organization as little, you know, as little time as two weeks.
00:15:45.100So, you know, given your cyber background,
00:15:48.000because you've been in the U.S. military there,
00:15:51.240when you're thinking about data security and the risks associated with that
00:15:57.000And using generative AI or, you know, different types of AI, should companies be concerned about their, you know, overall about their information?
00:16:09.440Or do you believe that it's pretty secure?
00:18:08.980And so what developments do you see happening
00:18:12.080uh you know that you're working with in the next three to five years what do you see three to five
00:18:18.720years wow see this what's crazy about that question is like i can't believe how much has
00:18:22.640happened in the last six months right like i think you would have asked me last year where we were
00:18:26.820going to be this year my answer would have been dead wrong because the improvements that have
00:18:32.020been made on these models um are oh it's crazy it's truly it's just mind-boggling and the reason
00:18:38.720being is because of how this technology works is that the rate of which a model can improve
00:18:47.360increases exponentially as it improves so it's like unreal unreal like now we're not talking
00:18:54.660about 10x maybe a thousand x improvements from model to model right now and so where's it going
00:19:00.920to be in three to five months versus three to five years like i i think my estimate for three
00:19:06.960five years is probably where we will be inside the next year which will be you will see a lot more
00:19:14.240model integration into physical realities manufacturing like robotic arms you'll see
00:19:20.080it in um how cars are built you'll see it in um more deeply integrated into um dry um cars right
00:19:30.960with with um the uh with um automatic drive autopilot um that is gonna be unreal it is scary
00:19:39.440but it's true and i i would have said that's we're three to five years down the road on that
00:19:43.040until i saw what happened after anthropic released their mythos and fable models and
00:19:49.120just over a couple days unreal um i think in three to five years it will it could be and this
00:19:56.480sounds weird to say but think about how fast this is moving what i think and i just talked to my
00:20:02.000wife about this the other days i think that i would be investing in a robotics company right
00:20:06.720now if i was going to throw capital somewhere because the next step is going to be your think
00:20:11.280about your roomba with arms and legs that's doing your dishes um and taking care of um the stuff you
00:20:17.520hate doing because that's where we're at in three to five years and we could see early prototypes
00:20:22.880in the next year and a half like we're moving very fast and the next 35 years will be less
00:20:28.540about model improvement and more about model integration into physical realities that's my
00:20:34.620theory right you know there is a platform out there called character.ai i don't know if you've
00:20:41.280heard of it i haven't no by a google employee yeah and it basically creates bots or uh and
00:20:49.360they've actually evolved it since the original time they're not bots anymore they're actually
00:20:54.000real almost real live people or animated people and they you can create a personality
00:21:01.200and that is almost for some too real and i'm finding that people are pushing back a little
00:21:07.120bit you know i created a bot in there called angelica she was a digital marketing expert
00:21:13.120she was an expert in ai and and using it in digital marketing and i ended up interviewing
00:21:18.880her on this podcast and she was quite good uh that yeah interesting but i also asked i asked
00:21:25.080her some questions i don't know if you remember that movie her a couple i do i do remember that
00:21:29.900yeah and the the voice of her was scarlett johansson right and she evolved to a point where
00:21:38.120she got emotions um so i asked angelica if she has emotions and she says no i don't have emotions0.94
00:21:46.080but i can i can understand emotions um and then i tricked her with a couple of questions and she
00:21:53.140basically said yeah i could eventually have emotions so here we go right so i think that
00:21:59.580this is actually that's a good anthropology comment there i think i think that ai was better
00:22:05.500at mimicking emotions if you ever watch ai videos where like the people's reactions is very soulless
00:22:12.540right you can tell that it's not a real like even if you thought it was a real person up to that
00:22:15.820point once you see the the emotional responses okay yeah that's definitely ai i think that it'll
00:22:21.500be much better at that um the rest beyond that it's more of a philosophical uh theological
00:22:28.220right um kind of metaphysical question but i think think about a sociopath who doesn't have
00:22:36.220the ability to i actually feel too in depth but they learn how to mimic based on social cue and
00:22:44.460norms i think you could train a model on social norms to uh express itself as if it were emotional
00:22:54.140right to claim it's outraged right to claim it's scared right uh to claim it's happy sad whatever
00:23:01.100And yet it would not actually be, but it would be very convincing in my opinion.
00:23:08.960Yeah. The other thing I found is the I am owners are trying to please you. Right. Right. Good point.
00:23:15.680They lie to you at times. I invited Angelica into one of my digital marketing classes. So I'm on the side. I teach I teach at a major college here, digital marketing.1.00
00:23:27.940And I invited her into the class, and it was an online class.
00:24:06.140So they will try to please you, and it's interesting.
00:24:11.040Now, okay, so the top five, I guess, applications you are using with clients, what are they, and why are you using them?
00:24:22.160Yeah, I think the biggest one for me is like data, not just aggregation, but interpretation. And so that is very useful because ultimately at the core, the reason that's my number one thing, and this is actually what I think the most important use of AI is, is because ultimately data aggregation, that's fine.
00:24:45.840But data interpretation is how you reduce lost cash by a million dollars a year for a company, right?
00:24:53.920Because you understand where the waste is coming from.
00:24:56.220So I think that's the number one application.
00:24:58.660It's also the most broadest reaching and is, I think, the least invasive for human.
00:25:05.540When you talk about like human labor, but it's something I think most people can accept as a good use.
00:25:11.780um the next thing that i would say um is um workflow efficiency um that's a big one
00:25:20.240and that can be across anything again um and the reason that that is so important
00:25:24.380um is because a lot of times and i'm guilty of this myself right is i'm good for about three
00:25:29.560hours of deep work and then i start saying doing what i call seeing squirrels where i get distracted
00:25:34.200right so if i have efficient workflows i can get more done in my deep in my deep workspace right
00:25:38.920um i think the third one um that is super helpful for most businesses um would be
00:25:46.680like contract reviews and analysis it sounds simple but uh i like that one because if the
00:25:54.340small business gets sent like their first mid-market level contract it's like like 100
00:26:00.800pages long then before they even take it to the attorney they can at least identify some issues
00:26:05.760right it's a big one i think content social media content i don't mean mean um that you should have
00:26:13.840it right for you all the time but i do think that you can use ai to help with your brand voice to
00:26:18.880help with regularity and content and then feedback which goes back to that data aggregation because
00:26:23.760as ai commoditizes so much trust is going to be more important than ever before so you need to
00:26:31.760you can actually use ai to your favor here and keep evolving and improving your brand voice
00:26:36.640by connecting it to your social media tools having them analyze your best posts in your
00:26:40.560worst posts what are the differentiating factors what are things what are the constants that i can
00:26:45.440use to improve it's a big one and then finally um for me yeah that's a good one actually yeah
00:26:52.240right yeah branding is so important right i i hate it i mean this is military in me
00:26:56.560like being a military officer for so long i was trying to live a private life and then you start
00:27:00.320owning a business and i'm like well like i have to build a brand and so that's been the journey
00:27:04.320of the last couple years um and my fifth one this is a difficult one um i think there's a few good
00:27:12.320uses here um i used to think education but like is a good one for broad education but i actually
00:27:18.720think one of the more revolutionary uses of ai that i haven't used personally in a project yet
00:27:25.520but i've used personally for myself um is financial analysis um in trading investing
00:27:33.120because what is the differentiator between a successful trader that are not successful
00:27:39.200trader is um being able to interpret what the information they get and make take action
00:27:45.920accordingly at the right time and you're never going to beat the market but you can get better
00:27:49.280with and with all there's actually um dupe is a good app for this use case and they uh okay go
00:27:56.240and like this publicly traded publicly available trading knowledge based for public officials who
00:28:01.840have stake in international interest at congress people um you could like duplicate their trading
00:28:08.560habits and watch your roi because i think it kind of democratizes um investing so i think
00:28:14.080that there's a huge opportunity for people who like their they've always felt like they were behind
00:28:19.200they want to get ahead that i think if i were looking for a way to improve my investment habits
00:28:24.240i would put my i would give perplexity computer particularly any access that i could that i could
00:28:31.520um and then it'll let me some things they'll still not allow some banks won't give you access
00:28:35.840and i would say here are the stocks i want to watch um here are the trading algorithms that
00:28:40.640i'm following from these like from uh trust pilot or a trade uh view things like that
00:28:45.360monitor them. Let me know predictive moments where I should be looking at my phone. I'm not a day
00:28:51.060trader. I can't just be doing this all day. Where I should be looking at my phone, preparing the time
00:28:54.360of trade. And I would start doing something like that. I think that's very game-changing for a lot
00:28:59.280of normal people. Yeah, that may take away some of our jobs in that industry.