РАСШИФРОВКА 🔗
WEBVTT
00:00:00.080 --> 00:00:04.639
An activation layer may be the thing standing between a company's pilots and its results.
00:00:04.879 --> 00:00:07.679
You're gonna end up with what I call random acts of AI.
00:00:07.919 --> 00:00:12.160
Today, some companies are like bolting a Ferrari engine onto a horse-drawn carriage.
00:00:12.320 --> 00:00:15.519
Most leaders, from my experience, think they're further along than they actually are.
00:00:15.759 --> 00:00:17.039
AI is not standing still.
00:00:17.120 --> 00:00:20.640
It's someone's job, it's everyone's job, it's no one's job.
00:00:20.879 --> 00:00:26.079
The companies succeeding with AI are the ones that are doubling down on what makes us the most human.
00:00:26.239 --> 00:00:33.359
If you have 25 people in the room and you ask 25 people to draw a cat, you are gonna get 25 very different looking cats.
00:00:34.000 --> 00:00:40.560
Give me an example of one of the uh practical applications of your model in a uh in a particular company.
00:00:40.880 --> 00:00:44.479
One of their elco preps went from eight hours to three minutes.
00:00:44.719 --> 00:00:53.840
Welcome to the AI Accelerator Podcast, where we cut through the noise and get to what actually works when people and AI go to work together.
00:00:54.079 --> 00:01:03.520
My guest today is Melissa Reeve, creator of the hyperadaptive model and author of Hyper Adaptive Rewiring the Enterprise to Become AI Native.
00:01:03.840 --> 00:01:14.079
She studied the Toyota Production System in Tokyo, served as the first VP of marketing at Scaled Agile, and co-founded the Agile Marketing Alliance.
00:01:14.319 --> 00:01:29.359
Today we're going to get into a really exciting topic: the five stages of becoming AI Native, what dynamic governance actually looks like, and why an activation layer may be the thing standing between a company's pilots and its results.
00:01:29.680 --> 00:01:31.120
Melissa, welcome to the show.
00:01:31.439 --> 00:01:33.280
It's so great to be here, Matt.
00:01:33.599 --> 00:01:34.719
I'm really excited.
00:01:34.879 --> 00:01:38.319
Can you share your background from your own perspective?
00:01:38.480 --> 00:01:39.760
How did you get here?
00:01:39.920 --> 00:01:49.599
You have a tremendous background in agile and um organizational transformation, uh, change management, so many different aspects of it.
00:01:49.760 --> 00:01:58.480
Can you sort of just give us a quick walkthrough how you got here, what you're passionate about, and how hyperadaptive folk um layers into that?
00:01:58.799 --> 00:02:09.120
What I what I love about the moment we're in today is it feels like having a nonlinear career path is becoming more uh acceptable.
00:02:09.280 --> 00:02:16.479
And so it used to be that if you didn't have a very linear path, people didn't really know what box to put you in.
00:02:16.639 --> 00:02:30.159
And I'm very grateful for that because my my time in Japan actually started studying Japanese literature and language, but I was also studying Japanese business because this was in the 1990s.
00:02:30.240 --> 00:02:33.919
And when you think about that period of time, it was all about Japan Inc.
00:02:34.159 --> 00:02:35.199
How are they doing it?
00:02:35.360 --> 00:02:37.439
How are they dominating the global market?
00:02:37.599 --> 00:02:42.960
And of course, you can't talk about Japan Inc without talking about the Toyota production system.
00:02:43.280 --> 00:02:54.319
So I was fortunate enough to be studying in Tokyo and touring the Hino Motor Company plant where I saw the and cord uh in action.
00:02:54.719 --> 00:02:58.879
And at the time, I had a way of looking at the world.
00:02:59.120 --> 00:03:02.879
I actually didn't have the language around systems thinking.
00:03:03.039 --> 00:03:08.319
I didn't have words that now we know support agile ways of working.
00:03:08.560 --> 00:03:19.360
But it was it was always hardwired in my DNA to take this experimental approach, to take a test and learn approach to continuously improve.
00:03:19.599 --> 00:03:29.280
And so when I came across, there was a white paper by a guy named Jim Ewell that I came across in 2011 talking about agile marketing.
00:03:29.360 --> 00:03:36.080
I thought, aha, this is what I want to be doing with my career in my marketing roles.
00:03:36.319 --> 00:03:44.960
And I just happened to live in Boulder, Colorado, where Scaled Agile was coming into its own and looking for a VP of marketing.
00:03:45.120 --> 00:03:56.159
And all of a sudden, all these pieces of my career, the lean, the TPS, the agile marketing, kind of culminated into that role.
00:03:56.319 --> 00:04:06.240
And what was especially cool is that the first half was as the VP of marketing, the last three years I was there, I was developing intellectual property.
00:04:06.479 --> 00:04:33.680
And that was really the springboard to some of the work and the deeper work around not only business agility, but um, I was running the podcast, so I got to interview dozens and dozens of practitioners, hear what was going right, what was going wrong, and all of that fed into the book Hyperadaptive, which I see as this extension of a lot of that thinking for the age of AI.
00:04:34.560 --> 00:04:35.519
Wow, fantastic.
00:04:35.680 --> 00:04:37.680
So let's go into that a little bit more.
00:04:37.759 --> 00:04:42.639
Uh, can you give people the plain language version of what is a hyperadaptive organization?
00:04:42.800 --> 00:04:48.480
How is it different from the agile transformation that uh many of us are very familiar with and we've already lived through?
00:04:48.800 --> 00:04:49.199
Sure.
00:04:49.439 --> 00:04:54.319
So I like to say that most organizations today are what I call linear organizations.
00:04:54.399 --> 00:05:01.040
So strategy to execution, concept to delivery, we know that AI is going to compress both of those dimensions.
00:05:01.279 --> 00:05:06.959
So an AI-native organization is one that can sense and respond in
Load full transcript
near real time.
00:05:07.199 --> 00:05:25.199
So the question becomes how do you get from a linear organization into this hyperadaptive space, iteratively and incrementally, because any of us that have lived through digital transformation or agile transformation knows that it doesn't happen all at once.
00:05:25.360 --> 00:05:26.720
It's a jagged edge.
00:05:26.879 --> 00:05:30.800
You have to be opportunistic about where you start.
00:05:31.040 --> 00:05:43.439
And um in that we we also know, and this is supported by some research from Harvard, that it's not necessarily top-down or bottom-up, that the most successful patterns are middle out.
00:05:43.680 --> 00:05:56.480
And I think that just points to the fact that we've learned a lot along the way, and I've worked to really integrate those lessons learned into this model to help organizations make these transitions.
00:05:57.279 --> 00:05:59.439
Yeah, middle out, I love that expression.
00:05:59.600 --> 00:06:00.560
I haven't heard that before.
00:06:00.639 --> 00:06:02.399
That's that's uh fantastic.
00:06:02.560 --> 00:06:11.600
Um, you you write that uh that companies today, some companies are like bolting a Ferrari engine onto a horse-drawn carriage.
00:06:12.079 --> 00:06:15.360
Can you walk us through what that actually looks like inside a real company?
00:06:15.600 --> 00:06:18.639
What does that collision feel like, like on a typical workday?
00:06:18.879 --> 00:06:23.360
Yeah, I mean, uh, you know, I'm gonna use a marketing example because I've got rich examples there.
00:06:23.600 --> 00:06:37.519
It's somebody who can now produce a blog that used to maybe take two or three days of writing, and they can they can put it in in two hours, you know, even with the writing, but then it gets stuck in review for three days.
00:06:37.759 --> 00:06:44.639
It's your advertising campaign that you've built very, very quickly and then is in legal for two weeks.
00:06:44.879 --> 00:06:51.759
And so this really speaks to the process redesign that has to happen inside organizations.
00:06:52.319 --> 00:06:57.120
And until you start really looking at that process, we've known this for a long time, right?
00:06:57.279 --> 00:07:11.839
Business process re-engineering, agile ways of working, which by the way is why I think agileists are so well suited for this moment, you know, that we're just experienced in helping people through change into new and better ways of working.
00:07:12.079 --> 00:07:31.199
Um, but it's that kind of thing that uh, you know, we've got this really fast technology, but until you think about how to use it in the context of processes, until you think about what the highest value use cases are, you're gonna end up with what I call random acts of AI.
00:07:32.720 --> 00:07:34.879
Yeah, that's that's uh fascinating.
00:07:34.959 --> 00:07:50.079
And and you're right, agileist are very, very um capable of supporting the organizations and workforces through this type of transformation
that uh the opportunity for transformation that's in front of organizations today.
00:07:50.240 --> 00:07:59.279
Let's talk about the stages of of uh hyperadaptivity that's in your that's in your book and that you uh that you outlined.
00:07:59.439 --> 00:08:06.319
You mapped five stages from laying the foundation to full hyperadapt uh adaptivity, if I can say that right.
00:08:06.560 --> 00:08:07.759
Hyperadaptivity.
00:08:07.920 --> 00:08:12.720
And uh most leaders from my experience think they're further along than they actually are.
00:08:13.759 --> 00:08:24.720
Describe the five stages and then sort of um uh walk walk us through how that how an organization can sort of or a leader can start to think about like what stage they might be at today.
00:08:25.120 --> 00:08:25.519
Sure.
00:08:25.680 --> 00:08:34.320
So I think the the biggest uh we start with the foundation, and I don't think there's anything really magic about the the stages themselves.
00:08:34.399 --> 00:08:35.279
I'll go through them.
00:08:35.440 --> 00:08:38.559
You know, the foundational stage is stage one.
00:08:38.720 --> 00:08:41.919
You start to spin up your AI councils and AI leads.
00:08:42.159 --> 00:08:48.639
I see organizations doing some of this, but they they're following again these old processes.
00:08:48.799 --> 00:08:54.080
So the AI council might be a committee that meets not frequently enough.
00:08:54.320 --> 00:09:00.799
The AI leads might just be your power users that you haven't really thought through that role.
00:09:01.039 --> 00:09:05.759
And um and then the second stage is all about task augmentation.
00:09:05.919 --> 00:09:13.600
And this is the process piece that we were we were pulling at before, which is how do you start to deconstruct those processes?
00:09:13.919 --> 00:09:20.799
And in this stage, I recommend that organizations start uh spinning up what I call the activation layer.
00:09:20.960 --> 00:09:33.360
And so this is a network of these activation hubs that start to connect with each other so that best practices and success patterns can be shared, so that we've got always on learning happening.
00:09:33.759 --> 00:09:36.799
If we have time, we can talk about the AI learning flywheel.
00:09:37.120 --> 00:09:39.120
Stage three is a gentic AI.
00:09:39.200 --> 00:09:46.159
And this is really, I know that's a term that has a lot of different meanings to a lot of different people.
00:09:46.559 --> 00:09:51.679
I use it to say it is the automation of entire workflows.
00:09:52.000 --> 00:09:56.000
So you're not you're not just spinning up an agent to do one thing.
00:09:56.159 --> 00:10:00.000
You've got agents that are covering most of a workflow.
00:10:00.240 --> 00:10:05.840
And why that's so important is because that's where the roles really start to change.
00:10:06.240 --> 00:10:20.320
And we've seen this with DevOps, we've seen this with factory automation, that when automation takes hold, we go from building the task or sorry, doing
the task to building, monitoring, and maintaining the automations that do the task.
00:10:20.559 --> 00:10:28.320
And when you think about that shift, that's a lot of role reinvention, that is a lot of upskilling and reskilling.
00:10:28.559 --> 00:10:30.559
And not everybody's a builder.
00:10:30.720 --> 00:10:38.320
We need to understand who's the builder, who's really good at monitoring and catching the agents when they go rogue.
00:10:38.559 --> 00:10:40.240
And we're already seeing hints of this.
00:10:40.399 --> 00:10:47.200
So I spin up what I call the AI impact hub that's responsible for helping the organization through this stage.
00:10:47.360 --> 00:10:56.720
And we're doing it at small, like small parts of the organization because it is so disruptive, before we start to scale that in stage four.
00:10:57.120 --> 00:11:02.639
Now, stages three and four have another level, which is moving towards value streams.
00:11:02.879 --> 00:11:10.080
So this is where we're really starting to move from a functional orientation into a value orientation.
00:11:10.480 --> 00:11:21.519
Stage five, we've got orchestrated value streams, we've got agentic workflows all over the place, we've got a really focused telemetry that's monitoring the system.
00:11:21.919 --> 00:11:29.759
Our our our funding models look different, our talent models look different, and now we've got a hyper adaptive organization.
00:11:30.480 --> 00:11:31.759
Wow, that's fascinating.
00:11:31.919 --> 00:11:45.759
I want to I want to double click into um the people side of this because you mentioned something that's a really important aspect that I haven't seen um addressed um in the marketplace that well.
00:11:46.000 --> 00:11:58.399
So as the people are going from what I would call like a human-centric uh job responsibilities, um, you know, performance measurements, et cetera, to a human plus AI more hybrid version.
00:11:58.480 --> 00:12:06.480
Like you said, uh some of their tasks and their workflows and value streams become more automated, or you know, percentage goes toward AI.
00:12:06.799 --> 00:12:14.159
Um, so the human percentage of doing that one activity goes down so that frees up their time to do other things.
00:12:14.879 --> 00:12:18.080
Role redefinition perspective, like I would view it.
00:12:18.159 --> 00:12:20.240
This is like rewriting everybody's job descriptions, right?
00:12:20.320 --> 00:12:21.679
And what roles are and things.
00:12:21.840 --> 00:12:23.759
How does your model address that?
00:12:23.919 --> 00:12:25.039
Does it address that at all?
00:12:25.200 --> 00:12:28.960
Like for the individuals, for the individuals in the workforce, like how does that factor in?
00:12:29.039 --> 00:12:30.000
Is HR involved?
00:12:30.159 --> 00:12:30.879
What is that?
00:12:31.200 --> 00:12:35.360
Yeah, and this is where we're breaking out of our boxes.
00:12:35.600 --> 00:12:38.159
And I see so many people, it's our comfort zone.
00:12:38.240 --> 00:12:39.200
What
is HR doing?
00:12:39.279 --> 00:12:40.879
What is finance doing?
00:12:41.120 --> 00:12:47.759
And what happens with AI is all of us start to start to bleed into what I call our adjacent competencies.
00:12:48.000 --> 00:12:53.440
And so there's a real mindset shift that needs to happen as we make these shifts.
00:12:53.679 --> 00:12:55.600
Is HR still HR?
00:12:55.919 --> 00:13:00.159
And so that's why I talk about these AI impact hubs.
00:13:00.320 --> 00:13:11.600
I haven't given it a label of this is HR, this is finance, but the goal of this impact hub, and it's probably a network of these hubs, is looking at the impact.
00:13:11.759 --> 00:13:18.320
If I am in finance, how is automation impacting the folks in in finance?
00:13:18.559 --> 00:13:21.120
What are those roles evolving into?
00:13:21.279 --> 00:13:23.679
And what does that transition look like?
00:13:24.080 --> 00:13:34.240
Do the frontline people are they are we able to upskill them or reskill them into a new role that's more AI plus human?
00:13:34.480 --> 00:13:40.159
Are we helping to transition some people out who are like, you know what, AI is just not my jam anymore.
00:13:40.399 --> 00:13:45.039
I really love doing the task and I'm going to do it in another way.
00:13:45.360 --> 00:13:49.039
Can we start to build uh these systems?
00:13:49.440 --> 00:13:51.120
Unilever has done this.
00:13:51.200 --> 00:13:56.639
Uh, their program used to be called Future Fit, and now I think it's called You Work.
00:13:57.360 --> 00:14:06.480
And uh what it is, is you as an individual share your purpose, you identify your skills, and then there's AI matching, right?
00:14:06.639 --> 00:14:20.720
So we're thinking about the initiatives and the skills required, we're matching them to the people who have those skills, and all of a sudden we're looking at a human and it's in his or her totality rather than just saying, is this person a finance person?
00:14:20.879 --> 00:14:22.799
Is this person a programmer?
00:14:23.840 --> 00:14:24.720
That's fascinating.
00:14:24.879 --> 00:14:37.440
So it's actually it's like a rematching, it's like a reinvention or an evolution of of uh of how the uh individual contributors in the workforce are seen and how they're how they contribute.
00:14:37.679 --> 00:14:38.720
Yeah, absolutely.
00:14:38.879 --> 00:14:44.159
Yeah, there's somebody in my network who is a salesperson, he's been a very experienced account executive.
00:14:44.320 --> 00:14:50.639
He had an undergrad in or a minor in computer, uh, computer programming, and he's a musician.
00:14:51.279 --> 00:14:56.799
And so, you know, he had to choose a box and he chose sales for a long time.
00:14:57.039 --> 00:15:07.840
And what I'm seeing is because of AI, he's able to draw upon that that computer programming background, and that's that's a new role, uh a sales programmer.
00:15:08.240 --> 00:15:16.799
Yeah, very cool, very cool because there's so much you can do now without having a tech background and just communicating with natural language.
00:15:16.879 --> 00:15:28.799
It's the first time, I think, in in the world history, uh documented history, anyway, where we have this ability for everybody to uh take advantage of technology um in a very, very significant way.
00:15:29.120 --> 00:15:29.759
Absolutely.
00:15:30.080 --> 00:15:30.720
Yeah, it's good.
00:15:30.960 --> 00:15:35.519
Let's talk about you mentioned um something that that perked my interest uh a minute ago.
00:15:35.600 --> 00:15:39.039
You talked about an AI learning flywheel in the organization.
00:15:39.279 --> 00:15:41.440
Can you expand upon that a little bit?
00:15:41.600 --> 00:15:46.000
What is your vision for what does success look like for an AI learning flywheel in an organization?
00:15:46.240 --> 00:15:49.840
Yeah, and I'm gonna I'm gonna talk on a few layer levels here.
00:15:50.080 --> 00:15:56.080
So I call it a bi-directional flywheel, and it's always on because AI is not standing still.
00:15:56.240 --> 00:16:04.240
So if you think about Claude just drops fable or maybe mythos will come one of these days.
00:16:04.559 --> 00:16:10.080
And we as an organization need to understand what that means for our people.
00:16:10.320 --> 00:16:15.360
How that's generally happening today is everybody shares that cognitive load.
00:16:15.440 --> 00:16:19.759
It's everybody's responsibility to figure out what it means for him or her.
00:16:20.159 --> 00:16:25.200
What we want to move to is something like this activation hub is out there.
00:16:25.360 --> 00:16:31.600
This network of activation hubs is out there saying, hey, I'm the activation hub for finance.
00:16:32.159 --> 00:16:35.120
You know, Claude Five just or Fable Five just dropped.
00:16:35.279 --> 00:16:38.080
Here's what it means for our part of the organization.
00:16:38.559 --> 00:16:46.960
We're gonna interpret that, we're gonna get it into the hands of the AI leads who then get it into the hands of the AI practitioners.
00:16:47.200 --> 00:16:49.360
So that's one direction of the flywheel.
00:16:49.840 --> 00:16:55.840
The other direction of the flywheel is hey, there's something really interesting happening on the front lines.
00:16:56.080 --> 00:17:10.319
The AI lead notices that happening, and then they surface that success pattern to the activation hub, which then goes horizontal to the other activation hubs and says there's something really interesting happening here.
00:17:10.640 --> 00:17:20.960
And I don't think enough thought has been given to this human complementary infrastructure that goes alongside the the AI that is new channels.
00:17:21.039 --> 00:17:28.079
And I'm not saying that these are net new positions, but they are new channels for inform where information should flow.
00:17:28.720 --> 00:17:30.240
Yeah, it's uh fascinating.
00:17:30.319 --> 00:17:38.640
If you think about the the the human uh network of communication throughout an organization, everything you just shared
uh makes a lot of sense because it's different.
00:17:38.720 --> 00:17:41.440
It's it's like learning a new skill or a new habit.
00:17:41.519 --> 00:17:48.880
You're you're you're wiring uh new um uh you know uh channels in your in your in your brain, etc.
00:17:49.119 --> 00:17:54.480
So it's the same thing to rewire an organization and their communication patterns and how how that works.
00:17:54.799 --> 00:18:13.279
Yeah, I was uh I was uh having a conversation with somebody who was a Boy Scout, and he said in Boy Scout training, when there was an emergency, you couldn't just say some you they were trained to say not to say someone call 911, because when it's someone, it's no one.
00:18:13.599 --> 00:18:19.920
But if you look at somebody and say, I need you to call 911, it happens.
00:18:20.160 --> 00:18:38.000
And so we're missing these named roles with named responsibilities, whose whose responsibility is it is to do things like keep track of AI models, to disseminate information to the frontline practitioners, to surface success patterns, to measure those success patterns.
00:18:38.240 --> 00:18:43.359
And without that, it's someone's job, it's everyone's job, it's no one's job.
00:18:44.960 --> 00:18:46.640
Very true, very true.
00:18:46.799 --> 00:18:50.480
Let's talk about governance for uh a minute, Melissa.
00:18:50.559 --> 00:18:54.240
The governance is is a word that usually kills momentum in the room.
00:18:54.319 --> 00:18:58.799
It's like someone else has to do it, AI using governance policy, et cetera.
00:18:58.960 --> 00:19:01.039
You know, you call for dynamic governance.
00:19:01.200 --> 00:19:05.200
What makes it dynamic when you uh use that term?
00:19:05.279 --> 00:19:10.400
And is it is it just another approval gate, or like what is that, how does that factor into the overall model?
00:19:10.799 --> 00:19:14.079
Well, I think it speaks to both cadence and layering.
00:19:14.240 --> 00:19:23.039
So when I think of governance uh in the the old way, it typically looks like a group of individuals who meet on a regular basis.
00:19:23.200 --> 00:19:26.720
It's usually quarterly or maybe even annually.
00:19:26.960 --> 00:19:34.079
They develop the policy, the policy goes up on the internet, uh, maybe some e-learning is developed off of that governance.
00:19:34.240 --> 00:19:46.960
People, it gets e-learning gets sent out to the people who watch it at triple speed uh while they're multitasking and they're clicking through the quizzes, trying to pass, you know, whether it's one time or five times.
00:19:47.279 --> 00:19:50.720
That's not really what we're looking for in today's world.
00:19:50.880 --> 00:19:52.640
It's there's too much at stake.
00:19:52.880 --> 00:20:01.920
So what I advocate for instead is yes, you probably still need a central group of people deciding the high high-level guardrails.
00:20:02.079 --> 00:20:04.400
They should be meeting much more frequently.
00:20:04.480 --> 00:20:07.279
You know, it could be
every two weeks, it could be every month.
00:20:07.440 --> 00:20:11.440
These new models drop about every six weeks, but there's a lot of them.
00:20:11.680 --> 00:20:15.039
So again, this needs to be much more dynamic.
00:20:15.279 --> 00:20:23.759
I then see that those guardrails need to get translated by the people in the functions, depending on how big your organization is.
00:20:24.000 --> 00:20:27.200
What do these guardrails mean for us in in finance?
00:20:27.359 --> 00:20:28.880
What kind of data do we have?
00:20:29.119 --> 00:20:32.480
What do these guardrails mean for us in marketing or legal?
00:20:32.640 --> 00:20:44.079
So they're interpreting that, getting that governance information into the hands of the activation hubs, into the hands of the AI leads that can answer those frontline questions.
00:20:44.319 --> 00:20:51.759
And then alongside that, there might actually be a governance uh GPT where people can query.
00:20:51.920 --> 00:20:53.920
I'm thinking about doing this with AI.
00:20:54.079 --> 00:20:56.640
Is this red light, yellow light, green light?
00:20:56.799 --> 00:20:58.559
Why would I do it or not?
00:20:58.799 --> 00:21:06.000
And to me, it takes governance off of the internet and into the hands of people at the moment they're needing it.
00:21:06.400 --> 00:21:07.279
Yeah, I love that.
00:21:07.359 --> 00:21:07.759
I love that.
00:21:07.920 --> 00:21:12.160
Dynamic governance is a uh is a very good uh term.
00:21:12.319 --> 00:21:16.400
You know, it goes right along with the uh hyper adaptivity, right?
00:21:16.559 --> 00:21:23.599
Because it's uh the the adaptive uh the muscle of being adaptive is really what's the new superpower that's coming out there.
00:21:23.920 --> 00:21:38.559
And um speaking of humans and machines and technology, et cetera, uh I know from your uh research you had said that the companies succeeding with AI are the ones that are doubling down on what makes us the most human, right?
00:21:38.640 --> 00:21:41.680
So this is an insight that sort of uh came out.
00:21:41.839 --> 00:21:51.920
Um, what was uh unique about that insight or unique about how that applies into your overall model uh that you maybe weren't expecting when you first uh went down this path?
00:21:52.240 --> 00:21:56.720
Well, I use this analogy of AI as like drawing a cat.
00:21:56.960 --> 00:22:05.279
If you have 25 people in the room and you ask 25 people People to draw a cat, you are gonna get 25 very different looking cats.
00:22:06.160 --> 00:22:20.720
And I think it speaks to the creativity that AI demands because in this is what also makes it so difficult for enterprises to embrace AI because it's not deterministic.
00:22:20.799 --> 00:22:25.440
It's not like PowerPoint where you can say, click here and this will happen.
00:22:25.680 --> 00:22:32.160
And so the question really is how do you unleash the creativity and teach people to draw?
00:22:32.480 --> 00:22:37.519
But also then once they understand, uh, you know,
this is Shuhari, right?
00:22:37.759 --> 00:22:48.559
Like once once they understand the rules, then maybe you can get to the place where you can break the rules and uh and unleash another whole nother level of creativity.
00:22:48.720 --> 00:22:53.200
And you've got people doing things that they've never been able to do before.
00:22:54.000 --> 00:22:55.920
Yeah, that's very, very uh true.
00:22:56.079 --> 00:22:58.880
That's a great way to um to describe it.
00:22:59.119 --> 00:23:05.839
Well, um, going into some examples, so a question for you about um uh uh measurement.
00:23:06.000 --> 00:23:09.519
So, okay, they're they're they're moving down the path of the five stages.
00:23:09.599 --> 00:23:11.200
Maybe they're starting at stage one.
00:23:11.440 --> 00:23:17.599
I imagine most companies that I've seen out there will be starting at stage one, but they start down the path of this model.
00:23:17.839 --> 00:23:28.720
How do you sort of um um you explained earlier, like sort of what success looks like going through these stages and what what the organization sort of feels like looks like um uh through that?
00:23:28.880 --> 00:23:32.799
How do you measure progress going through these stages um as you go?
00:23:32.880 --> 00:23:35.759
Can you give some examples of maybe measurements for yeah?
00:23:36.000 --> 00:23:42.400
So so one of the disconnects I see right now is everybody so hyper focused on ROI.
00:23:42.799 --> 00:23:50.960
And I get it, you know, you're paying for licenses, there's natural pressure in an organization to focus on ROI.
00:23:51.119 --> 00:23:55.039
I think the part that people are missing is that we're also innovating.
00:23:55.279 --> 00:23:59.200
And innovating and learning naturally has a J curve.
00:23:59.359 --> 00:24:02.559
So we're gonna dip before we see the acceleration.
00:24:02.799 --> 00:24:09.359
And stages one and two of the model are all about the dip, the slowdown before you can speed up.
00:24:09.680 --> 00:24:13.039
And so, what do we do when we have the dip?
00:24:13.119 --> 00:24:19.839
We we turn to things like innovation accounting, which has different metrics than typical accounting.
00:24:20.000 --> 00:24:24.079
We look at things like learning velocity, how fast are we learning?
00:24:24.240 --> 00:24:26.640
What are those experiments looking like?
00:24:26.799 --> 00:24:44.400
Sure, we can also look at things like time saved, but my guess is as people are figuring out how to use this technology, how to prioritize the best use cases, how to get organized so we don't have the random acts of AI, you're not going to see the immediate ROI.
00:24:44.480 --> 00:24:46.400
And I talk about that a lot in the book.
00:24:46.640 --> 00:25:01.359
However, once we get that momentum going, the flywheels going, the people in place alongside the technology, we we learn how to prioritize, we start to see the hard numbers come through.
00:25:01.599 --> 00:25:04.559
Because now we're starting to optimize for flow.
00:25:04.799 --> 00:25:14.480
And I love giving a little shout out to Mick Kirsten and his new book, uh, outcomes, sorry, outputs to outcomes, because he is, he he said it best himself.
00:25:14.640 --> 00:25:20.079
He says, Melissa, my book is the operating manual, the the how of your stage four.
00:25:20.240 --> 00:25:34.000
And he gets incredibly granular in how to do that and how to move towards this value approach, value streams, and get the flow going in your organization that can be very concretely measured.
00:25:34.640 --> 00:25:35.359
I love that.
00:25:35.519 --> 00:25:36.000
I love that.
00:25:36.079 --> 00:25:49.599
And speaking of others outside of uh of ROI, very, very uh perceptive, like learning velocity and uh and and guiding them through the metrics that that matter throughout that part of the journey.
00:25:49.759 --> 00:25:54.559
Um, can you give um an example um as we're winding down, just a couple more questions?
00:25:54.720 --> 00:26:07.599
Um can you give an example of um uh one one of the uh the the practical applications of of your um your model in a uh in a particular company?
00:26:07.680 --> 00:26:16.079
Like what was sort of the the before, during, after, maybe sort of short back in the napkin view of what you saw, of what you saw happen.
00:26:16.160 --> 00:26:19.839
I know it's still early days, it's not like we're 10 years into this transformation.
00:26:20.160 --> 00:26:21.440
No, it's all it's all good.
00:26:21.519 --> 00:26:32.559
And what surprised me is when I started to or when the book came out, because I'd given it to a couple of early readers, you know, that's that's part of the the deal.
00:26:32.880 --> 00:26:38.559
And one of the early uh customer stories was Greenfield Savings Bank.
00:26:38.720 --> 00:26:44.960
So this isn't a huge organization, I think it's about 600 people, community bank in in New England.
00:26:45.119 --> 00:26:46.559
Maybe you you've heard of it.
00:26:47.119 --> 00:26:54.079
And um, and it was the CIO, and she was responsible for rolling out AI.
00:26:54.319 --> 00:27:07.039
And we happened to share an Uber uh back from a conference we got to talking, and her aha moment was recognizing that it was, of course, more than a technology.
00:27:07.359 --> 00:27:10.640
And you have to think this was about 14 months ago.
00:27:10.799 --> 00:27:16.000
We were all super hyper focused on licensing and rolling out the licenses.
00:27:16.160 --> 00:27:23.200
And so that one mindset shift got her thinking about her change agents.
00:27:23.519 --> 00:27:27.039
Um, she started to get very human-centered.
00:27:27.200 --> 00:27:36.240
She picked one to two people per department to really find those the worst pain points and match one AI solution to it.
00:27:36.480 --> 00:27:46.400
She gave an example of how how one of their elco preps, which is in the banking world, went from eight eight hours to three, three minutes.
00:27:46.640 --> 00:27:48.000
So it was a huge win.
00:27:48.079
--> 00:27:54.480
And of course, we know from Cotter, when those wins start stacking up, we start to get more and more people moving in.
00:27:54.960 --> 00:28:04.319
But I think the most powerful part of that story is that she was able to pull her whole plan of AI transformation forward by six months.
00:28:05.680 --> 00:28:13.359
And so things that she thought she was going to be doing in the first half of 2027, she was doing by June of 2026.
00:28:13.519 --> 00:28:24.400
And so there's real acceleration when you put the people side of this in place and you invest in the people side uh and you get those named roles with named responsibilities.
00:28:24.720 --> 00:28:26.160
Yeah, so powerful.
00:28:26.319 --> 00:28:28.160
That's uh that's excellent.
00:28:28.480 --> 00:28:45.039
Melissa, as we uh as we wrap up, final question, you know, for a leader listening, business leader, C CEO, CIO, um uh a leader of of an organization that uh wants to do more with AI, what what should they be thinking about?
00:28:45.200 --> 00:28:49.519
What should they be doing, you know, after after listening to you?
00:28:49.680 --> 00:28:57.279
Like what's the best way they can practically start to apply uh some of the insights framework or model essentially that you had talked about?
00:28:57.839 --> 00:29:07.119
Yeah, so I think it is those named roles with named responsibilities, and it's those complementary organizational investments that are surrounding the technology.
00:29:07.440 --> 00:29:16.960
Most organizations that I talk to have the licenses, they've they've rolled them out, but it's really diving into some of what we've talked to.
00:29:17.279 --> 00:29:19.599
Do you do you have AI leads?
00:29:19.680 --> 00:29:28.720
And are those your AI power users, or are they people that you've you've handpicked because they know how to influence without authority?
00:29:28.960 --> 00:29:41.680
They know how to move resistance, they know how to prioritize hand or high-value use cases, they know how to measure success, and we're programmatically supporting them.
00:29:41.920 --> 00:29:50.240
That's what I'm talking about with those organizational investments, because it's it's getting really noisy inside these organizations.
00:29:50.559 --> 00:30:02.480
They have the pieces, oh, we should fire up a Slack channel, we should name AI leads, but without a little bit more structure around that, it's it's still all about that cognitive load.
00:30:02.559 --> 00:30:12.799
And we're expecting everybody in the organization, in addition to their day job, to be doing things like measuring success, to be doing things like staying on top of the models.
00:30:12.960 --> 00:30:16.000
And it's too much, and we're not getting the returns.
00:30:16.559 --> 00:30:17.680
Very, very well said.
00:30:17.839 --> 00:30:20.000
I see the exact same thing in organizations.
00:30:20.160 --> 00:30:24.480
And and thank you so much for your for your insights and all your experience that you shared.
00:30:24.559 --> 00:30:29.759
Uh Melissa, your I I have um uh a lot of respect for your your career, your backstory.
00:30:29.839 --> 00:30:33.759
I encourage everybody listening to this to uh uh to connect with you.
00:30:34.000 --> 00:30:36.319
Best place to connect with you on LinkedIn, you know.
00:30:36.880 --> 00:30:37.920
LinkedIn is great.
00:30:38.079 --> 00:30:42.160
The the URL for the website is hyperadaptive.solutions.
00:30:42.400 --> 00:30:43.920
Love talking about AI.
00:30:44.160 --> 00:30:46.960
Such a pleasure to be on the the show here with you today.
00:30:47.200 --> 00:30:48.079
Oh, so great.
00:30:48.240 --> 00:30:48.640
So great.
00:30:48.799 --> 00:30:50.079
Thank you so much, Melissa.
00:30:50.240 --> 00:30:52.000
Looking forward to our next one we do together.
00:30:52.319 --> 00:30:52.880
Same here.
00:30:52.960 --> 00:30:54.000
Thanks so much.