All Things Digital Transformation – Digital CxO Podcast EP76

Hello and welcome to the digital CXO podcast. I'm Amanda Ani, and with me today is Mike Baard. How are you?

I'm doing great. Happy to chat as always. Same.

com website, we have an article by Nathan Eddie. This is about an Infosys survey and there's some interesting information coming from that. Less than half of contracted cloud services are being consumed with the majority of CPGN retail respondents struggling to monitor cloud costs.

So what are you hearing from business leaders out there as it regards to cloud cost? Cloud computing is still too damn hard. That's the problem.

And fundamentally it requires developers with, uh, infrastructure expertise and a lot of them don't wanna have that and they find the infrastructure gets in the way, or I need to go find a DevOps team and a bunch of software engineers who can manage that cloud environment on behalf of those developers. And, and I'm hoping this is gonna get better sooner with the help 'em ai, but, um, right now, I mean we all talk about like, oh, we're just gonna move everything over to the cloud willy-nilly, and A, it's hard to move stuff into the cloud and B, it's hard to manage and when you get there and c it gets expensive in a hurry. A lot of folks I talk to are also repatriating some of their workloads in the cloud, not because they don't like the cloud per se, but they did the math and they found out that sometimes the on-premises environments are cheaper to run over the long haul.

Um, back in the covid days, there was no other decision to be made, so a lot of workloads wound up on the cloud when maybe they shouldn't have been there in the first place. So the world's becoming more federated. This is all relevant to digital CXOs though in this regard.

Right? Uh, most if not all of these digital initiatives depend on some sort of cloud computing capability. And if it's too expensive and we don't have the people to manage it and we don't have the skills to uh, invoke it, it limits the pace of the digital transformations.

And this is, uh, becoming a, a fundamental problem. I mean, at its core it's really an IT problem, but the impact on this is all on the pace of digital transformation, which is probably not going as quickly as everybody would like. Yeah, and it's interesting because, you know, all I heard back in the day when everybody was migrating to cloud was how much more affordable it would be, but I think they're now sitting back and realizing there's certain things that would best benefit from not migrating to the cloud, whereas others would.

So it depends on the reason behind it, and some things would be more practical to stay put. Yeah. Your mileage will vary by workload considerably.

So you need to really think through exactly what's happening with the workload, what the requirements are. Um, there's a tendency to kind of just say, you know, if you're a digital CXO, we're gonna default to the cloud. Um, truth of the matter is more data is being consumed at the point where it's being created and that's pushing more compute out to the edge.

And uh, a lot of the digital applications that you're trying to drive need to be running near real time and soon have AI models embedded in them. And all that stuff is gonna happen out at the edge. And there is this thing, it's called network latency and the laws of physics.

It has not been suspended and it still impacts the decisions as to what work will to run where. Um, and I don't think digital CXOs need to be an expert in all that stuff, but I firmly believe they need to have a good appreciation for what's going on here. Yeah.

And a good team behind them that they can collaborate with and, and can kind of keep them up to date on everything. Yeah. Unless they just wanna pay the bill for everything no matter what, which I think is probably the shortest path to unemployment.

So moving on, you had an a, a interesting interview with Ryan Matzner. He's the head of fueled about, uh, AR and VR goggles and how those are gonna be integrated into workplace. So can you share a little bit more about that?

Yeah. His point was just that by all new technologies early on, it seems like it's expensive and who's gonna use this, but the price points for these things will come down. And if you're really thinking about this in a one year to two year timeline horizon, if you wanna go build some sort of augmented reality type of application, you need to start working on that now because, uh, with the assumption that the cost of the infrastructure is gonna drop to a point where it's more affordable.

But if you're gonna wait for the price point of the goggles to reach something before you start co building these things, you're gonna be late. 'cause everybody else is already starting to experiment with this stuff today. I'm not saying that, you know, you should put this at the front of the, we're going in production tomorrow list, but, um, it's pretty clear that these goggles are gonna have a bigger role.

I mean, you know, he's talking about two things. One is gaming, that's kind of obvious, but I know, yeah. Am I gonna use these goggles to manage, you know, business processes?

Probably not. But are there tons of, uh, digital simulations and, and all kinds of interesting things that can be done on the engineering side and field support and all kinds of things that, you know, today we are still trying to do, uh, with a heavy manual lift that these things will help bottom me, I think the time is now. And so his point is, hey, you expect to be in this game in a year or two, maybe you start playing with this stuff tomorrow.

Yeah. And I mean there are so many use cases, for example, in the training realm. So, um, I know I went to a university here a while back and got to experience healthcare training via, um, VR goggles.

So there's a lot of use cases coming down the road, Right. So I'm curious, but uh, you know, how did they simulate the grumpy patient who doesn't wanna be touched? Yeah, it was really interesting.

Like you felt you were in the hospital room and there's the patient in the bed and all the, the equipment and uh, it was interesting. So Sounds beautiful to me. Yeah.

So moving on, um, uh, you're, you had another interview with Ugo Ori of Digitate and this one's talking about how middle managers really need to change the way they're managing in this day and age of ai. So can you explain more? Yeah.

I mean this is kinda a common sense and b there's not enough common sense out there. So it seems to me when I go talk to folks out there, there's a bunch of end users who are randomly using AI to test stuff and play with stuff and create stuff that's interesting. And then there's a bunch of, um, c-level execs who are, you know, basically telling boards and shareholders that AI will transform the world tomorrow.

Somewhere in between that is the rank and file middle managers, you know, otherwise known as the people actually get stuff done. And they're the ones that kind of have to step back and say, well, how are we gonna reinvent these processes? And how, where will AI fit in?

Where are all these so-called digital assistants that are gonna become, everybody's coworkers gonna fit in alongside their humans. What exactly are the humans gonna wind up doing in a workflow that adds value? This battle's gonna be won and lost at the middle management layer.

It's not gonna be won at the board level and it's not just gonna be won by a bunch of random, uh, employees, you know, playing with stuff, right? And it needs a plan, it needs an execution strategy and ultimately that lives and dies with the middle managers. I think everybody who's a digital transformation officer at this point has already figured that out.

They've experienced it firsthand. They probably had some project go awry simply because the folks tasked with executing it. Really this thing is poorly constructed and doesn't really match up with anything that we are doing today.

I think the same thing may be happening with ai. If we don't get those middle managers involved today, it's not gonna happen tomorrow. Absolutely.

And I mean AI is really, uh, affecting all processes and, and all departments and probably all levels of management honestly. Uh, as it rapidly advances and goes into every industry, Hey, but if you are a middle manager and you aspire to be something more AI might be the best thing that ever happened. So next up we have, um, an article by Adrian Bridgewater.

This is, um, on digital CXO as well, and it's about, uh, data GPT democratizing the use of data analytics across the enterprise. Yeah. So what are you hearing about that?

This is an example of a use case that's pretty compelling. Uh, part of the problem with all the analytics stuff that we've invested in over the years is that it takes a small army of people to set it up and uh, and then the people who consume it have to have a really good understanding of what that's in that dashboard and what it's telling them. And um, it's kinda, you know, for lack of a better phrase, it's still a glorified canned report.

Um, chat GBT and these other tools allow you to interrogate the data without having to be a, you know, master SQL jockey or whatever it is. You can just start asking questions and we're looking at the democratization of analytics and we should be able to make better and more informed data-driven decisions as a result, assuming of course that the thing that was used to train the LLM was reliable data in the first place, which is sometimes a big assumption, but, um, I think we're on the cusp of this. And so Adrian's article is one more example of, you know, of a company and a product and a moving us in that general direction.

There's gonna be a lot more of this stuff. I mean, you can just look around, it's pretty much, uh, it's already here. It's just not evenly distributed just yet and it's maybe not perfectly optimized, but I think I'm looking forward to an age where, you know, instead of bolting on analytics or or running it in after the process, it's in the process and it's part of our everyday work life just because we can type literally natural language that says, Hey, what's up with this process here and what percentage of this is gonna happen by Friday?

Yeah, I mean I imagine that would make things so much easier, but like you said, it's gonna be about accuracy. I know some people have been using it to say write some code for them, but the code doesn't always work, um, properly. So, you know, it's helpful, um, as a tool and I'm sure it'll get better and better.

Uh, but it sure would be nice just to ask and have that information right away. And here's the thing that will happen. The machines are gonna teach us how to ask questions in much the same way that search taught us how to query data and look for things.

As I play with all this stuff, it's gonna become apparent that the difference between one user and the other user is their, their ability to navigate the logical flow of asking questions and getting the prompts together in a way that is compelling and interesting. Too often we sit like we're watching television when we look at analytics and we're just waiting for, you know, the thing to unspool in front of us, but we're not directly engage with that data and we question the data all the time, but we have no way of playing with it to really get at the veracity of it. So I think we're gonna be moving to a time soon where, uh, they say that the sign of real intelligence is not knowing everything.

It's your ability to ask good questions. I'm hoping we're gonna have a little leap in humankind here. Absolutely.

And the more that people will play around with these GPTs, the more they'll refine those prompts. Mm-Hmm. Yeah.

And I'm hoping that we don't get too dependent on prompt libraries, right? I mean, there's a bunch of things you can do to automate a set of workflows and whatever, but there is no substitute for playing with this yourself. Yes, exactly.

So now we're gonna move on to an article from Ed Waddle, I believe is how we say his last name. And this is on tech strong ai and it's about, um, AI for social good and how there's so many great aspects of AI and healthcare for sustainability purposes. And so yes, there are pitfalls in bias and data privacy, but there's um, a balance there that outweighs as far as social good.

Yeah. So what are you hearing? I mean, AI is a tool.

It happens to be one of those tools that cuts both ways. Um, there are issues, there are threats to jobs and industries and everybody's cheese is moving, but there's a whole raft of things that are gonna happen or that can happen now that will make our lives better. And, uh, we should not let our fear of one preclude us from looking at the positive sides of the other.

And, um, more and more we should play to my earlier point, um, a lot of things that we previously would not have considered doing for the good of the order, as they say. Um, we can now do because the lift isn't gonna be that heavy or it, it's more attainable. And maybe we should just take a step back for a minute and say, Hey, is there a list of things that we could do for our fellow humans that with a little help from ai that would be good for everybody?

And, uh, you know, I liked this article just 'cause it kind of gives us a list and starts thinking to people in that direction. Yeah. I was watching a show the other day about these people that got stuck, um, deep down in a cave and you know, when they try to send, um, rescue, they're putting their lives at risk.

Um, it's very difficult. So I foresee things like rescue missions with AI and attached to drones, um, collecting data and being able to just go into those dangerous situations. Um, so I've seen a lot about AI and drones technology doing a lot of things.

Yeah, I see the same thing. I mean, if you're in the fire department or the police department, you know, we're asking you to assume a level of risk that, um, you know, your loved ones would probably prefer you didn't. So if there's another way to do something that involves using AI and robotics, um, by all means that, you know, I don't think we need to keep putting people in harm's way, especially to rescue somebody who may have done something silly in the first place.

I agree. Well, that brings us to the end of today's show. Mike, do you have any last thoughts to leave our audience with?

Um, just, you know, don't be afraid. Uh, ai Yeah, it's scary and there are things going on all over the place that, you know, can be problematic. Just remember, it's a tool, it's probabilistic, so don't have too much faith in it, right?

It's, it's a machine that's guessing based on a bunch of data that it has about what the next probable thing is. Um, as such, it is a, an enthusiastic helper, shall we say. Um, sometimes overly enthusiastic.

So, you know, you need to keep an eye on it, you need to teach it and train it. And if, if it's easier for you to give it a name, by all means go ahead. But frankly, uh, it's here.

The question is what to do with it. That's exactly right. And I wanna thank our audience for staying tuned to our show every week.

And please share in the comments what's the most interesting to you. And until next week, have a great day. See you later folks.