AI Governance, Intelligent Systems & The Future of Learning | Alapeti “AL” Ware | AI Accelerator Podcast

AI Governance, Intelligent Systems & The Future of Learning | Alapeti “AL” Ware | AI Accelerator Podcast

AI is evolving fast but without control it creates inconsistency, risk, and missed opportunities.

In this episode of the AI Accelerator Podcast, host Matt Zembrowski sits down with Alapeti “AL” Ware, Co-Founder of 8P3P, to explore how AI is shifting from tools to true intelligence infrastructure.

With a background in XR, metaverse development, and enterprise AI systems, Al shares how he is building an intelligence control layer that governs AI outputs, personalizes learning, and integrates seamlessly into existing systems without replacing them.

Together, they break down what it really takes to make AI useful in education and enterprise, why most implementations fail, and how organizations can move from experimentation to real impact.

In this episode, Al reveals:

◼️ Why most AI implementations fail in real-world systems
◼️ The problem with inconsistent AI outputs
◼️ What an intelligence control layer actually is
◼️ How AI can personalize learning for every student
◼️ Why traditional grading systems are broken
◼️ How AI supports teachers instead of replacing them
◼️ The role of AI in workforce training and onboarding
◼️ Why AI should never replace human decision-making
◼️ The importance of governance and policy in AI systems
◼️ Why “rip and replace” is a mistake in enterprise tech
◼️ How to integrate AI without disrupting existing workflows
◼️ The shift from AI tools to AI infrastructure
◼️ Why LLMs alone are not the future of AI
◼️ How adaptive learning improves long-term outcomes
◼️ The opportunity to reshape education using AI

Key Learnings

✔ AI is inconsistent without proper control layers
✔ Most companies are adding AI instead of integrating it
✔ Personalized learning is one of AI’s biggest opportunities
✔ AI should assist humans, not replace decision-making
✔ Governance and policy are critical for AI success
✔ Education and workforce training follow similar models
✔ Integration is better than replacing existing systems
✔ Focused AI systems will outperform generic tools
✔ Human oversight remains essential in all AI workflows
✔ The future of AI lies in infrastructure, not interfaces

💬 Al’s Most Powerful Quotes

“AI should never be the final decision-maker.”
“An 85 percent does not mean a student understands the subject.”
“We don’t replace systems, we control the intelligence inside them.”
“Most companies are just putting a sticker on AI.”
“The future of AI is systems connected to LLMs, not LLMs themselves.”

Follow Alapeti Ware

Website: https://8p3p.io
LinkedIn: https://www.linkedin.com/in/drgenius/

Follow Matt Zembrowski

Website: https://leadingaiagility.com
LinkedIn: https://www.linkedin.com/in/mattzembruski/
Email: matt@leadingaiagility.com
Phone/Text/WhatsApp: +1 978-618-5778
Facebook: https://www.facebook.com/mzembruski
Instagram: https://www.instagram.com/thelifeofmattz/

Transcript

SPEAKER_00 0:05

Welcome to the AI Accelerator Podcast, where we turn AI from a corporate buzzword into an unfair advantage. I'm Matt Sembrusky, and this is the show for leaders who are done talking about AI and ready to execute. Today I'm joined by Al Ware, who's a battlefield-tested leader, turned deep tech pioneer, who has spent two decades leading elite teams under fire in the last 15 years pushing the boundaries of AI, next gen computing, and quantum workflows and all sorts of cool technology that we're going to start to dive into here.

SPEAKER_01 0:50

Thank you. Yeah, I'm excited too. I'm looking forward to this conversation.

SPEAKER_00 0:54

Yeah, I really appreciate you being here, Al. And I had a very abbreviated um overview of your uh your background and so forth, but I'd love for you to share a little bit more, if you could, with the audience. Just tell them you and I have had a chance to acquaint ourselves on an earlier uh earlier meeting that we had, but I'd love for you to share for our audience a little bit more about your background, um, what you're passionate about. You have so many just different skills and uh and passions. So I'd love for you to just uh explain your background a little bit more and about how it led up to what you're doing now with your your current company.

SPEAKER_01 1:28

So initially I started off. Uh you ever heard of a virtual world called Decentraland? Decentral land? Correct. No, so it's a it's a web 3 world where it's basically you know like crypto enabled, and you have to enter into this world, and it's like basically virtual, it's the metaverse. And I started off in that space creating digital avatars, and it was just one of those things where it was like, all right, I know that I want to be in the virtual space and AI, but I also knew that I needed to learn every aspect of what I wanted to get involved in. So, you know, eventually I moved on to where I entered into meta, uh, spatial, and I started creating immersive worlds for big enterprise companies. And then AI, before AI became a thing, I was already uh building AI solutions within these worlds to where they would have educational instructions for like students for school districts, or you know, just showing other corporations how to implement AI into these virtual settings. So that basically led to me diving into, and at the time I didn't know how to research and publish my own research. So I had to teach myself how to do all those things, and that's when I started to realize like how big of an opportunity there was for people like myself to really dive in and fine-tune and and and actually build the ecosystem to make it better than what it is now. And that's where you know it started to lead to like patents and it started to lead to like a much bigger calling where I was like, you know what, I could turn this into a company, which you know, AP3B came about. And I always tell people, you know, they always ask me, like, how did I learn so fast? And I just tell them it's a combination of like years of experience, uh just life experiences, work experiences, and just that, you know, that attitude of never wanting to give up and just putting in the work every day 24-7, and in and actually enjoying what you do because there's days that I go to bed at like two in the morning, one in the morning, but it don't feel like it's like me working, it just feels like I'm playing with toys, literally. So, yeah, I mean, that's you know, background a little back, a little bit about myself and you know what we've done and how we came about.

SPEAKER_00 3:56

That's exciting stuff. Al, there's so much more there, and I know I know a little bit about this, but I want to I want you to I want you to to dive in more because I'm very curious. You just touched on a lot of things, a lot of things about technology. Not all of our audience is super technical. Um and I want to I want to go into it a little bit more. So because you're you're passionate about serving, right? You're passionate about serving people, and I know with your with your new initiative, there's a lot there about um um uh childhood education and um and um uh early uh early childhood education as well as other aspects of all the technology um knowledge that you have with uh with STEM and the different learning. So talk a little bit more about your your um your company, your company's current like mission and what you're going after. Because I know I I looked a little bit about it on your on your website, but it's really exciting stuff. So I want to just sort of start at the top and then we can we can double click.

SPEAKER_01 4:53

So my company AP3P, we are an intelligence control layer company now, and basically what our job is and capabilities is controlling all of intelligence and providing that intelligence that a lot of these companies and educational systems need. So the initial start of this started back in 2024 when we were building for a nonprofit organization called Simple Foundation in Omaha, Nebraska, which was a government-funded project. And that's what initially opened my mind to like school districts. Where I was like, all right, we can actually implement AI into anything to where we can help these students learn in a different environment, a completely different way of like how things are done traditionally. And AI wasn't really where I wanted it to be because it was just slow, laggy. It didn't they didn't have the things that I needed it to have for it to work the way it is now. So that's what led the research and really developing this system to where it's at now, where now our system is a governance and policy where, as in education, let's just say that a student K through 12 is you know going through you know the public school system or private or or even homeschool, their record travels with them, just like a traditional setting. But their record travels in a way where it remembers where they left off as far as how they learn, so it adapts to their learning style. So now it's like, all right, so we know that let's just say John he struggles in this particular math, and at the end of the year they know that he struggled in this math, where it's gonna keep up with that pace and it's gonna let uh faculty know that hey, he may need uh intervention, he may need to you know reskill or redo this particular thing that he's having a problem in, and you know, that help them along the way because traditionally, if you get a grade like an 80 or 90, it's good on paper, but it's just a system of record that doesn't really show what they're lacking in. So we were like, we have to govern AI for one to force AI to give the same output every single time. Because right now, as you know, if you give if you tell AI to do a certain thing, it's never going to give you the same output. It's always a different output. And we figured out how to force AI into one streamlined output every single time. And we don't, and this is not us owning the systems, this is just us only the only owning the control layer. So we could tell we could go into any system and we can control that systems, AI system, we can control the LMS, LXPs, we can control those systems to operate under our control and ensure that the teachers they receive that same output. And it's the same for workforce because we know that learning and training is the same. So for the workforce, you know, imagine someone is onboarded, and once they're onboarded, you know, they take a series of tests, and there's like companies have like at least three or four LMS systems, which is a learning management system. And in these systems, they have to figure out all right, this person passed, and it's just a again a system of record, and it doesn't really let them know that all right, they they passed, but they failed these critical points. So now you have this person who initially passed, fell critical points within this assessment. Now you put them on the field, he or she out on the field, and they cause the company to have to do rework, then they have to get a supervisor involved, then they have to retrain. So it's basically spending money that they don't have to spend if they took that initial step in understanding uh where they lacked in in the very beginning. So with our control layer, we we give them immutable uh audits where the company, everybody can see like why those decisions were made, what decisions were made, how were they made, and a human has to approve it every single time. Now, the reason why we we did it like that is because we know that both education they need proof and enterprise need proof, especially when it comes time for budgets, and when it comes time for like promotions or evaluations or trying to understand the dynamic of their entire company or school system. So that's how we came about with this system, and from that, we were able to implement two patent pending systems for that particular uh system that we're that we're that that you know we currently are now, which is infrastructure, which is basically a system that lives between systems. So we don't own the we don't own the UI or the workflow. The only thing we we own is just the intelligence, and we provide that intelligence. So it allows us to focus on one thing and allow those companies to do what they do best, which is own the UI and the workflow. And what we found out is that the IT departments love us because we're not telling them to rip and replace now. We're telling them, hey, just API, simple connection, you keep all your systems, and we just ensure and we just power your systems through us. And you know, it's a it's a win for everybody.

SPEAKER_00 10:40

I like that a lot, Al. I want to talk um from a let's say it could be a typical uh a theoretical uh case study, but just from a software perspective or from a systems perspective, you know, to be clear, um, they have their LMS, they have their UI, and maybe they're maybe they're connected together, maybe they're not, but some other another application, and you're putting your intelligent layer, intelligence layer in between the two. So uh what's an example of how the experience will be different for the users? And let's I, you know, let's let's stay on the uh K through 12 from the school perspective, because I think that's um I think that's where you're um initially there's a bunch of momentum going on there with your company right now. So I know I know you also mentioned the workplace, uh, which is which is also very important for the adults out there, but for K through 12, um how is the experience different for like you you had mentioned, okay, um before uh someone's getting an 80 or a 90 or a 95 or something like that or a 75 on on a test, and that's all you have for tracking. You don't have actually the way that that student um what they needed more help with. Maybe they should maybe they showed up with a 95 in the test, but they missed some critical thinking aspects that might might might um get them later. So explain the before your system's there and then after your systems there. I'm trying to understand that before and after uh contrast for the uh K through 12.

SPEAKER_01 12:11

So before the system is before our system, right now, take a test, they pass, they get a good grade, and everybody's like, Well, you did really good, you know, you pass, move on to the next subject. With us, it's more so they pass, but maybe they miss certain questions, and everything stays the same, so they don't notice a difference, but they miss certain questions that are actual that the same set of questions. So let's just say that they don't truly understand how to uh word problems in math, right? And you know, word problems is something that a lot of people have trouble with. So let's just say they pass everything but the word problems. So now the system picks that up and says, Hey, this person passed, but they are missing critical areas within um you know problem solving. And our recommendation is that they take these 10 questions to help assist them in understanding problem solving. So then the teacher will get an approval or deny and say, hey, these are the set of questions that we think that this student's gonna need. And they have the option to either accept or reject. Now, once they accept or reject, it shows in the in the log, which is like this decision history. And in that log, it will show exactly um, you know, if they accept it, it will show, okay, they accepted this, and then in the within the log, it's gonna show why our system chose those set of questions. Why did it recommend those recommend that they take these particular questions and it just shows the overall from beginning to end of that audit trail. So that way, when it comes time for like a school district to figure out um, okay, so did this student improve or not, they could look back at the audit trail and see that yes, they didn't they did improve because the system you know was able to target exactly the the problem areas. It addressed it, not only addressed it, but it kept the audit of everything. And the audit right now is probably one of the hardest things that people can't keep track of. Because, like I said, it's just a system of record. All school systems, it's just a system of record. It it will tell you they pass, but it doesn't tell you the next steps as far as like what this person should do next and why they should do it next.

SPEAKER_00 14:40

Yeah, and it's all reliant on the capability of the teacher. Because I mean, teaching it's challenging. If you got 24 students or whatever, and you're trying to keep track of all these 24 separate little paths and journeys, you know, you it's that's almost impossible as a human. You do a good job at that. I mean, you know, the best teachers are phenomenal at this, but it's still a lot to be asked. And what you're talking about is more of like an individualized path. It's the audit is checked. Does it change the um, does it like uh for the for the student, does this uh you mentioned okay, if they accept the recommendations like, hey, this student needs to work more on the word problems, needs more of assistance. Um what happens there? Does the teacher then just have to self-prescribe uh different assignments for that particular individual if they want to work on that skill? So like how does it how does it change behavior?

SPEAKER_01 15:30

So, right now, as of right now, uh the questions are already provided for them. So they all they they they will be able to see the set of questions and they will be able to um hit once they hit accept or they hit approved, it will send to that student's profile. So the student will be able to log in and then they'll be able to go in there and take that assessment. And in within those questions, there's hints to help them along the way to like, hey, like uh think about when such and such was this. Like, what do you think that these numbers will eventually become? Like, it was it would give you hints like that, and once they take that assessment, then the the system is gonna say uh they improved, or it'll say it's critical, but it gives a percentage of that approval rating of how they did, and over time it will show like them improving over time to keep up with their record. So it's like you said, like a teacher don't have to, it's so many students that teacher have to keep up with because it's not just 24 students, like, think about middle school and high school, it's like 24 students times what three or four classes.

SPEAKER_00 16:41

Yeah, I mean forget it.

SPEAKER_01 16:43

Yeah, so it's like instead of having a cookie cutter system in public schools, now they can have a system where each student is somewhat individualized, and you know it it helps prepare these students because the well, one thing we we focused on STEM and um we focused on STEM, which is math and science, because through math and science, it opens doors of opportunities within a workforce because you have through there the ability to um reason, think through problems, uh critical thinking, like all those things that are really important in the workforce and just everyday life stems from um math and science. So, and you know, what we learned math, students fall off math in third grade here in America. So that was one of the things where it was like we have to make it fun for them and enjoyable, and but we can't overload them. So the system as of right now, I have it where it teaches them like five questions at a time, just to give the cognitive load and the science behind it is like you want to give them a break in between for their brain to just digest what they just uh learn.

SPEAKER_00 17:58

That makes a lot of sense. So you do the same in the workplace, then it's you it acts as an intelligence layer, right? You know, in inside the uh learning management or training or whatever they're calling it uh in particular workplaces, and then um it would uh each independent um employee or staff member would have their own login into the application. So based on how they're doing with whatever training they took, uh, it would then give them like a few questions, handful of questions, like here's some extra material, extra questions for you, extra assessment, extra tips, things like this. So it would give them customized, hyper customized actually, um uh ongoing uh support to uh continue their learning journey.

SPEAKER_01 18:40

Right, right. And there's two optional enterprise because uh they can have the option where the questions are provided, or they will see where they lack, and then they can provide that training for them. Because we also know that when it comes to enterprise, they have a particular way of like training or teaching. So we don't want to take that ability away from them. That's just optional, but at least they'll see, like, all right, you can approve or reject. And if you approve it, then we know that hey, the assessment is going to be taken care of.

SPEAKER_00 19:12

That makes that makes good sense. Let's talk about your um journey for a minute with you have a lot of tech background with machine learning and AI and you and quantum and so forth. There's there's a lot, there's a lot out there. What was the what was the first thing for you where, and this is probably um most people, most people listen to this podcast didn't really start paying attention to AI until Chat GPT dropped, like late 2022, right? But for you, it sounds like it was earlier. What was your what was your first aha moment with with AI and technology that really just got you like, wow, I gotta I gotta crack the code on this and just and bring it out to people, create something with this.

SPEAKER_01 19:51

I would say 2017. 2017 was that year mark for me because AI, oh, it was really bad. But but during that time, I was like, oh, this is like the potential of it, because at the time you were only able to create 200-word essays, whatever. But it was but to me, it was just like huge because I was just like, all right, if we could create this, like what are the other opportunities of what we can do? And you know, I kept trying to force the system and trying to break it, like trying to get it to do things I needed, I needed it to do, but I just couldn't. And it wasn't until 20 what year was it? 2022, when they started to offer the LLMs as a way for us to download it ex internally into our own computers, and then we were able to really break it apart and figure it out. That's when that's when things really started rolling for me, was at that point because I was able to just pull it apart and figure out like the dynamics of it, like what makes LLMs the way the way it is. And I've also learned too that I don't see LMs as being like the future, really. I I think this the future is gonna be more so with systems that connect and l that connect to LLMs, because the systems is gonna streamline very specific tasks or specific things that a lot of these companies are gonna need because you know you figure like you know, a big funnel, and you got to stream like that funnel to something just extremely small and fine that is specific to what you needed to do exactly. I think that's where the big play is gonna come in at for the future of AI.

SPEAKER_00 21:42

Yeah, I agree. I agree with what you're saying, and and um the LLMs are super powerful, but they're not focused and they're not hyper focused on an individual's needs. And uh, I think what you've built that what you're building, uh, if we zoom out a little bit, is a very, very smart play because. If you um the way the market's working, like you're saying, LLMs are generic, they go across everything and they have a lot of value, a lot of horsepower. But the deep vertical integration of an application that you're building uh for your company is really a smart move because it's those deep vertical plays that really have the ability to um you know to change the game out there uh for people because you're leveraging that ability to focus in on a particular user's needs and what matters, what makes an impact for them, what delivers the outcomes to them. And then you leverage just the horsepower of the LLM that you need, but you're you're connecting all that, you solve all that with technology, right?

SPEAKER_01 22:37

Yeah, and and also, too, to your point, a lot of these in a there's a lot of companies they build LLMs, but they're just uh like you you see it, they're just uh like a sticker slapped on another chat GPT. And with us, we're actually uh AI is secondary, uh, primary is governance and policy for us. So AI is just used in the background as a way to explain things in a more detailed manner, but everything is controlled through infrastructure, through governance and policy for us. So I wanted to change, I love AI, but I wanted to change the game a little because I know that success is going to come from companies that truly enhance AI without using AI. So it was just something that you know I wanted to really just dive in into.

SPEAKER_00 23:33

Yeah, so on on that point, let's look at that a little bit. Because in workforce development, you know, back into the office now, away from the uh the school for a little bit, you know, what should AI never be allowed to do when it relates to human, you know, I know in so many of the materials that you talk about for your company, you know, AI shouldn't should not be allowed to do, should not be allowed to replace human oversight. What are some things that like you feel like the where that line is drawn? Like AI is good at doing this part, but we still need the people to do that part. Where is that line drawn for you?

SPEAKER_01 24:07

For me, the line is drawn, like you said, the human oversight. AI should never be the final decider of anything, it should just be an a layer where it's like, okay, this is what I recommend. But and that's what it is right now, it's just a recommendation tool. And but that again, that recommendation tool, it's just it's a different output every time. So, you know, that that's one of the problems, is like a lot of these LMS systems, they have really good LMS systems, but everybody's starting to implement AI and it's just a different output every time. So for me, it's it's it's beyond you know where AI should fall. For me, it's more so how do you control AI? Fully control it to where you force AI to do what you want it to do. Uh that's that's where I'm like more focused at.

SPEAKER_00 25:05

Yeah, and you get more individualized uh benefit from that too. Um what what what mistakes what mistakes do you see uh companies out there making today uh regarding AI technology on the learning management side of things? Is it like companies that don't know about you yet? Like what is the um are they trying to implement AI just generically, like, oh let's just slap chat GPT on top of our LMS and and see if that helps or whatever.

SPEAKER_01 25:36

I could give you an example, like take us for instance. Before we started off with uh before we became you know the intelligence layer, initially we started off where we created these photorealistic avatars that look real, just like me and you, you and I. And we were I was we we are able to talk to these digital avatars, and they're AI enabled on with 100 milliseconds, and these are AI instructor assistants that we created. So it's like it's talking to you, it's teaching you the lesson, it's giving you a breakdown of everything that you need to learn. And through that, you know, we were like, Oh, this is like one of the most advanced uh tools that we've created. But then, you know, once we started to do the pilot and we start doing user testing, we found out that it's more of a problem to implement it into these enterprise companies because again, they have to rip something out to replace ours. And uh I feel like a lot of companies where they're failing at is they're so quick to just look at the next shiny tool because we were the next shiny tool, and they're willing to rip and replace something, but sometimes that's not the best option. Like, and and what I've learned too is that I can save companies a lot of money without having them to rip anything out. We just add that layer on top of everything that they have. So I feel like you know, enterprise companies should look at before ripping something out, like think about the complications of their IT department, the the tech that how much money it'll cost, like the potential damage that it could cause, you know, all those things you don't really hear from tech companies really talking about it, or those are just things that you don't really see until you're actually, you know, working with those teams. So it's just something that you know I noticed, and something that you know we we're really you know focused in on.

SPEAKER_00 27:35

So, question for you as we're as we're wrapping up, I really want to understand a little bit more about your you your your software and everything that you've designed there, uh has the potential to impact so many people's lives in a positive way. What is your what what is your big mission for this year? I guess what excites you most about what you're doing with AP3P coming up this year.

SPEAKER_01 27:56

I think the most exciting thing is now, well, for one, we got accepted into uh the veteran hero accelerator program, which is through Pin Fed Foundation. And you know, we leave in about a week to Bentonville, Arkansas. And we're just excited because it's like that final validation that we're on the right track, and now we're starting to see investors talking to us, we're starting to see doors open, and honestly, it's easier to talk about, and I think it's so much easier to talk about is because I truly love this actual infrastructure that we created, and I'm just having fun with it. And and and uh it's just every day I wake up, I'm just like, I can't believe we built this because back to the education side, we actually built an entire educational platform, and it was just only for to understand the users. I just built this just as a and I built this as a system where it's like, hey, let's just get users on here, let's understand how they learn, let's understand. And these are students, and the reason why I did that is because I know K through 5 is the hardest to actually streamline and have accurate information. Well, that app that we created actually turned out to be a success, and people actually like it. So now it's like, well, we accidentally created you know, two products really. We created AP3P, the infrastructure, and then we and then we created Sunrise STEM quest, which is an whole entire STEM learning application for for students, K through 5. And I'm just excited because I didn't expect that to happen. It was just more for user testing. So those are just things that we're just you know, we're we're excited to to walk into and just see how this year plays out for us.

SPEAKER_00 29:46

Yeah, it's fantastic. It's it's it's it's a blessing to be that passionate about what you do and to make a positive impact on others' lives there. Right. K through five is a big market that is such a powerful place to impact change for for their lives, plus for the future of our world, right? That's a really important, uh, very malleable place in uh in their lives as they grow. D um so as we as we wrap up, what's a good what's a good place for uh how can how can listeners find out more about 8p3P or or you and and and what you're up to?

SPEAKER_01 30:19

Well, uh if they want, they could go to www.8p3p.io and then they can also reach me at al at 8p3p.io. Yeah, like um, you know, more than happy to answer anybody's question, you know, see you know their mindset or where they're at and where they may need help with. So yeah, reach out whenever.

SPEAKER_00 30:46

That's great. We'll include that uh in as well as the reference to the um uh to the uh to the sunrise program that you have there. We'll include all that in the show notes so people can, if they're if they're not just listening, but they're actually online uh as they uh attend this podcast, they can just click the link and go connect with you there too.

SPEAKER_01 31:03

Okay, sounds good.

SPEAKER_00 31:05

That's great. Thank you so much, Al, for your time today. Really enjoyed meeting you or meeting you again, but uh hosting you here on the podcast. Really appreciate all your insights and uh and knowledge that you share with everybody today.

SPEAKER_01 31:17

Yeah, thank you. And uh yeah, it was an honor to be on this show and thank you for having me on here, and I enjoyed it. Thank you.