Innovating Leadership:
Co-Creating Our Future

Resources:

You can contact Greg for more information at https://www.linkedin.com/in/gsmoran/ .

Our host Maureen Metcalf posts a newsletter every week on LinkedIn. You can subscribe here.

Maureen’s latest book is Innovative Leadership & Followership in the Age of AI. You’ll find details about it at https://bit.ly/LeaderInAI, or check out the Kindle version at https://amzn.to/44buVz8. The audiobook version is now available at  https://amzn.to/4dTCleZ.

Her other 10 books are available on Amazon here.

Binge listen to more wisdom from Greg:

– Three Steps for Successful Negotiation

– The 7 Leadership Mindsets You Need Right Now

– How Missing Ethics Makes a Mess of Organizations

Guest(s):

Guest(s) Bio:

Greg Moran is a director, founder, advisor and operating executive with extensive global operations experience (U.S., Europe and Asia). He has a strong market focus with deep technology experience forged at business giants such as Ford and Chase which he now applies to organizations of all sizes – start-up, scaling, restructuring, sales, and finance/operations.

Transcript
(auto-generated)

Maureen: [00:00:00] This is Innovating Leadership Co-Creating Our Future. I’m your host, Maureen Metcalf, the founder and CEO of the Innovative Leadership Institute, where we help leaders become future ready. Today I am delighted to have Greg Moran joining us. Greg is a coach. Executive advisor and board advisor, and today we’re gonna be talking about how AI is evolving and impacting organizations. So Greg, you’ve mentioned that over the last few weeks how app, the app dev world has changed dramatically. Can you talk about those changes and why it matters?

Greg: Yeah, absolutely. Uh, it’s interesting when you, you look at the, at the arc of technology evolution, there are often things that sort of have, have been out there as a promise for a really long time. And then when they arrive on the scene, they can arrive kind of suddenly. And that’s a little bit of what’s [00:01:00] happened here.

We’ve talked for years about the idea of using AI and AI assisted coding, uh, to really help accelerate the work of developing full functioning applications. And there’s been little bits and pieces of it, and there’s been fits and starts and some progress made, but in the last three to four months.

There’s been rapid acceleration as the AI companies have really shifted their focus from building out better models to, uh, building out application ecosystems off of those models and working closely with other companies that are developing tooling to help automate, uh, the development of applications.

So. Specifically some things that have really been game changers. Uh, one model context protocol servers. So these really hit the mainstream late last year, and now they’re, they’re, we’re seeing MCP [00:02:00] servers delivered almost on a daily basis. And easy way to think about a model context protocol server is that it’s an abstraction layer between, uh, an application that you may be trying to write.

The various APIs that you may wanna integrate with. So, uh, think for example, if you wanted to build an application that knew how to talk to Salesforce, knew how to talk to, uh, uh, Oracle databases, knew how to talk to ServiceNow, and knew how to talk to, you know, any other, uh, set of commercial applications that we all know and love that are in the cloud, an MCP server.

Already knows how to interact with those APIs, so you don’t need to know that anymore. You just need to know how to talk to the MCP server and the MCP server will translate your prompt or your request into the calls that need to be made to all these various different services that you’re interested in.

[00:03:00] Interacting with that alone is a huge game changer. And then when you combine that with. Automated tooling on the front end that allows you to simply, uh, do what’s called now Vibe Coding, where you’re creating a thoughtful prompt for your application development environment. Usually tools like Cursor and there are many others that can just take a prompt and they’ll generate all the code for you using various different models like those from, uh, anthropic, those from an open ai, et cetera.

So now you’re sitting there and you simply put in a prompt and within seconds you’ve got all the code and you can test that code in real time and then you can go back and readjust your prompt and make it a little bit more specific or tune the outcome and you end up with in the space of maybe an hour or two.

A really [00:04:00] good working app that you can then sort of do the things you need to, to harden it and get it deployed, uh, as an app in the world. And, uh, that’s a real game changer. I.

Maureen: I, and can you just for context, for those of us who haven’t developed apps, how long did it take previously and what kind of funding? Level of effort, cost.

Greg: Yeah, you’re pointing to one of the things that I think is gonna be. Really rethought, and that is the whole concept of seed rounds in the, in the tech startup world, right? So the concept of a seed round was typically that you had a great idea, but you needed enough money to hire some programmers to develop MVP or minimum viable product, something that could be tested that you could potentially.

Try out on a couple of likely customers and see what they thought of your [00:05:00] idea. And that oftentimes is months and months of work. So you might have a seed round of a few hundred thousand dollars or even a couple of million dollars to, to get your MVP built, I think. I think that kind of goes away in a lot of cases.

Not every case, but in a lot of cases you’re now in a situation. Where you don’t need a seed round, you can bootstrap the company to the point where it’s at MVP and simply be looking for investment to do the things that you don’t know how to do or don’t have the capabilities, uh, and network to do things like setting up distribution for your app.

So it really accelerates and changes the economics of the startup world, particularly the tech startup world in a profound way. One of the things I read just tail end of last week was that 25% of the tech startups in the Y Combinator, which is the largest [00:06:00] incubator, uh, in the, in the technology world, have code bases that are 90% AI generated.

Uh, I think that number is gonna be going north rapidly over the next few months.

Maureen: And so for I understand for the startup world, this is an absolute game changer people working in fields that aren’t. Tech based, how? do we care? How does it impact us?

Greg: Yeah, so, uh, it’s a great question. I think, uh, every company today, uh, uh, even a company that is not a technology company. Typically is gonna have a set of applications they rely on to run their business. And you know, if you’re a medium sized business in today’s world, the preponderance of those apps are gonna be.

Cloud-based applications that everybody else is using too, because that technology [00:07:00] isn’t one different, differentiates you, right? If you’re a services company or a manufacturing, a process manufacturer, or even a discreet manufacturer, you’ve got some standard apps that you’re running for your MRP systems and your accounting systems, et cetera.

But the bulk of those businesses, even medium sized business. Have a couple of things that they do that are unique, that differentiate them in the marketplace. And oftentimes around those unique capabilities, they’ve built up some homegrown applications that help them, uh, continue to be able to innovate in a way that separates them from their competitors.

Even in that space, this is a very relevant conversation, uh, and even if it’s using very proprietary technology because now. You can purchase proprietary large language models that are uniquely oriented towards your industry, and you can run them [00:08:00] privately so you don’t have to worry about ip, and you could hook them up to these application development environments that would allow you to then iterate way more rapidly on the capability that you have in-house.

If you grow that up to the enterprise space, you’re in kind of usually a different sort of problem. So at the, at the mega enterprise space, you typically can’t be served by the generic sort of cloud-based application platforms, and so you have a lot more custom code in-house to integrate all these systems.

And to add custom features to it. What I see happening now at the enterprise level is the ability to deliver features on a timeframe that enterprises just aren’t used to. You can deliver something in days that used to take months. So the whole concept of business analysis and a [00:09:00] product function in the company begins to.

Go away in light of the fact that the long pole in the tent is really figuring out what you wanna do, not the building of it. And it used to be the other way around. We spent a lot of time making sure we really knew what we wanted to do before we started building it. ‘ cause it was really expensive to build and deploy.

What if the build and deploy cost becomes the smallest piece of the equation? And now it’s really about making sure that. You, uh, are enabling your organization to iterate very quickly and do experiments that allow you to using a very practical and a very real prototype, evolve your way rapidly towards what the next, uh, solution is.

Even if that means cutting off an entire branch of code and replacing it with a new branch of code just to get access to one new feature. [00:10:00] Because the newly generated code may be way better at all the things in that particular functional aggregate than anything you’ve built previously.

Maureen: So what it sounds like, or at least what I hear, is the coding is easy or much easier, and yet. If through the lens of an executive, I’m still trying to run my business more effectively. So I may have better code, but now I have processes to change. Humans have to behave differently and function differently to meet my users’ need or or their stakeholders. So now it sounds like the long pole in the tent goes from. actual app to the humans who are needing to be a lot more agile than they were

Greg: It could well be, although, uh, you know, in, in my experience when you. When you are trying to introduce change, [00:11:00] particularly at the enterprise level. The people who are the ultimate recipients of the change. While they may need some change management, they may need to be trained on the new application or whatever it might be.

They generally have been desperate for the change for a long time, and they’ve been waiting in a backlog of things getting justified and business case and developed and tested. And so I think what you’ll do is you’ll find that organizations are able to evolve way more rapidly. And to your point, I think it’s gonna put more pressure on people to be thinking of what are the things that we could do to be better as a company?

Because the barriers to introducing that change are getting lower and lower every day. So that agility has to go up, but it it, it’s beyond the agility that’s about just absorbing change. It’s now. You need to be a change agent. If you wanna be an [00:12:00] effective part of this company because you’re at the front line.

I don’t need you just doing, I need you telling me what doing better looks like. ’cause I can get, I can get that for you fast.

Maureen: and, and I’m just thinking of the range of humans we work with in our workplace. For some, this is absolutely a welcome change and for others that that speed of changes is not comfortable.

Greg: I a hundred percent agree with that. The, the, the days of being able to say, I’ve learned how we do this thing. And I’m just gonna be the person that does that for the next 20 years until I retire. I think that’s gone. There just isn’t going to be a tolerance for standing still, because the barriers to figuring out how to do it better are just getting lower and lower and [00:13:00] lower right.

The unity economics of improvement are getting really, really good. I mean, I look back on things like, do you remember how much energy it took to, to decode the first human genome? And, and, and now it’s a minute, right? Like it’s literally a minute and very little compute power. ’cause we’ve, you know, completely changed the unit economics.

Now imagine. That progress took several years. Imagine now with the tools that we have, including ever expanding compute power, ever expanding compute models, and you look at, you know, things like quantum computing that are just coming around the corner and the ability to do, you know, more computing in a minute than even our most powerful capability today can do in days and weeks.

Right. Like it just becomes a, a different sort of a ball game, uh, to be a knowledge [00:14:00] worker, right? And the knowledge worker is really the person who is conceiving of what’s next. And to use the vernacular sort of vibe, coding their way to that next place very quickly, and then validating that they got it right very quickly or evolving it very quickly.

Right, so now you’re in very rapid change cycles.

Maureen: And just, still trying to process the. Let’s take Columbus, Ohio, ’cause we’re here. Nationwide Insurance, Cardinal Health. How do Worthington Industries steel manufacturing? So we’ve got, uh, pharmaceutical distribution, we’ve got insurance, and we’ve got manufacturing. How do companies like that take this idea and quickly become more effective at the enterprise level?

Greg: [00:15:00] So, um, I’ll take a couple of examples. Let’s say that, uh, in the case of insurance, uh, you know, insurance is increasingly, uh, you know, focused on putting capability in the hands of the policy holder, right? And let’s say that historically you’ve had an app that you do a new release on the app twice a year.

What if you were releasing new capability on the app daily, right? What if you were,

Maureen: don’t know as an end user, if I wanna have to figure it out every day when I go on it.

Greg: okay? So you gotta be smart about how much change you introduce, but you know, I can make a small change. And deploy it to you. And this happens today, uh, it’s called AB testing, and I’m gonna release certain feature a certain way to some users and a certain feature to other users. But now the lead time to get those features built is really, really [00:16:00] low.

So you could be doing that type of testing at a much, much more rapid pace, right? Because, you know. A, a company the size of, uh, nationwide, or a company the size of Cardinal Health, they’re going to build out their own MCP servers. And so now developers don’t need to know how to talk to all those apps and the MCP server is tested so that they can trust that any, any call that goes against that MCP server is gonna be handled properly on the other side.

So I don’t have a lot of testing lead time, and I don’t have a lot of things that have to be checked and double checked. You’re just focused on trying and if you get feedback that says, you know, the 200 users you tried that feature on, hate it, it’s gone tomorrow and there’s a better version of it.

Maureen: So

Greg: I think at, at a company like Cardinal Health, when you’re looking at [00:17:00] systems that are used to optimize very, very complex supply chains, and introducing change into that world is very hard.

Uh, imagine change, being able to be introduced into that world much more easily because you’ve got these layers of abstraction that allow you to simulate what it might look like. Very, very easily when today, that’s very difficult to do.

Maureen: So I can update the backbone infrastructure. Then that runs not only a small company, but a large complex company in relatively short order.

Greg: Yes. And doesn’t mean you’ll choose to deploy that. But you can experiment and the barrier to doing the experiment is so low that it becomes worth doing, right? Like if the experiment’s gonna cost you six months of coding time and a team of 10 programmers, and it’s gonna be, and you’ve [00:18:00]gotta build a business case for that, that that’s gonna, you know, likely justify spending $10 million on this potential change and you change those economics to, we could try this.

Potential idea out in a week and put simulated data through it and see if it performs better than the existing system. Now, the unity economics of, of experimentation go way down.

Maureen: So,

Greg: sort of like, uh, you can look at it a macro where, uh, macro way, uh, in the physical world with what Elon Musk has done with rocketry, when you reduce the cost. By a hundred x and then by a thousand x, you now can be experimenting with things that we used to never even contemplate experimenting with.

He doesn’t mind if a booster blows up [00:19:00] because the booster doesn’t cost that much relative to the revenue stream of the company and the potential revenue stream where it used to be if you lost a booster. That was the end of your company. Like, you know, so it was, it wasn’t a measure twice, cut once. It was a measure a thousand times.

Right. And cut once. Well now it’s, no, that’s my iteration. So if you’ve been following like his, his new big, huge, big rocket, he’s blown up, I think six of ’em now. Right. And he is unconcerned. Like he posts on there. He is like, yeah, we got way further than I thought we would this time we learned so much. You know the next one’s gonna blow up much later in the flight, but it’s because you, unity economics are so different

Maureen: Mm-hmm.

Greg: that experimentation makes sense in a place where you never used to experiment.

Same thing is happening in the application development world. You can now experiment in places where before you would’ve never experimented.

Maureen: So does [00:20:00] this solve the question? I’m thinking of large legacy companies, again, the ones we mentioned, and or any other large legacy company that has. Let’s say it would take a billion dollars to replace their legacy system. So, um, fill in the blank. Company has built custom systems over the last few decades. Does this now start to help those companies compete with uh, newer, more nimble companies?

Greg: I think it does, and I think, but I think it’s gonna be a very hard mindset shift, which is why it’s worth talking about, because it’s gonna take a while for people to really believe. Right. But if I were at a large enterprise now, and I had a. A complex custom legacy system that I was dependent on. I would seriously think about getting [00:21:00] three or four teams started on replacing that legacy system with new generated code and see where we end up after six months, maybe two months, right?

And say, you know, this is what this system. All right. I have a dog barking in the background. Uh, so I’m gonna go try and deal with that for a second,

Maureen: Okay.

Greg: uh, so that we can pick this up again.

Maureen: Ah, okay. And I am not wearing a kil or, or a skirt pants for me. [00:22:00] I’ve got a green necklace we went to, ahead.

Greg: Here’s the culprit.

Maureen: oh. I think the cover photo needs to be Greg and the Kilt.

Greg: I’ll get right on that. Uh.

Maureen: So, Greg, you were talking about large companies with multi-billion [00:23:00] systems have an option now to at least experiment and consider replacing those systems in. A number of months rather than and billions of dollars.

Greg: Yeah, I mean I think I, uh, I think, uh, as I contemplate that, if I was at a big enterprise now I would be looking at standing up MCP servers that know how to talk to all of our various endpoints. I. And I would be looking to experiment probably with multiple teams as to how I would replace that, what used to be a multi-billion dollar system.

And quite frankly, with these new capabilities, if it’s on your books for a couple of billion dollars, it’s probably not worth a couple of billion dollars anymore. Because the barrier to building a replica of it is now extremely low, and so the asset value [00:24:00] of that thing that you spent maybe hundreds of millions of dollars building is evaporating rapidly because it can be replicated in a matter of weeks and months.

Versus, you know, years and, and, and tens of years. And so, uh, even at the enterprise level, I would be paying very close attention to this and pushing my teams to think about how do I reenter the question of these dependencies that I have on legacy systems in a new way.

Maureen: So I hear significant financial consideration in the, in the asset value on my books. And replacement value, especially as it would drive massive efficiencies in some instances where I’m entering data into three systems that don’t talk to [00:25:00] each other well, or don’t talk to each other at all. databases across multiple lines of business, as an example. So I may be able to drive massive business efficiency in six months rather than. Uh, lots of years and millions and hundreds of millions of dollars.

Greg: Yep, that’s exactly right. And uh, we’ll see this technology really get approachable even at the end user level. So I can imagine somebody who is tasked with. Entering data, similar data into multiple systems, being able to even write their own little code, even though they’re not writing code, they’re just talking to an interface and saying, Hey, I need to be able to get this same data into two systems at the same time.

Can you write me a little snippet of code that’ll run on my local workstation that will just do that for me? And you know the answer is [00:26:00] yes, you can have that. Uh, and you know, there’s entire business models that have been premised on this idea of workflow automation using ai. Uh, the problem is that, you know, they generally were.

Doing little snippets of the addressing little snippets of the problem. Like, I’m gonna read this form off of the internet and I’m gonna figure out what it’s looking for, and I’m gonna go look for the data in this system. And now you can bring that all together at a much, much more rapid pace.

Maureen: And so pace of business transformation, I wanna remain relevant. need to be engaging in this. Now, whether I am a relatively small enterprise, and especially if I’m a large enterprise.

Greg: Yeah, I mean, I think the pressure on big companies to deliver improved unity economics [00:27:00] is gonna be extreme, and I think the pressure on smaller companies to be more agile and to disrupt, I. Is also gonna be very high because the barriers to replicating what you do, uh, are, you know, they haven’t gone away.

There’s all kinds of, of barriers to being successful in a given business. And they haven’t all gone away, but one pretty important and pretty expensive. One has just gotten lowered pretty substantially. So I wouldn’t,

Maureen: I.

Greg: is not gloom and doom in the sense that, you know, like your relationships still matter.

Your culture still matters immensely. The quality of your leadership still matters immensely, right? The mindsets and the investments that you make in education still matter. It’s just that there’s this one historically, very expensive and slow thing that is now way less expensive and way slow, way less slow.

Maureen: I [00:28:00] and, and what I hear then is the mindset of the leaders and their teams has to shift as well, because that taking on the mindset of a scientist. Is now crucial, and we’ve been talking about this for a long time. You and I have, uh, I have on our show. The most effective leaders have to take the mindset of a scientist and continually experiment. We’ve now hit a point where that’s not nice, it’s required.

Greg: Yes. And, and, and the old, you know, sort of standing mentality that people in big companies have around it or around the engineering function is gonna shift. ’cause it used to be, oh, every time I ask it for anything, it takes months and it’s way over budget. Right. That’s probably not gonna be true for very long anymore.

Maureen: Well an it becoming a very different kind of business partner because they can,

Greg: [00:29:00] Absolutely.

Maureen: and cost that much money.

Greg: That’s right.

Maureen: So, so as we wrap up, what do you want our listeners to be thinking about as they are considering to deploy this new kind of tools?

Greg: Uh, you know what I would say is if you’re a technology leader. You need to be learning about these technologies rapidly so that you can be a leader in your organization at leveraging them for the benefit of the company and the people that work there. If you’re a business leader and not particularly technically savvy, I.

Learn enough about it that you can ask the hard questions and hold your technology leaders accountable for bringing this new capability to bear on behalf of your company. It’s time, it’s real, it’s now you’ve gotta be working with it, uh, or you will be behind the curve. And I, as I [00:30:00] started this has been, it’s felt like a time compression.

It’s really a tipping point phenomenon, but it feels like time compression because it’s converged very rapidly. And so if it feels like, wow, this is moving really fast, it’s because it’s converging really fast and whether you’re on the business side or whether you’re on the technology side, you gotta be getting on top of this fast.

Maureen: Beautiful. Thank you very much, Greg. And to our listeners, the call to action I think is crystal clear that for our tech leaders, the pace of change has now hit the tipping point. if you are not enabling your business to be successful. Your business may not be. It is now yours to do for the business leaders with your technology [00:31:00] organizations to ensure that you are working collaboratively. To enable these changes is also crucial and, and I wanna use this minute to also say our latest book, innovative Leadership and Followership in the age of AI starts with 10 mindsets and behaviors that I think are absolutely relevant to this conversation. So things like understanding your functional area, understanding how AI. Enables it. And also then the things like agility, abundance mindset, and the ability to communicate that we as leaders, behavior and the way we think about leading, must enable the agility for our businesses to be ful. So thank you, Greg.

Greg: Happy to be here as always.

Maureen: And let’s also end [00:32:00] with how do listeners find you

Greg: Uh, if anybody’s interested in reaching out to me, uh, I’m on LinkedIn and, uh, easily accessible through LinkedIn and, uh, I’m happy to chat with anybody that wants to talk about this topic or others, uh, in a little more depth.

Maureen: and your website is.

Greg: Uh, I think the easiest way is LinkedIn. So, uh, uh, uh, GS Moran, uh, is my LinkedIn profile and, uh, that’s the easiest way to get in touch with me.

Maureen: cool. Thank you.

​