🎙️ Backstage Tech by George Helgesen

Steve Mast: The 3 Steps to Real AI Adoption | Co-Founder @ Twenty44

• George Helgesen • Season 1 • Episode 15

Share episode

Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.

0:00 | 58:15

Steve Mast is an entrepreneur, AI advisor, and co-founder of Twenty44 — a company helping organizations move from AI experimentation to real adoption. Steve spent over 3 decades building digital companies, including Delvinia, where he built Asking Canadians and Methodify before selling to Sago in 2021. Now he's focused on one problem: why 45% of organizations have invested in AI, but only 12% of employees are actually using it.

In this episode, Steve breaks down why AI adoption is a people challenge, not a technology one, how Twenty44 diagnoses workforce readiness, and what it actually takes to move organizations from hype to habit.

Topics covered:

  • Why 45% of organizations have AI but only 12% of employees use it
  • Why most organizations are built on executors — and why that changes everything about AI training
  • How Twenty44's 3-step framework closes the adoption gap
  • Why embedding AI inside Salesforce drove more adoption than any standalone app
  • How to benchmark your workforce's AI readiness before spending on tools
  • Why AI is the most accessible technology in history — and why that still isn't enough
  • Whether the Twenty44 assessment platform will become a self-serve SaaS product

If you're an enterprise leader or founder trying to turn AI investment into actual results — this episode is the playbook.

👉 Follow George:

👉 Follow Steve:

SPEAKER_00

Welcome to Backstage Tech, a podcast for software founders, investors, and product leaders. Today we're joined by Steve Mast, entrepreneur, AI advisor, and the co-founder of 2044. Steve spent over three decades building digital companies and now works with leaders trying to turn all the excitement around AI into something that actually works inside their businesses. We're gonna talk about how 2044 helps organizations overcome challenges in AI adoption and their unique methodology. Before we start, make sure you hit subscribe to get notified when the next episode drops.

SPEAKER_01

Hey Steve, how are you doing? George, I'm really, really well, and it is an absolute pleasure to be here with you in beautiful Lisbon.

SPEAKER_00

It's my first time here and it won't be the last. Thank you, Steve. I'm so excited to have you as a guest on my pod. And thank you so much for coming to Lisbon from Toronto. Actually, I was really surprised when we met in the airport and I asked you if if you want to have if you want to sleep in the hotel. You said, no, no, no, no, no. We're gonna go to the city. George, you can give us a quick tour. I hope I didn't keep you overwhelmed with all the information, like Ashtell Dinata per capita, everything else.

SPEAKER_01

Oh no, I I wouldn't have retained any of the information, but uh I will say uh you could you could quit your day job and become a tour guide.

SPEAKER_00

Steve, why don't you introduce yourself just uh briefly and then we can dive into some more specific topics?

SPEAKER_01

Absolutely love to. Uh I think brief is going to be the challenge. Uh it's uh I am known to uh I'm known to talk, let's put it that way. So I uh we already mentioned I I live uh in Toronto, Canada. So uh it's a beautiful city. Canada's a beautiful place in the world as well. You know, I've called it home my entire 55 years of my life. I have uh married, got a beautiful wife, two wonderful kids, 23 and 20. They're getting close to being off payroll, so to speak, which is uh, you know, really interesting point in time for my wife and I now because we're at the point and stage of our life where we're gonna be able to maybe travel more, do more things, those kinds of things. But spent most of my career around technology. I actually went to school to be an architect. Um, I love design. Uh, Lisbon has just been an unbelievable experience, like exploring the city. Uh, anywhere my wife and I go, we're we're big fans of walking the city. I'm a you and I were talking about it. I love to jog and run. So uh it's a great way to kind of see the place and see the city, but I love architecture. I'm a huge architecture design fan. Got into uh uh computers at a pretty young age. Like I am I'm old enough to remember uh when you know there was no computer on your desktop, literally. There was no, you were still doing everything pretty manually. So the rise of that was just an unbelievable thing for me because it just completely transformed my life, absolutely transformed my life. I mean, I was never a great student, but computers were just like this miracle thing for me. I just, I was like a duck to water to it. Interesting, kind of fun story when I was really young. My dad was a bank manager for TD Bank, and uh he was there for 37, 38 years or something like that. And we would move to small little towns before we ended up in Toronto uh when I was still pretty young. But uh one day he took me to the bank uh on a Saturday, because banks used to not be open on Saturdays, but took me to the bank and he said, okay, you just play down here. I'm gonna go and do some paperwork and work in my office or sure, no problem. Whatever. So I don't know, I might have been like six years old or something like that. And I went behind the uh the teller stations and I was just playing it over. And I noticed these buttons, and I started pressing the buttons. And I noticed there was another button at another station. I just kept pressing these, but I couldn't figure out there wasn't making a sound. I was so curious. So I just kept pressing these. So uh anyway, yeah, triggered the silent alarm. We had an entire SWAT force outside, and uh yeah, I never set foot in that branch again. But um it it really was an example of I love to push buttons, I guess metaphorically in some ways as well, but I was never afraid to press buttons on computers and things like that. So uh so that really got me into technology. And then I also was uh a big fan of building uh companies and build building things. I guess that was my architecture kind of background. Right out of school, I founded a uh 3D animation company with another gentleman who was a friend of mine that went to school with. Uh, we we didn't last very long, we lasted like nine months or something like that. And uh we got bought by a video game company. So uh quickly transitioned to that. We built their entire animation group and then eventually founded uh a pretty large gaming group and I became like a producer and started producing uh video games. Really exciting field. Again, just transformed things again, exposed me to all kinds of new things. And then when I left there, the company went public and I was able to um uh cash out some shares and my wife and I did a little bit of traveling, left there, and I'm like, okay, I got to go find a home so uh or something to do. I was still pretty young. So I ended up at this company called uh Delvinia and uh became a partner there. And that was really a consulting company that moved into more of the product space, which I think we're gonna talk about a bit more. We sold that in 2021 uh to a US company. I stayed in that company for two and a half, almost three years as their chief product officer. So big global company, really exciting, really interesting with that. Left there in 2023. Uh, and no, I guess it was 2024, actually. And then I was trying to figure out what I was gonna do next. And AI was obviously uh gaining steam. I did quite a bit of planning work at the company that bought us uh around the AI space and some strategy work. So I was really interested in it. So founded 2024 with uh some other folks and uh kind of focused around like how do we help organizations adopt AI. So told you it wasn't gonna be brief. I told you it wasn't gonna be brief. But yeah, I've been very blessed in my life, continue to be blessed, uh, continue to meet amazing people like yourself, which is um something I really, really enjoy. Uh as much as I'm into AI and technology, uh I am I am I'm a people person for sure, first and foremost.

SPEAKER_00

Thank you, Steve. I hear it's been an exciting journey. And I'd really like to know more about how you decided to focus on helping enterprises adopt AI because right now it's it's a it's a major topic in a lot of industries. Could you tell me more about this transition?

SPEAKER_01

I guess I'll kind of work sort of backwards with this and talk about sort of 2044 a little bit. But the main sort of premise behind it was organizations need to really start to transition and figure out how they're going to ultimately not just adopt AI, but really be able to apply it to actually get a real return of it. And this has always been a challenge with any major technologies that have sort of come forward. I will say when I first got into it, I was, it was almost from a selfish point of view. I was like, okay, I've, you know, sold businesses, I've uh built these things. Do I really want to go back to uh consulting and starting to dive into this AI thing? And it was a bit of a selfish perspective because I was like, yes, I do, because I don't want to let this unbelievable point in time in technology and point in time in history where AI passes me by. I think I quickly realized about six months into it, as we were starting to work for organizations and just the struggle organizations were having. And some of them were just almost not putting their head in the sand, but they were very much saying, like, we'll figure it out when it becomes something we need to really figure out. So my attitude shifted and I was like, I can't let the world, or I can't let organizations let this pass them by. So it became almost like uh, almost like a uh I wouldn't call it philanthropy, but it would definitely become it became a cause versus like, I'm just gonna do this to start to consult and sort of do these things for myself, right? So the big thing was it's not, it wasn't so much of a transition to 2044 around sort of the consulting, and there's a bit of a twist around where we're headed with that business as well. Uh when I was at Delvinia, uh, I was the president chief innovation officer, and my business partner always had the idea that we should build products ourselves, or we should build businesses that have real repeatable revenue models around it. I was very much so running all the consulting business. So companies like RBC World Bank and TD Bank and some of these big financial organizations we did a lot of work with. But they were always uh, it was consulting, right? So, other words, it was time and materials. That's how these business models work. So the more time I spent in front of whiteboards drawing, you know, ideas and sort of solving their problems. And by the way, those early days of problems, like I'm really going to date myself. Like it was, can you create a landing page for me that we can send a banner ad to? Like, I mean, that was so uh, although we did work on uh RBC World Bank, we were one of the first projects we worked on them was helping them launch online banking, which was uh just an incredible experience, like uh incredibly transformational. Like you're moving people out of branches and ATMs, these kinds of things to actually try this online world. So that was that was a great project, but it was all consulting based. And then along the way, we started developing some different products that were helping us in the consulting business, particularly around understanding people's digital behaviors. There was, there was just wasn't any data, there was no research out there around really under very similar to what's going on with AI right now, like understanding the behaviors that people use AI. So we created an organization called Asking Canadians, and eventually we took it into the US. And Asking Canadians was it was just a community of people that uh signed up to ultimately do research studies, mainly surveys for the most part. And we used it at first for our consulting business, but it quickly transitioned into we were helping market research companies all over the place. So the Ipsos, the Cantars, where we were supplying them access to people we had in our database. So we got that database up to about a million people. And then we started, and it became literally a product. So we really became a product company. Uh, and then we started moving into building automation on top of that. So we built a product called Methodify, which was really it automated the entire process. So you weren't a market research specialist. You could be someone who just uh, I need to talk to, I don't know, moms in Lisbon who chew bubblegum kind of thing. Well, we could find them, but we could also give you, say, for example, a methodology, a pre-packaged methodology. Say you wanted to, I don't know, test a logo or test a new idea. You could access them and then we would give you the questions all pre-packaged and then give you the report and the data all within the one system. So, and then eventually it became like a full SaaS model where it was self-service. And that was a whole other experience that we ended up having with that. So we really transitioned from a consulting company to a product or platform company, and then sold all that to a company called Sego in 2021, as I mentioned. And then uh I'm right back at it where it was consulting again. So I mentioned on the 2044, but the twist is we're now building a product and a platform to do all the assessments with it. I mean, I think one of the things is once you get, and you probably know this is more than even I do, but once you get product in your blood, and especially if you understand like SaaS business models and it's hard to get them out of your system. Like they are fantastic, not to mention the margins and all those kinds of things, but it's just a lot more fun building something and then watching people be able to use it versus the consulting businesses in a way. And I would argue too, AI is gonna start to eat consulting businesses lunch, even our business where we're consulting on how people to use AI, eventually that's gonna take over and AI is gonna be able to do that for us, right? So the transition was consulting product back to consulting, and now even 2044, we're building our own product and platform uh within that business.

SPEAKER_00

Do you see a flywheel model here? You start with consulting, then you see a repeatable pattern, which you implement as SaaS. Users start using SaaS. Uh, there is also a consultancy component to it, like uh the team that helps you onboard, team that explains how to use the product. And now that you see new needs from your customers, you implement it and it works as a flywheel.

SPEAKER_01

Yes, the short answer. Uh I would say within our Asking Canadians and Methodify businesses, even though the two were not mutually exclusive, they had different ICPs, like different client base. We like we really built Methodify to go after clients. So, like banks as a primary client that we were going after. And Asking Canadians was servicing the market research industry. The truth is, Methodify, even though it was a SaaS platform, I would say uh 60% of the revenue came from services. So we would obviously charge them for all of the consulting work that we would do. And you're right, it was very much so one would continue to feed the other, right? So the more we would do around customer service and onboard them. So we actually had a system where we would take the organization, they may have their own market research methodologies or approaches to use. We would then adopt them and put them on our platform for them and we would customize everything. Well, that would just become like an unbelievable business model onto itself, where it was like more and more groups would come forward and say, Hey, can you put our MPS on this? Can you put our employee satisfaction survey on it? Can you put our like it just went on and on, and other groups would start to come forward with it? So, and then the can, you know, the customer service people would step in and sure, we can do that. So it just became a an absolute flywheel, is the best way to describe it. So I think that's one thing with SaaS companies, and I wouldn't say it's it's for everybody, but I would say there's there's ultimately kind of like this dirty little secret that they all know that uh a lot of their revenue is service-based, not purely platform-based.

SPEAKER_00

Speaking of consultants, a lot of them show up with the technology first. And I know that your approach is people first, which is which you already mentioned to me. Why does that work better?

SPEAKER_01

I wouldn't say it's necessarily better. I just think in this particular point in time, uh with AI, it's less of a technology challenge, and it's much more of uh of a people or a change management problem. So if you there is uh there's a company called Section uh in in the United States, which is run by a guy named Scott Galloway. And well, sorry, it was founded by Scott, it's run by another gentleman. Section does uh really training, particularly around AI, but they also do a report every year. They call it the proficiency report for AI. And they've run it for a couple of years. And the interesting about it, they just came out with the latest one. And really interesting is that uh 45% of organizations that they went out and looked at or surveyed or spoke to, 45% of the organizations, and you could probably extrapolate this across every organization, have invested in AI in some way, right? Mainly on applications, tools, maybe some infrastructure, maybe some data readiness, but it's really about the applications. But uh conversely, only 12% of the employees are actually adopting AI. So 45% of the organizations have it, but only 12% of the or of employees inside the organizations are really adopting it. That's a people problem, right? That's a training problem, that's uh an awareness problem. Uh, and and it's such a different way. I don't know, I'd love to hear your perspective on this too, but it's such a different way of interacting with, I get really, I get really excited about this. I feel like I'm a kid again, but it's such a different way of interacting with a machine that we've never had in in sort of history. So, like for decades now, the way we've interacted with computers or iPads or whatever, uh it is, you know, file, open, and then we start to whatever, use the spreadsheet, build a PowerPoint, do whatever we're doing inside of a CRM, whatever it is, right? That's how we've interacted with computers, right? Uh, even writing code has been very similar as well. Now we have this interaction where it's just tell me what you want me to do. And and it sounds like, well, that's easy. People have, you know, been managers and said to other people, like, hey, can you go and do the following things for me? Here's the list of to-dos, and you get them to do things, right? But not everybody is a manager, you know, like when you look at organizations, right, which brings up a whole other problem, which we'll get to a little bit later. But most organizations are made up of like 85%, maybe even 90% of the organization are executors. They're doing stuff, right? You might have some managers, supervisors sprinkled in there, but that's only really like maybe 10, maybe 20% of the organizations are actually managing people. The rest of the organizations are really executing, right? So it's not a natural thing for people to do, is to start to, you know, like uh actually speak to this thing, as particularly a machine, and tell it what you want it to do. And it, and in fact, we're really not very good at it. Like explaining to uh particularly AI about how to do something is a lot more comp the prompting is a lot more challenging for people than I think anybody really expected, right? So to truly get the most out of it. If you're just doing like kind of replacing it for search functions, sure. Hey, where are the best places to eat in in Lisbon? Well, you could use Google search, but people are using AI all the time for those kinds of things. So it inherently it is a people challenge, not necessarily a technology challenge. There's obviously problems from a governance perspective, and there's obviously society, political government. There's a lot of challenges related to AI being adopted, and it's going to change the way we we live, work, all those things that people talk about. Until we get past this, that sort of, I'll call it sort of the application gap, that 45 to 12 percent. And that's where really we sit at 2044. We're trying to close that gap, really.

SPEAKER_00

I've never thought about it in such a way that using AI is basically being a manager of a digital coworker of another digital colleague, so to speak. And that's so interesting. 45% of organizations are invested in AI versus 12% only using it in their operations.

SPEAKER_01

And the interesting about that too is like so out of the data we've collected with the many organizations we work with, because we run um an assessment, that's the platform we're building. We're building basically a platform that anybody can access and run their own assessments and audits on their on their business and kind of get a scorecard. But even our data, a year ago, we were seeing a trend closer to like 20% within organ, like employees applying it on a daily. In fact, some organizations were even getting up to like 50%. Now, were they applying it effectively and were they actually getting some return on it? That's a whole other sort of discussion. But but at least they were starting to apply it. It is dramatically dipped, I would say, kind of late last year, and particularly into this year. So part of it was, I think, you know, that classic hype cycle that Gartner has. I think we were just, we were climbing that sort of hype cycle very quickly. And now we are definitely down into the trough of disillusionment, if you will. So, which, you know, we're gonna come out of that and we're gonna get into the, I think they call it the slope of enlightenment or something like that. But it's gonna take time. And again, trying to close that application gap is really something where it's the big challenge. Like right now, I think if I think of kind of the place that we really try to focus on, we try to keep it really simple, where it's it's moving from hype to habit. That's what we're really trying to ultimately do. That's change management. That's really what it is. And the interesting about that, too, is I'm probably gonna get in trouble for this for a bunch of people that are maybe in IT or technology that are watching this, because obviously it'd be a lot of them. But a lot of technology or IT people are not particularly good at the change management part of it. They might be good for in their own worlds and their own, but you think about trying to, if they're charged to get a system, right, like a CRM system adopted inside of an organization, it's usually not, it doesn't fall on them. It's somebody else that it really falls on to actually do that transformation or do that change management piece. In this case, it's sort of like AI is sort of just like thrown into their because it's technology. So it's kind of thrown, and it's like go get it. And then they can't seem to figure out how to get this this close this adoption or this application gap. And I think it's because they're just not thinking about it the right way. They're taking it from a, oh, this is what we've always done technology. This is how we've always got people to adopt new technologies in the organization. This is very difficult. Different. It's it's just it's radically different from the user experience, from applying it to workflows, getting people to actually be able to truly get the most out of it, um, understanding the risks associated with it. It's just very different. It's very different.

SPEAKER_00

Steve, you've lived through different waves of digital transformation. And I wonder how does the AI adoption wave compare to the previous ones, like cloud or SaaS? What is what do you think is different and and what is the same?

SPEAKER_01

There's probably more that's the same versus different, um, believe it or not. Um you know what? I maybe I'll maybe I'll take that back a little bit because there is we just talked about a lot of things that are different, but let's talk about things that are the same. So I am a firm believer in uh 20-year innovation cycles. So I'm sure you're aware of that. In fact, actually, the reason we call it 2044 is the company was founded in 2024. And the whole premise behind it was okay, well, it takes about 20 years for some of these new technologies to be adopted. So we'll call the company 2044, and that'll be its last year of operations because you know, like there's a lot of theories out there and models out there that say like the world will adopt AI, I think somewhere up north of 85%, right? So probably even going to be higher than that. So it's like, okay, we'll do 20 years. So, and and you kind of see that playing out right now. And again, some people will definitely argue with me that no, it's happening faster and it's and it's bigger than anything in the past. And sure. But for the most part, we're right in the middle of that, call it sort of what is this? And that excitement and that hype phase and experimentation, and it's just overinvestment. And, you know, look at the stock markets or, you know, we're kind of in that sort of phase where it's everybody is super excited about the possibility of it. And quite honestly, we're just seeing like this very early days of the possibility of what this could look like and be. And then we're starting to see how this is going to translate into physical devices like robots and things like that, where it's not just on your phone where you're using, you know, a general LLM to answer questions or help you with strategy or something. But now it's gonna translate into uh some kind of machine that's actually going to, as my wife puts it, do my laundry, please. So uh it, but that's all happening in its definitely, but we're still in this very early days of experimentation, right? So that's usually about the first five years. Well, we're kind of right in the middle of that. Maybe we're maybe a little closer to the tail end of that. And then we go through this long period of about 10 years of really just organizations trying to figure out how we truly operationalize it. So we kind of get through something, we realize, okay, yes, we maybe have overinvested, uh, went too far here, went not enough over there. And we go through this operationalized phase, right? And that's where a lot of automation starts to happen and those kinds of things within organizations. And so that's usually about 10 years. And then the last five years is where technology almost fades into the background. It almost becomes invisible. It almost starts to become this thing of like, I don't know about you, but I can never remember not having a smartphone. But that was only 20 years ago, right? So uh it just becomes like uh electricity or water out of a tap, right? Like we just don't ever remember not having that, right? So now is the is AI going to uh buck that trend and accelerate it? And it's not gonna be 20 years, it's gonna be less. I think in some places, in some situations, you're seeing it right now with consulting firms, prof any professional services firms. Like if you look at law firms, particularly there was a big study that came out, and I would argue, I think general contract law was already transforming anyway. Like you can go online and find a pretty simple contract or even like a will and put your name in it. And, you know, so so can AI sort of transform professional services business a little faster than other places? I mean, there's a lot of question about, you know, what's gonna happen with kind of white-collar worker. There's a there's a great study out there right now by Sinterest that's talking about sort of this whole paradigm that's gonna happen. And it talks about it's gonna be in the next sort of five years this happens. So is AI gonna be transformed faster than other technologies like cloud or SaaS? Maybe. I've got my money bet on. No, it's it's gonna take time. And the reason is, as I go back to what we talked about before, uh, this is a people challenge, and we are humans and we have a way of adapting new things, new technologies, and it seems to always play out this way. So, because of that factor, we are going to push and pull. We're going to, governments are gonna be hind and putting regulations in place, uh, companies are gonna be hind and putting policies in place. They're gonna, IT is gonna turn these things off and say you can't use those things because of these purposes. And it's just going to take time for everybody to start to really sort of see how this adopts. And also, too, the technology is evolving at like just ridiculous speeds. Like how much further ahead the LLMs are or just AI applications are from what people have as actual use cases. Like when we do our assessment, I would say probably 80% of when we ask people how you're using it, probably higher than that. I mean, the guys would probably have the actual stat for me, but it's somewhere in the neighborhood of 80 to 90 percent. The use cases are it's like an advanced search engine for them, or it wrote me an email. These are such basic functions, they're missing the point. And it's because they just don't know necessarily how to actually true, truly use the technology. So I think it's the same from that perspective. I think what's potentially different is what we talked about before around when when particularly ChatGTP when it came online. Because I I remember um I I really remember the day someone texted me and said to me, uh, remember that thing we were talking about, the generative AI thing? And uh, well, it's here, check it out. And the first thing I did was I pulled out my phone, downloaded the app, and I started using it. And then the stats started coming in. They were reporting how fast it was being so like the fastest technology in history to hit 100 million users, like it took a week or something like that. Again, this is because it was built on top of it. But it really struck me. I remember going home that night and I was just fascinated. I was just using it for all kinds of things. And then I sat back and I went, how many smartphones in the world? How many people like it's one of the first technologies? It's not a steam engine. Not everybody had access to trains, not everybody had access to planes, not every like this thing became accessible overnight. It is one of the most accessible technologies. I I would argue it's probably the most accessible technologies in history. That's different. SaaS, like not everybody had a Salesforce license, not everybody had a, so there was there was, you know, IT had to get involved. Well, I guess they have to get involved a lot with AI as well, but they had to get involved to turn these things on. Like, if you got a smartphone, you got access to the world's information in a way that you've never had the world. Like, this is not Google. This thing is thinking for it's a companion, it's it's helping you do things that you never imagine. I got, you know, like take a picture of your fridge and tell me what to make for dinner, and it gives you a whole recipe and tells you what to make, tells you how to do it, shows you a video. Like, it's not the same. It's not the same. And that was literally almost overnight that happened. So that's really different.

SPEAKER_00

You mentioned the journey from hype to habit. So I'd be really curious to know what happens in between of those and what specifically 2024 does for the clients to make this adoption happen. Now you're talking about something I really want to talk about.

SPEAKER_01

Um yeah, so uh again, this comes out of years of consulting and and even a lot of transformation and change management inside of organizations. And I go back to, I remember even when social media came out, there's a lot of parallels and similarities to, well, mobile, internet in general, to get people to really truly adopt it and understand it. But the thing that and really what we do is we we have a three-step or three-phase process we take people through. But the thing I think everybody misses is the first step. And I think this is where everybody is is very quickly applying tools. And by the way, I think lots of companies have have developed like overarching visions uh or vision statements. Maybe they have a manifesto of some sort, maybe they have some operating principles, which I think are really important. I think it's actually more important to have really clear operating principles uh than it is in even a vision. And I'm I'm a guy who loves doing vision statements and things like that. But so lots of companies have that. But the thing they miss is they haven't diagnosed the organization. Like, how ready is the workforce and how ready is the organization? And I would argue a lot of people have actually figured out like our infrastructure's in place. We have, we have these kinds of, we're a Microsoft shop, so we can put Copile on its like they kind of figure that part of it out. The data readiness piece, I think everybody's realizing very quickly how that's a problem. Uh, but I think organizationally, you kind of have a sense of what sort of it needs to be done. But the workforce, everybody seems to miss the idea of assessing the workforce and really understanding how ready the workforce is to adopt this and being very realistic about it. So that's the first step in our process is we really diagnose the problem, essentially, is what it is. And we run, and this is honestly, I would say this is right now, like 80% of the work that we do is we run a workforce assessment. And it's a series of questions. But beyond the series of questions that we ask uh individuals and organizations, by the way, it can be a team of one, it can be a team of 20, or it can be thousands of employees. Like it doesn't matter to us. But the thing that we're really looking for is we're we're trying to really measure and understand particularly how good you are at applying AI. So we look at four different dimensions. So we look at knowledge and skills, right? Or your ability to actually turn uh functions into skills is a big one. General openness, which is interesting. I would have thought in this day and age, people are everybody's open to it. That isn't the case at all. But sort of general openness, the ability to apply it, that's really where we get into testing, like proficiency. So we give like uh through the assessment, we actually ask people, how good do you are you at prompting? Because that's still the problem.

SPEAKER_00

I was about to ask, is it a hard skill or is it a soft skill?

SPEAKER_01

Well, well, it's interesting because people always rate themselves, which is interesting, people always usually rate themselves low at things or lower than what they think they are, or most people do. In this case, almost everybody rates themselves way higher than they actually are. So they rate themselves like a seven out of 10. I think that's probably the average. And then we test them right afterwards. We give them some prompting tests, and then we give them back and we tell them right away, you're about a three out of 10, right? But it's important to know that. And this is again by diagnosing the organization and really understanding like, here you here's where you're at. Now, what do I need to do? What courses do I need to take? How do I need to apply this? What do I need to do? The other thing we test people too is like being able to spot and identify within their processes or their workflows where AI could actually support and help them. People are really bad at that, surprisingly not good at it, even though it's like, well, I'm doing this process day in and day out. And how come I can't figure out where to put AI into this, right? So I was talking to uh a colleague of mine, and he's a big fan of Manus. And he's uh, which is a great, you know, it's the Meta's application. But the thing it does really well is you can just like map out a workflow or a series of steps and then tell it to make it a skill. So it repeats it, right? So we've been trying to like organizations trying to create agents, and it's the same kind of thing. It's a set of tasks, whatever. But it's surprising, even though people will do that same set of 10 steps every time, it's surprising how hard it is for people to say, AI could actually do step four, five, and six for me. How come I'm not getting it to do that? And I think a little bit is we are stuck in this world where AI is, well, let's face it, it's still doing things like hallucinating, it's still doing things that that sort of affects the trust factor of it. Like I've I've been on the soapbox for a long time that AI has a huge PR problem. And it's a very North America thing, maybe even in Europe as well, where uh if you go to APAC countries, or particularly if you go to like Hong Kong and China, if you ask, you know, the population how they feel about AI, 80% of the population will say, uh, we're ready for it. It's gonna improve society, it's gonna do great things, we're gonna grow businesses, all these kinds of things. You go to North America, it's like 30% of the population say that. And you think about it. Every time you open a news article, everybody is like how bad these companies are. And to some degree, yes, I agree. Some of these organizations are doing some, you know, bad things, and tech companies and tech leaders have not done themselves a lot of service, but they talk a lot about how AI is almost evil. And then you look at movies for decades, right? And I'm not just talking about Terminator, like, name me a movie where AI has come out in the positive side of things, like it's always dystopia every time. And actually, I was thinking about it one day. I'm like, what's the first like time that like machines were really kind of talked about in storytelling? And actually, I go back to Frankenstein. Frankenstein was told as humans creating effectively artificial life. Well, what happened there? Frankenstein got a little pissed off and went out and like destroyed the entire town, right? Like so it didn't end well, right? So, and we just continuously tell these stories and we're fascinated by these stories. So, if you think about all the compounding negativity about it, of course, people are gonna be skeptical and they're not gonna have a positive outlook on how to apply it. But we have to switch that. We got to get into a better narrative that it's gonna have positive outcomes. Because if we don't, other people are just gonna create the narrative and they're gonna create the outcomes that they want versus the outcomes that you want. So we have to get a hold of this narrative. Um, anyway, back to the 2044. So we do that diagnostic. From there, once we've got a pretty good sense of where the workforce is at, it's pretty simple from there to really create a effectively like a roadmap of what you need going forward. And by the way, that roadmap often doesn't involve uh, hey, let's go out and get all these AI applications and start putting it in. I think that's uh often a mistake inside of organizations. You need some of them, but I think it's more about where do people need to be trained, workflow identification, helping them identify those opportunities. And then I think the big one too is really the ROI part of it. It's like we really are stuck right now where businesses still really struggle with like what's the investment you're gonna do. So one thing we do for organizations, we go through the workflows and we literally do a diagnostic on uh where you're gonna get the most return out of doing that. And most of the time it's efficiency plays, right? So, and then the last step in the process, once we go through, it's really that we talked about before, like the the hype to habit. It's really creating like daily habits. So we have a whole series of really sort of helpful things that you're applying it every day and you're getting stronger, right? So, like once you this is the change management piece. And I think this is the step that everybody starts to forget about. When you start to form those habits, it becomes one thing. So one of the simplest things I tell people to do is when you open up your browser, Chrome, Safari, whatever it is, what's the first tab that shows up? Right. Still 80%, 90% of people will say, well, often it's the company's homepage or something like that. But next to that, it's usually like your favorite search, like Google search or something like that, which is fine. That's what mine was for a long time. Switch that. Put your favorite LLM in there, right? Now, if the company may not let you do it, but watch your behavior change. Watch how you start to learn and apply it in very different ways, not just traditional kind of searching and looking for things. You start to change your habits every single day. So that habit forming is sort of the last stage that we do. So there's the diagnose, which is the assessment piece, which is a lot of what we do. There's uh really kind of the roadmap. And then uh the final one is really habit forming.

SPEAKER_00

And I definitely hear the compound effect here. Yes. I wonder how do we help organizations actually execute on the roadmap?

SPEAKER_01

We're not actually a development shop. So we often get asked, hey, so now we've got this roadmap of, and you know, uh there's usually like first phase if they don't have things like uh policies and the things we talked about before. Like if you don't have like principles and we'll we'll sometimes help them do those things, but more often than not, organizations have actually kind of thought though through those things. The big thing is like, and again, I was talking to my colleague about this the other day. It's like, how do you get into it's it's like it's the software mentality or software engineering mentality about like ship things, like quickly ship things, create things and quickly ship things. We seem to be kind of stuck a little bit when it comes to AI, where it's like get into a little bit of thing of, and some some of it's like based around usage, but some of it is like start to stand up these skills, start to stand up these agents, start to push them out there, get people to use it, refine it, make it better, right? So start to move towards that sort of space. So we typically will help organizations do that. And then we have partnerships with companies like yourself, like Pro Coders, companies like that, which would then come in and do development for us if there's like an application that needs to be developed. I would say though, what's really interesting is and some of the biggest successes we've had of organizations that, and this is really smart people, are well, let's not let's not embed a new chat thing inside of the organization, and people gotta now go open up a new tab or you know, go to this application, whatever. Why don't we just embed this inside of our existing applications? And there's a client we're working with, which it's not like it's a well-known branded company, but it's a company called Garland. Garland makes like high-end uh restaurant equipment, stoves, things like that for like high-end restaurants, but then one of their biggest clients is Tripulti. So they make all the equipment for Tripulti, right? So you can imagine how many restaurants they have. The thing that's leading, the person that's leading the AI uh adoption initiatives there, he's created an application, an AI application that lives inside of Salesforce for the customer service rep. So we identified it as a low-hanging workflow, like an easy one. The data was in good shape. Customer service, they actually scored really high on the proficiency score. So it's like that's the area they I think you're getting the most traction. But instead of creating a separate customer service AI application, he just embedded it inside of uh Salesforce. Guess what? People are using it because it's you're in a familiar area, right? Might be a little different in terms of how to use it, but you're in a familiar area, right? And it's something where people are getting a lot of use out of and seeing like this is changing my life. I go into this and I say, here's the problem the customer is having, gives me all the part lists. It even actually uh orders the technicians, sets a date because it looks on their calendars, sends all the information to the tech so they can go out. Like all of that was stuff that that customer service rep had to do before. That customer service rep is way more productive than they were in the past. Like I love that use case. Embedded inside of things that you're already used to doing, I think is brilliant.

SPEAKER_00

I really love that approach, Steve. You're speaking exactly about not creating a new habit, but integrating into the habit that already exists, right? Yes, that's exactly it.

SPEAKER_01

I mean, like the if you think about it, like we create these work streams or workflows, and you really forget about it that that's been operating for like 25 years inside this company. It sure, maybe we've got a new CRM system or a new kind of piece of software or new ERP or something that you've applied. Okay, I get all that. But, you know, really the process has been the same for 25 years, right? Like client calls, you pick up the phone, what's the problem? Enter it into a database, call the tech, send them out to the thing. They don't know where their address, call them again. Like there's this process that we go through, and that's just how we've always done it, right? And I think that's a big thing too. Like that's a manufacturing, which manufacturings are famous for. They get into habit forming, right? And for good reasons. Like they do things systematically because they need to, right? Like there's all kinds of rules and regulations around what they're building, and uh it has to be built to a certain spec. Like, you know, like you don't want the stove catching on fire and burning down the place. That's bad. It's bad for the restaurant, bad for the people inside the restaurant. So, so changing the way you do things, uh, and maybe there's compliance, that can be challenging, right? Like banks, uh, healthcare companies are really governments are really challenged with AI because it changes the conventional way of how you've done things in the past. Well, these are organizations that are heavily regulated and there's massive compliance. Like Europe is like heavily regulated around so so all of a sudden you're saying, like, hey, just start to adopt this thing. Well, if it's not inside of an existing thing, it it starts to become one of these things where it's like, well, we need to now completely change the way, but in this case, like, no, no, don't change anything you're doing. Do it the same way. Just click this button versus that button. That's the only change you have to do. Wow, that's that's it. Yeah, that's it. So, uh, and I'm really oversimplifying this a lot, but that we've seen the biggest effect on organizations and people like you just you just see adoption rates, that application gap I was talking about, like from 12 to 4, you just see it start to spike inside of organizations. And by the way, the one thing back to the assessment thing or benchmarking, it's not a one and done thing. Once you set a score or benchmark in a company, right? Say you go out and do that application, you embed it in things, you need to then train the people and then rerun that benchmark. So you constantly are trying to upskill, you're trying to constantly get like we were a 38, now we're a 62. Okay, next is to get to an 85, like that adoption. And the only way to do that is you need some kind of way to assess and benchmark the people inside of an organization. It's a huge missed thing inside of organizations.

SPEAKER_00

If I'm an enterprise leader watching this right now, or if I'm one of the company directors or a manager of a company division, what is the first step I should take to assess my team's AI readiness?

SPEAKER_01

Uh well, hire 2044. Um, yeah, I I mean, look, we're not the only folks out there running uh some kind of readiness assessment, but I would say start with like we're gonna set a benchmark. That's the first thing we're gonna do, right? So uh again, this is that you've got some baseline principles, we've communicated it, we've got some policies in place. Like that stuff is table stakes. You need those things in place. If you haven't done that, you need to kind of take a step back and get some of those things set up. But then really to go back to the workforce, you really need to sort of like, let's set a benchmark, let's let's score them, let's get a real good understanding of how ready the workforce is to adopt AI. And by the way, don't just do it with a series of questions, like a survey, like an assessment or audit is very different than just a survey of questions. Survey of questions is self-reported data. Hey, how good are you at uh at writing a prompt? I'm a seven out of 10. Great, they're all ready to go, right? Test people. You have to, in this case, you need to evaluate people and test people. So that's the only way you're gonna actually get an actual behavioral score. We're actually working on something else too, where we want to um embed uh live stream data from the application. So say they're running Copilot, for example. Uh, there's one thing us saying, like your score is whatever, a 30 out of 100 kind of thing. Fine, that's your score, but what's your usage behavior like? So we want to be able to bring that data into our overall scoring system as well. Because you may be using it a lot, but your proficiency is really low. Like you're not very good at using it, but you're on there every day, right? So, so that tells me, uh, well, obviously you're trying, but you're you're really not uh being taught the right way to do it. It's sort of like it's it's back to the habit forming. It's to make sure you're teaching them good habits, not bad habits. And so we're starting to see a lot of that creep in as well. But the only way to really understand that is you have to start to embed usage data into it as well. For sure, start with like benchmark your team. Like, where is your team at? And then do it again, like benchmark, figure out what the score is, what do we need to fix, train them, teach them, reset, like in other words, do another benchmark. Do we did we move the needle up? Did we move it now? Where are we? And just continue that process again and again and again. That's the simplest advice I give somebody.

SPEAKER_00

And at 2044, do we have a platform or a product helping you doing this?

SPEAKER_01

Yeah. Yeah. So we um, it's interesting. We uh um a little plug for uh the podcast that you and I did with my other colleague Ian. Um, so we were talking about vibe coding and lovable specifically. So um, yeah, one of my colleagues at 2044 has been playing around with vibe coding uh almost from day one and uh and started to develop a platform for ourselves that literally is going to run the assessments. Incredible technology, it really is. Like this whole vibe, I can't speak highly enough about it. I get there's a lot of issues around production ready and those kinds of things. But uh anyway, yeah, we created a platform. We call it T44. We also have something called the AI44 as well. So essentially the platform allows you, it's not SaaS ready yet. We use it for our like a white glove service, like it's our own platform. We can go into the platform, we can adjust things like you know, of those four areas around the workforce. So, say, for example, we feel that ability to apply AI is more important than, for example, how open people are about AI, which I'm not sure why they would have one, but let's say that was the case. We can then have uh we can change the overall ratings or the index across those four categories and weight it higher. So we can say this is worth 50% and the others would be weighted, you know, divided by three kind of thing. So that's one thing. The other thing, too, is we do within the platform is we've developed where it'll actually provide you the recommendations, like the to-do list. So you score 30 out of 100 or whatever as an individual. Maybe the company scores 42 out of 100, whatever. Well, what do I do about that? You've given me a score. What do I do? So we can come back in and consult with you, or you can kind of look at it and sort of figure out what you need to do. But what we've provided is some specific tailored recommendations to you as an individual. And then those recommendations can actually be the company's recommendations, meaning, so say you've already got AI adoption leaders inside the organization. Some have that, lots of big companies have that, and you've already developed training programs. So Accenture has extensive training programs. What we can do is we can point you as an individual and say, okay, you're really stuck with the thing you scored really low on is identifying workflows. So we have a whole course, a whole training section for you to learn. We can point you directly at that and say, go take course 268 and pass this and then rescore yourself or retake the test to see if you've actually passed that. So we've built all that really flexible functionality within it. And then the final thing, too, is because we have like a set of variables and questions that we ask, we're able to sort of store, because we've done a lot of these assessments, we're able to store that and index you personally or index the company against other organizations within our database. So we can say companies like you score a 62 and you guys are a 42. I gotta tell you, there's nothing that gets a CEO's attention faster. And it's like, what do you mean we scored less than our competition? Right. So um it's almost gamifying it a little bit. But it's just a way to say, like, where do you fit within sort of the overall uh sort of ecosystem as well? So yeah, that's where back to what we were talking about before, around like we kind of really are moving from being a pure play consultancy to to really having our own sort of assessment platform. And ultimately for us too, I think it it's gonna be ultimately a behavioral database where we're gonna have, you know, like eventually a lot of data around actual changing people's behaviors within our like how did we move somebody from a 32 to a 62? Well, we'll have all that data to show this is what they did, this is how they did it within our database.

SPEAKER_00

Do you think this platform will become a DIY tool one day or SaaS? We've been having a lot of conversations about that.

SPEAKER_01

Uh, it's a great question, by the way. You know, like again, kind of going back to what I said earlier about my passion is about building things and the it, you know, like you look at these things as annuities, right? Like they're something where it just continuously sort of flows money into the organization. And that's and and obviously the more hands-off you can be, the more SaaS-based. I don't know if we're prepared 100% to get to the point where it's going through, because again, I've built platforms going through all of the uh, you know, the SOC2 and like going through all that, maybe, maybe. We're not ruling it out. I think it needs a bit of white glove service to it. I don't know if you can uh right now totally just hand it over to somebody and say, like, here's the here's the application. Also, too, I would sort of like take it one step further, like some of the training applications, we're actually building training agents to teach you how to prompt better. So instead, if you don't have a uh a prompting AI course in your company or you don't know where to point, we've got a little AI application that or AI agent that can teach you how to prompt better, right? So that'll all be built into the platform. Well, in order to really do that, we probably have to get to a SaaS platform where someone can log into it, but we're not there yet. So do we want to get there is the strategic question. So I'll let you know in about a month.

SPEAKER_00

Okay. I actually have a rapid fire around for you. Excellent. Ten quick questions. You don't have to overthink, you just pick one of the one of the options. Okay. Operations or strategy? Strategy. Exit the business or cube building. Can I say depends?

SPEAKER_01

Nope. Uh oh man. This is not rapid fire. I'm gonna say I'm gonna say exit. Claud or chat GPT.

SPEAKER_00

Claud. Clot or base forty four. Hire developers or white code.

SPEAKER_01

Depends. Um acquired cottage or city.

SPEAKER_00

Depends. Uh probably cottage. Podcast in the car or in the gym. Both. I would say the car more. And now a few Portuguese ones. Pastel de nata or pastel de bacanal. Second one. Because I look messed up actually trying to say it. Espresso at 10 p.m. or port wine at midday. Oh my gosh. Uh both.

SPEAKER_01

Um my gosh. Probably the espresso. And that was a little bit you. That's a little bit you. Like, I'm not a I'm not a coffee after, like, uh, I do my morning coffees and and I'm good. But yeah, you you'd had the expresso that I'm like, okay. And and actually every night I've been here, I've done that. So uh and I've slept like a baby.

SPEAKER_00

So yeah, espresso. We did it in a very Portuguese way because here in this country, dinner starts at 9 p.m. And it's not common to be always on time. So that's keeps you awake. That's why you drink it. It's just it's just a part of the culture there somehow.

SPEAKER_01

Somehow picked. Yeah. I have to say just one thing as well. Uh, just that other night when we went out for dinner and uh had a just a lovely time. But uh first time ever having barnacles. Um yeah, uh, that was a great experience. They were uh it it doesn't sound delicious, but it is delicious. Yeah, I was I was really surprised. Yeah, excellent. Thank you for recommending that for sure.

SPEAKER_00

I'm glad you liked it. Steve, thank you so much for joining me on the pod today. I was really pleased to have a conversation with you, and it was great to hear your perspective on the AI adoption and how technologies are evolving.

SPEAKER_01

Thank you very much, George. It's been an absolute pleasure. Uh, you are an incredibly gracious host, uh, and we definitely need to return the favor and get you over to Toronto, Canada. Absolutely. Thank you so much.

SPEAKER_00

If you enjoyed the episode, hit like, subscribe, and drop a comment. And by the way, check out more Backstage Tech episodes on my channel. If you'd like to join as a guest, send me a DM on LinkedIn or email at george at georgehelgison.com. Thanks for watching, and I'll see you in the next one.