Latest / CXOTalk / CIO Advice on AI: How Chief Information Officers Can Lead in 2025 | CXOTalk #898
Transcript
- 0:00Many companies are stuck in AI pilot purgatory.
- 0:05Endless experiments, but nothing scales, nothing gets done.
- 0:10Today on CXO Talk number 898, Tim Crawford, a top CIO
- 0:17strategic advisor, reveals how CIOs and IT can breakthrough to
- 0:23real ROI. I'm your host, Michael Krigsman.
- 0:27Let's get into it A. Lot of the work I do is advisory
- 0:31with CIOs, but then I do a fair amount of advisory as an
- 0:35industry analyst with vendors as well.
- 0:38Tim, you spend so much time with CIOs when you look at AI, What
- 0:45is the impact of AI on the CIO role today?
- 0:50It's massive from a number of different lenses, number of
- 0:54different directions. When you look at the role itself
- 0:57and what the CIO needs to think about, there are impacts to how
- 1:02they work, how they operate, how they conduct themselves, how
- 1:06they inform themselves. And then also outwardly how they
- 1:11engage with the others within their organization, whether that
- 1:14be the enterprise leadership team, whether that be the board,
- 1:17whether that be customers, as well as their own organization.
- 1:21And so AI is playing a pretty significant role from a number
- 1:25of different dimensions for the CIO both today, but then as we
- 1:29go forward in terms of giving them the insights that they're
- 1:32going to need to be successful. What does this mean on a daily
- 1:39basis to the life of a CIO and to IT?
- 1:44One of the things that that is coming up again and again is how
- 1:48do we get to the insights to make really good business
- 1:51decisions. And so there was a lot of manual
- 1:55labor, a lot of manual work that went behind those, those
- 1:59decisions, whether it be through spreadsheets and documents,
- 2:03conversations, kind of just experience even coming into the
- 2:09mix. And so today you can automate
- 2:12and bring a number of those insights based on core data to
- 2:17the surface that you never had access to.
- 2:19And So what that means is you can now make decisions faster
- 2:22and more accurately in terms of where you go and what you do.
- 2:28And that impacts both the organization, the IT
- 2:31organization, whether it's from a cyber perspective and
- 2:34understanding what threats are coming your way or, or how to
- 2:38prioritize those. If it's from a service
- 2:40perspective, being able to do things like ticket deflection.
- 2:44And we see that even in the contact center too, being able
- 2:47to use AI to address some of these core issues that we had no
- 2:52other choice but to put people in front of.
- 2:55And so from a day-to-day perspective, the CIO can
- 2:58leverage AI from a number of different dimensions.
- 3:01And they really should, because this is really one of the
- 3:04attributes that's going to differentiate those CI OS that
- 3:08succeed into the future and those that really get held back.
- 3:12Are we talking about driving efficiency, saving money, or are
- 3:20we talking also about innovation opportunities for the CIO to
- 3:27forge a new or really improved relationship with the business
- 3:33partners? Now let's take a moment to learn
- 3:36about Emeritus, which is making CXO talk possible.
- 3:40If you're a business leader navigating change or driving
- 3:44growth, explore executive education programs with
- 3:49Emeritus, a global leader. They offer programs developed in
- 3:53collaboration with top universities, designed for
- 3:56decision makers like you. There's a program tailored to
- 4:00your goals, whether that's AI and digital transformation or
- 4:04strategy and leadership. Find your program at
- 4:08www.emeritus dot org. At the end of the day, it's
- 4:13allowing us to do things as IT leaders that we couldn't do
- 4:16before. We just didn't have the means
- 4:18and which to do before. But before we get ahead of
- 4:22ourselves there, we can definitely use it from an
- 4:26efficiency standpoint to help understand what are some ways
- 4:30that we can leverage AI to to do things quicker to use the
- 4:37multiples that come from using AI tools.
- 4:40And this is where the copilots and even the chat GPTS come into
- 4:45play. Some of the not chat bots, but
- 4:48when you get into agents and agentic, there are some ways
- 4:51that you can use that from an efficiency standpoint to kind of
- 4:54catapult the process and automate some of your day-to-day
- 4:57work. But once you get past the
- 5:01efficiency pieces, you really cannot stop there.
- 5:04You have to keep your eye on the ball, which is really what are
- 5:08those big transformational ways that you can bring innovation to
- 5:12the table to transform the business.
- 5:14Because at the end of the day, tech is tech.
- 5:17And whether it's red, blue-green, yellow, it doesn't
- 5:21matter. It does not matter.
- 5:24And when I talk to CI OS, they just exemplify that point.
- 5:28It's not about the tech, it's what you do with the tech to
- 5:32address those business outcomes. And that's really where AI can
- 5:36can help kind of catapult and and move you forward.
- 5:39But what about this notion of AI pilots and your term AI pilot
- 5:47purgatory? What's going what?
- 5:50Yeah. What's going on with that?
- 5:53There are a number of studies that are out there.
- 5:55There's the IBMCEO study, which they run every year.
- 5:59They've been running that for a number of years now.
- 6:01And the latest one kind of touches on success of AI efforts
- 6:06and the perception that CE OS have of how successful those AI
- 6:11efforts have been. And, and in that particular
- 6:13study, only 25% of the, of AI projects, according to those CE
- 6:18OS in the study they believe are proving out the ROI and only 16%
- 6:22are actually going to scale. That's a pretty low number.
- 6:26And then just what in the last month or two, you saw the MITAI
- 6:31study, which a lot of folks want to quote because the numbers are
- 6:34so dramatic, says that only 5% of AI projects are proving out
- 6:42their worth. Take the the MITAI study with a
- 6:46grain of salt. There are some very specific
- 6:50aspects that it focuses on that you need to make sure that you
- 6:53understand when you're looking at those numbers.
- 6:55Don't just read the headlines on it.
- 6:58But I think, you know, kind of back to your question, Michael,
- 7:00it's, it's really a question of do you understand what it is
- 7:06you're trying to accomplish with AI or are you just doing some
- 7:11sort of of random experiment? And to date, a lot of folks are
- 7:15doing just random experiments as a means to learn, which is good.
- 7:20You know, we have to learn about new technology and new
- 7:22innovation. This is new for all of us.
- 7:25But they, they're not making that transition from
- 7:29understanding how to turn it from an experiment into
- 7:32something that is meaningful for the business.
- 7:36And I'd say the business very specifically, because this
- 7:40cannot be just another technology project.
- 7:43It has to have some business outcome.
- 7:46And when you are focused on those business outcomes and
- 7:49understanding the outcome that this this project or this
- 7:52experiment could potentially have, then that helps you
- 7:55understand very clearly what are the metrics you should be using
- 7:58to measure the success or failure of that effort.
- 8:02And then kind of working backwards from there, you need
- 8:06to quickly understand, is this going to prove out your
- 8:08hypothesis or not? And if not, dump it and move on.
- 8:12And this is where I think people get caught up is a couple of
- 8:16things #1 they're not making that that determination quickly
- 8:20enough #2 they're not thinking about what the outcome is that
- 8:26they're focused on. And they're not aligning it from
- 8:29a business standpoint. They're looking at it from a
- 8:31technology standpoint and then #3 is they're not considering
- 8:36that AI and the effectiveness of using AI requires another
- 8:40component that we haven't necessarily had to do in the
- 8:43past. In the past, when we brought
- 8:44innovation to bear, we just kind of dropped it in and people
- 8:48started to adopt it and be able to to use it and leverage it.
- 8:53AI is very different. We're, we're very quickly
- 8:55learning that without the training to help us understand
- 8:59how this changes how we do work, there's a very high likelihood
- 9:03that that AI effort is going to fail.
- 9:06And we've even seen this with relatively simple things like
- 9:09Copilot, where when you start to drop it in, it changes how
- 9:13people work. And if we don't help them along
- 9:15that journey through training and other means, it's really
- 9:19going to struggle. I want to remind everybody that
- 9:23you can ask your questions. If you're watching on Twitter X,
- 9:29use the hashtag CXO Talk. If you're watching on LinkedIn,
- 9:33just pop your question into the LinkedIn chat.
- 9:38And we love your questions. You guys are so smart.
- 9:41So ask your questions. Take advantage of this
- 9:43opportunity. So David Morales Weaver makes
- 9:48the comment that he scaled a pilot by proving ROI, not
- 9:54chasing features and it saved months.
- 9:58So that's a, that's a. That's a really interesting
- 10:01point right there. Absolutely.
- 10:03But again, he's probably looking at the ROI from the perspective
- 10:07of what is this really going to do for my business or for that
- 10:11that effort. This isn't, this isn't about
- 10:14technology. This you cannot look at this as
- 10:17tech for tech or tech for tech sake.
- 10:20You have to look at this from the context of what is the
- 10:24return on investment. And what that means is what's
- 10:26the business value that this change or this technology is
- 10:30going to bring. And you have to be able to
- 10:32demonstrate that very clearly. And I don't mean demonstrating
- 10:36it with a whole bunch of of dots that you have to connect.
- 10:40You have to be able to connect those in pretty short order.
- 10:42Now let's quickly hear from Emeritus, which is making CXO
- 10:46talk possible. If you're a business leader
- 10:49navigating change or driving growth, explore executive
- 10:54education programs with Emeritus, a global leader.
- 10:58They offer programs developed in collaboration with top
- 11:02universities, designed for decision makers like you.
- 11:06There's a program tailored to your goals, whether that's AI
- 11:09and digital transformation or strategy and leadership.
- 11:13Find your program at www.emeritus.org.
- 11:20Our mutual friend Isaac Saccolic, He says this the
- 11:26elephant in the room. Is it an AI bubble or are CIOs
- 11:31moving too slowly? We are definitely in a bubble,
- 11:35but I also think that we are also moving too slowly in the
- 11:40way that we are adopting the innovation.
- 11:43So what does that mean? So we're in a bubble in the
- 11:46sense that the AI conversation is incredibly frothy, right?
- 11:51Everything's now with AI. My toaster has AI in it now.
- 11:56But is that really going to move the needle for our business?
- 12:00No. Do we really care?
- 12:02Not really, but it's a, it's the marketing buzz.
- 12:05And so you kind of have to parse through that when you get down
- 12:09to kind of the, the, you know, core building blocks of what
- 12:14this could do and you start looking at the opportunities for
- 12:18efficiency. Are we really kind of embracing
- 12:20those? No.
- 12:21And here's the reason why. And I know Isaac really well and
- 12:24I know Isaac's probably rolling his eyes going Yep, Yep, Yep.
- 12:29And that is that we're not rethinking how we have to change
- 12:33our business and how we operate. And this is something that has
- 12:36to happen within the organization as much as with the
- 12:41CIO. You know, Isaac talks a lot
- 12:43about what it means to be agile within the organization.
- 12:47That has to play out with AI. We have to think very
- 12:52differently about how we work, how we operate, our processes,
- 12:55our workflows. You can't just take AI and apply
- 12:58it to the way you've always done things.
- 13:00You have to simplify, understand how you can optimize, and then
- 13:05figure out how AI can help catapult that.
- 13:08And that's very different. That is very different.
- 13:10But love the question. I totally agree and you run the
- 13:13CIO think tank, which is so you're you're actually walking
- 13:19the walk. But but let's go to Craig
- 13:21Brown's question. He says this once the Gen.
- 13:24AI pilots have started. How do you monitor processes and
- 13:29people to justify the cost of the J Gen.
- 13:32AI tools and show the value? How do you show the value and
- 13:37what metrics are key to monitoring and how do you
- 13:40compare it against current operations?
- 13:44I mean, it's basically what he's saying is how do you justify the
- 13:50cost, show the value and how do you compare it and measure it?
- 13:53Depending on the specifics of the type of work you're doing,
- 13:57the answer is going to change a little bit, but you have to you
- 14:01have to go back to let me use some examples.
- 14:05If let's say you're using, I mentioned Copilot earlier, how
- 14:09much efficiency is that really kind of driving into the
- 14:12organization? And what that means is you have
- 14:14to spend the time to understand how much time is actually being
- 14:18saved. You can't assume that you and
- 14:21you, you are going to make some assumptions along this process,
- 14:25but you can't start with assumptions.
- 14:27You have to start to understand actuals like how much time of
- 14:31the 10 people in in a given department, how much time are
- 14:34they really are spending? And you have to spend the time
- 14:37with each of them. And then you can extrapolate
- 14:39from there. But it requires more footwork to
- 14:43truly understand that. And then you can extrapolate to
- 14:47your organization of 20,100 thousand, 250,000 people when
- 14:52you're starting to build out discrete non efficiency types of
- 14:56projects, but rather innovative projects, those have very
- 15:00discreet outcomes because now you're talking about doing
- 15:03things you couldn't do before. And in those cases, you
- 15:06typically have some dollar value that's tied to the outcome.
- 15:11Maybe it's a a new product you can come out with, maybe it's a
- 15:15a new service line or a new market you're going after.
- 15:19And so they're very clear metrics that you can track as
- 15:22part of that. In addition to that, on the back
- 15:25end, companies like Aptio from IBM and the TBM Council, they're
- 15:31trying to get their arms around this as much as even just the
- 15:34fin OPS foundation from a cloud perspective and AI perspective
- 15:39of how do you start to understand what you're actually
- 15:42spending on these efforts to be able to factor into that value
- 15:48equation? Because that's really what
- 15:50you're asking here, is what's the value that drives the ROI?
- 15:55And a lot of folks focus on the cost, but cost for AI is not,
- 16:00it's not clear. It can be convoluted.
- 16:04And so there are efforts that are starting out, but we're in
- 16:07the very early days of this too. And so this is where I'd
- 16:10encourage you to figure out with your partners whether that's the
- 16:15SIS, the GSIS, but also the vendors, the enterprise vendors
- 16:19that you're working with. Work with them and help them
- 16:23understand what it is that you need help with in terms of
- 16:26clarity around this, because they're trying to figure it out
- 16:28too. Let's go over to Twitter to X
- 16:33and an interesting question from Arsalan Khan.
- 16:37He says do these AI fail? It's interesting.
- 16:40He says, do these AI failures? Well, we haven't really been
- 16:42talking about AI failures except I guess pilots not succeeding, I
- 16:47guess we could say is an AI failure.
- 16:49And yes, that MIT study is about POC's failing.
- 16:55Arsalan says this do these AI failures occur because of lack
- 16:59of vision of non-technical executives, or because tech
- 17:04executives just don't know how to convince their peers that AI
- 17:11is a holistic solution? Both are true.
- 17:15There is an understanding around what is the value of the
- 17:19project, what is it, the value of the experiment and you have
- 17:23to tie that to a business outcome.
- 17:26And one of the problems that that I often see is that, and
- 17:31this kind of goes back to something, Michael, that we were
- 17:34talking about kind of at the top of the show.
- 17:36But when you start talking about the role of the CIO and how the
- 17:40role is evolving, And I'm sure many folks haven't had a chance
- 17:43to read it, but I just published A blog post talking a little bit
- 17:47about this last night from my blog on avowa.com.
- 17:51But as the CIO role transforms from traditional to
- 17:56transformational, and what that means is from a technology focus
- 18:00to a business focus, what that does is it gets the IT
- 18:04leadership to be thinking more about those business outcomes,
- 18:07not technology outcomes. Many of those failed projects,
- 18:12they're technology projects, they're not business projects.
- 18:15And this is one of the problems is when you're looking at
- 18:18efficiency, it's efficiency of what?
- 18:21And So what is this project really intended to to accomplish
- 18:28and how do you get to that business outcome?
- 18:32But first you have to understand what those business outcomes
- 18:35are. And many folks in IT don't have
- 18:39the level of depth that they need to have around what those
- 18:43business outcomes should be, what their business objectives
- 18:46are. Like, for example, do you know
- 18:49if you're in IT, are you crystal clear on understanding what your
- 18:53company's objectives are, what the business objectives are that
- 18:56the executive leadership team is oriented around?
- 19:00And if the answer is no, well, OK, there's step number one is
- 19:03understand what the nature of your business is.
- 19:05Step number 2 is understand how your business operates.
- 19:08How do you make money? How do you spend money step #3
- 19:11is understand your customer. What does your customer do?
- 19:14How do they engage with your company?
- 19:17Why do they engage with your company versus your competition?
- 19:20Now some in technology might go, well, wait a second.
- 19:23I work in ITI don't work in marketing or sales.
- 19:25But guess what? With AI, you have to start
- 19:29understanding this. We've needed to understand this
- 19:32earlier, but with AI it's it's absolutely necessary.
- 19:37Why does AI force someone in IT or force a CIO to think more
- 19:44broadly across the company? Because I was under the
- 19:49impression that CI OS should have been doing this since so
- 19:54basically forever modern times I.
- 19:57Don't want to go down a rat hole on this.
- 19:59I have a whole talk track on the anthropology of IT and how we
- 20:03started by focusing on business outcomes in the 60s.
- 20:07And we kind of lost track of that.
- 20:09And we started to say, look, you know, we got this problem, we
- 20:12can figure it out solving these business problems.
- 20:15And we started to create this chasm between what IT did and
- 20:18what the business needed and the business to kind of lost touch
- 20:21of what technology was doing behind the scenes.
- 20:25We have now kind of come full circle where, guess what, that
- 20:28chasm needs to be closed. And the reason for that is
- 20:31because what is AI really doing? It's changing how we operate our
- 20:36business, whether it's from an efficiency standpoint, whether
- 20:40it's from how we engage with our customers.
- 20:42It is changing the way that we operate our business.
- 20:46And the only way you're going to know or understand the impact of
- 20:51that to your business is by understanding your business.
- 20:55And unfortunately, as you said, many folks in IT don't have more
- 21:00than just a basic understanding of how their company operates.
- 21:04You know how it spends money, how it makes money, how it
- 21:06engages with customers. And Simone Joe Moore on LinkedIn
- 21:11really wants to emphasize this point that we need to not focus
- 21:16on, quote UN quote, just efficiency and really think
- 21:20about the effectiveness for work for organizations and people.
- 21:26Completely agree. The real money, the real
- 21:28opportunity is going to be when you start to move into what I
- 21:32call innovation projects. These are things that you could
- 21:34not do before AI. That's different than an
- 21:37efficiency project. Efficiency projects will have
- 21:40value and they're a great way to learn, but very quickly you have
- 21:44to figure out how to shift gears and start adopting some of those
- 21:47innovative efforts, some of those things that you couldn't
- 21:50do before AI. And I completely agree.
- 21:54That's that's where the real gold is for companies.
- 21:57It's the efficiency is important, but the real gold is
- 22:00when you can do things that that will allow you to open doors you
- 22:05couldn't open before, provide insights that you never had
- 22:09access to in the past, allow you to engage with customers in ways
- 22:13that you couldn't before. And, and here's the kicker, it
- 22:17starts to differentiate you from your competition.
- 22:20And so this is where the big race is, is how do you start to
- 22:23differentiate your company from your competition and being able
- 22:26to address customers where they are and where they're going.
- 22:30So looking ahead, not behind, but also looking around the
- 22:33corner. Rocky Vienna asks about the
- 22:37security risks in what are using in essence, he says, public
- 22:43LLMS. Where do you see movement in the
- 22:45development of platforms to provide a security envelope
- 22:49around corporate use of AI? It's a very important question.
- 22:54We absolutely need to be thinking and talking more about
- 22:58not just security, but governance.
- 23:00And this is something that all of you should be hounding on
- 23:05your vendors and providers to be talking about is how they think
- 23:08about governance of data. So there are different aspects.
- 23:12Rocky's asking about security around the actual models
- 23:16themselves. And there's a, there's an
- 23:20interesting MIT study different than the MITAI study that talks
- 23:24about meek models. And so probably just look for
- 23:28MIT meek models. And it's, it's an interesting
- 23:32read talking about how the changes in big models versus
- 23:36smaller models are going to play out.
- 23:40When you start to think about security, you have to think
- 23:42about the data. You have to think about the
- 23:44outcome of how you're using that data and protecting that data.
- 23:47I think this is where there are going to be problems with big
- 23:52models that we haven't solved. For a good example, here's a
- 23:55great example. I make a public statement that
- 23:58says, you know what? Michael Krigsman is awful.
- 24:01He's a terrible person. Mind you, none of this is true.
- 24:04Oh, it is true. I didn't know it's true, but go.
- 24:06On no, no, no, no. I've known Michael for years.
- 24:08He's a great human being and love working with him.
- 24:13But the problem is when a model picks that up, there is no way
- 24:17to remove that from the hive. There's no way to back that out.
- 24:21So Even so, once we go down this path, the problem is you're
- 24:25going to start ending up with more misinformation and
- 24:27disinformation. And so that's going to be a
- 24:29problem. But you have to also think about
- 24:32governance around the data. You do have access to an
- 24:34especially proprietary customer data.
- 24:37And even some of those challenges, when you start to
- 24:39think about user governance models as well as AI agent
- 24:45models, as well as knowledge models, and then the different
- 24:48governance models that come from the different systems of record,
- 24:51how do you start to to bring all that together to make sense of
- 24:55it? And in my conversations with
- 24:57some of the largest enterprise vendors on the planet, they're
- 25:00still working that out too. So hang on, it's going to be a
- 25:04rough ride, but we've got some some pretty interesting
- 25:07challenges ahead of us. And of course, we're dealing
- 25:10with LLMS that have an insatiable desire and greed to
- 25:16suck up everything in their path.
- 25:21Yeah, and you have to be careful that the configuration of your
- 25:24interfaces to those LLMS exclude the sucking up of your corporate
- 25:30data. There's that too.
- 25:32And so I think this is where you have to think about your data
- 25:35strategy. You have to think about how
- 25:37you're protecting what is important to your organization.
- 25:40And that's probably a totally different conversation, but AI
- 25:44starts to factor into that when you start to think about the
- 25:47governance of what AI and especially as we move into a
- 25:50gentic with learning models, reasoning models, discovery A to
- 25:54AMPI could go on and on just on that alone around the challenges
- 26:00to govern data through those processes.
- 26:03Greg Walters has been patiently waiting on LinkedIn.
- 26:07And then we're going to jump back to Twitter because we got
- 26:09questions stacking up over there too.
- 26:12He says the AI failure seemed to occur when AI is dropped into
- 26:18existing processes instead of the processes adapting, being a
- 26:25jet digested by AI. But he says he he says a ray of
- 26:30hope that it seems that finally we're looking to focus on
- 26:33deliverables versus adherence to the process.
- 26:36But I, I will just mention I've heard the same conversation
- 26:40going back, you know, 25 years. This is not new.
- 26:43There's a whole nother talk track around just policy
- 26:47development. You know, how did we create some
- 26:49of these business policies and business processes in the 1st
- 26:53place? And usually it's because
- 26:56something didn't go right or we wanted it to work a very
- 26:59specific way based on the parameters at that point in
- 27:04time. Well, guess what?
- 27:05Those parameters have been changing in some cases over
- 27:09decades, but the company never changed how they did things.
- 27:13And so one of the things that that you're kind of alluding to
- 27:17and really needs to be thought out for success is how do you
- 27:21start to understand that process first?
- 27:24So set AI aside for a minute. How do you understand that
- 27:28process? How do you optimize that
- 27:29process? And then how do you bring AI
- 27:32into the mix? Because if you just simply layer
- 27:34AI on top of existing processes, you're going to end up with a
- 27:39mess pretty darn quickly, and especially as you move into an
- 27:42agentic framework. Arsalan Khan comes back.
- 27:46He says as a recovering enterprise architect, we should
- 27:51aim for AI. We should aim for AI taking our
- 27:55jobs. Some enterprise architects are
- 27:58OK with that, but some are not. How much of the AI race is just
- 28:02self preservation? The problem is this is where the
- 28:05marketing is really kind of settled in.
- 28:08AI is not going to, AI is not directly going to take your job,
- 28:13but AI could take your job. And what I mean by that is
- 28:17there's going to be a difference between those folks that use AI
- 28:22and those folks that that don't leverage AI.
- 28:25And the folks that leverage AI are the ones that are going to
- 28:28succeed in Excel. And the ones that don't, and
- 28:31this goes all the way to the CIO are the ones that are going to
- 28:35perish. It's the same reason why
- 28:37traditional CI OS are in decline and transformational CI OS are
- 28:42are in demand. It's because of what they focus
- 28:46on and how they leverage the tools at their disposal.
- 28:49And also how they can provide greater value over time to their
- 28:55particular role in their particular organization.
- 28:58Chris Peterson on Twitter makes a couple of interesting points.
- 29:01He says #1 the recent, There was a recent study from Anthropic
- 29:07that said only 250 bad documents are sufficient to poison almost
- 29:13any size of model at training time.
- 29:18So getting to your point about the bad data, basically poison
- 29:22data, and Chris asks this, he says that he agrees about the AI
- 29:28bubble and the value conversations that need to
- 29:30happen. How do we factor in the
- 29:33potential cost explosions when venture capital and circular
- 29:38deals go away? As customers, can we even figure
- 29:43out our sustainability metrics? I think we'll see more smaller
- 29:48models and leverage the large LLMS less.
- 29:54It's kind of crazy when you look at the whole open AI Oracle meta
- 29:59and they're just driving toward bigger, bigger, bigger.
- 30:01And this is one of the reasons why I kind of to the earlier
- 30:04question about a bubble. I think there, there is
- 30:08definitely a lot of bubble taste to to what we're seeing happen.
- 30:13At the end of the day for the enterprise, they need something
- 30:16that is more performative and also something that's more
- 30:20specific to their industry or or their particular company.
- 30:24And that's where you get to specific data and data, frankly,
- 30:28you can control. So that's one aspect.
- 30:31The second one, which you bring up around sustainability, this
- 30:35is a real problem. This is a real problem.
- 30:38And I've tweeted about this. I'm sure, Chris, you've seen me
- 30:41talk about this, but we have stopped talking about
- 30:46sustainability. And I think some of this is
- 30:49somewhat politically driven for the time being.
- 30:52But we're going to come back and look at this in spades, which is
- 30:56how do we start to understand the requirements from of natural
- 31:00resources and the impact that AI is having, whether that be from
- 31:03an energy perspective, right. Amazon got asked not to not to
- 31:08bring more workloads into Ireland or something along those
- 31:11lines. We're going to see more of that,
- 31:14whether it be for water reasons or power and driving costs.
- 31:19Communities that have data centers, they're seeing power
- 31:22costs go up. But that conversation is largely
- 31:26kind of subsided for the time being.
- 31:28We're going to have to come back to that.
- 31:29And I think that's also going to drive our need to get smarter
- 31:33about how we use this technology too.
- 31:36A question from LinkedIn from Edward Monroe who says how are
- 31:43CIOs approaching the shift from SEO search engine optimization
- 31:50to Geo generative engine optimization?
- 31:55And are there early priorities that can guide the development
- 32:00of a road map from this shift from SEO to Geo?
- 32:06Number one, the CIO doesn't care and I don't say that flip, but
- 32:12rather you have to look at what the CIO focuses on and you're
- 32:16talking about something that is farther into the organization
- 32:19and probably in the CM OS org specifically around how these
- 32:23technologies are used. Keep in mind the CIO moving
- 32:27forward is very much focused on business outcomes, not specific
- 32:31technologies. You know, I often say it doesn't
- 32:34matter whether it's blue-green, yellow, red, blue, it doesn't
- 32:38matter who, whose technology, whose badge, whose, which
- 32:41technology you're using. What matters is how is this
- 32:45going to move my business forward?
- 32:48But that particular question I'd have to punt a little bit on
- 32:51because that's that's outside of the realm of the CIO.
- 32:55From the CIO perspective, this is an internal set of operations
- 33:02of tasks and IT IT can help support that the infrastructure.
- 33:09But once you get inside marketing, really is that the
- 33:13job of the CIO or is that the job of somebody in marketing who
- 33:16has real deep expertise in that specific process?
- 33:20The other piece that you're starting to see more of is
- 33:22partnerships between organizations.
- 33:25So CIO and CMO coming to the table together, CIO and
- 33:28CROCIO&CFOCIO&COOCIO&CPO. This is something that came up
- 33:36this morning and our conversation within the think
- 33:40tank was around the the role of the CIO and the Chief Product
- 33:44officer. And so you have to keep in mind
- 33:49that the level in which these folks are focused and the things
- 33:54that are most important to them are very much into these very
- 34:00strategic business outcomes that move the needle for the
- 34:04organization and for the customer.
- 34:06When you get into the the how of it kind of into the the details,
- 34:12that's when you start working into the organization a little
- 34:14further. And also I would assume that's
- 34:17that's when you start working with detailed experts who may
- 34:22live inside it as well, but who have very narrow domain
- 34:27knowledge that's deeper on specific topics.
- 34:30That's right. That's absolutely right.
- 34:32So we have another question from Twitter from XA, really good one
- 34:37going back to the POC's to the proofs of concept.
- 34:44Elizabeth Shaw says how do you and when do you terminate an AI
- 34:51proof of concept, especially because doing so is rife with
- 34:57politics. So unpack that one for us.
- 35:00The issue there is politics sometimes can govern decisions
- 35:04and it's it's kind of the necessary evil that you have to
- 35:09work through. If you've ever worked in large
- 35:11enterprises, I have most of my career, you have to have the
- 35:15right balance. But this is also where
- 35:17relationships play a role. You have to think about what's
- 35:23the impact of spending the money and spending the resources,
- 35:27because it's not just about money, it's about resources that
- 35:30are going toward that effort when they could be applied
- 35:33somewhere else and ensuring that that effort is aligning with
- 35:37those business outcomes. And I'm going to sound like a
- 35:39broken record, but business outcomes and business objectives
- 35:44need to rule the day. Now, the politics will play a
- 35:47role in that. And you have to make sure that
- 35:50you have the right balance, healthy balance, and play the
- 35:53job of the politician in it. But you have to understand that
- 35:59they have motivations and incentives as much as you do.
- 36:03And so if you can understand what their incentives and
- 36:06motivations are of why they're driving that and pushing that to
- 36:09continue, that might give you some insights as to how you
- 36:13could drive the focus in a different direction.
- 36:16So you can pull the plug and move on.
- 36:19But on the sum, you have to figure out how to quickly
- 36:22ascertain is this experiment going to succeed or fail?
- 36:27And if it's going to fail, you need to pull the plug as quickly
- 36:29as possible. If it's going to succeed, you
- 36:32need to be able to demonstrate success in short order.
- 36:35I had a conversation with someone that said Oh well, we're
- 36:38going to be looking at a 12 month run rate before or 12
- 36:42months out before we can actually show progress.
- 36:44And I said those days are behind us.
- 36:47Like you have to show progress in very short order, in some
- 36:50cases weeks and months, not years.
- 36:53On this topic of succeeding with these AI projects or failing,
- 37:00Isaac Saccolic comes back with a real definition.
- 37:03He says failure occurs when AI exposes data, or lack of testing
- 37:10to validate results, or lack of change leadership.
- 37:14He's absolutely right. You have to think about that.
- 37:17You have to think about how failure is going to play out.
- 37:20And you know, to the blog post that I alluded to earlier that I
- 37:23wrote last night, there's a, there's a significant change
- 37:29coming. There's a study that just came
- 37:31out, executive leadership, 6 out of 10 executives across the
- 37:37spectrum, so globally, across the spectrum are expected to
- 37:40change roles within the next three years.
- 37:44That's going to drive a phenomenal amount of change when
- 37:47you start to kind of boil it down into technology and what
- 37:50folks are going to be looking to do.
- 37:53So change is in our future. If if you don't already see it
- 37:57today, expect more of it. If you are seeing it today,
- 38:01expect even more of it and it's going to come at you faster.
- 38:05And so you have to figure out how can you accelerate your
- 38:07organization and your thinking in your decision making to be
- 38:11able to accommodate that. I just recently I was giving a
- 38:15presentation to a group of executives and one of the
- 38:19statements I made kind of caught a few off guard and I said,
- 38:22look, today is the slowest your company will ever operate.
- 38:29Think about that for a minute. Today is the slowest your
- 38:31company will ever operate. We're only going to get faster.
- 38:37Strap in and hold on. On the subject of the human
- 38:43relationships that can help accelerate and make these
- 38:48provide the the fabric, if you will, that makes these projects
- 38:53relating to the business in in the strongest possible way.
- 38:58Arsalan Khan comes back and says on Twitter, should tech
- 39:02executives invite non tech executives to our echo chambers?
- 39:08What incentives are there beyond just cost and efficiency for non
- 39:14executives? So how do how do we broaden this
- 39:19pie, so to speak? I often talk about in my past as
- 39:23an IT leader, some of the first relationships that I built when
- 39:27I walked into a new organization were with my chief legal counsel
- 39:32and head of audit. Now, you might go, well, those
- 39:36are kind of odd for a CIO to build relationships with.
- 39:40But here's the thing. You're going to run across those
- 39:43folks from a legal perspective and from an audit perspective
- 39:46before you know it. And if you are at least building
- 39:49the relationship and setting the groundwork so you can
- 39:52understand, listen, and respect where they're coming from, then
- 39:58guess what? When it's time for you and the
- 40:00other person to sit down and have a conversation because
- 40:03you've been breached or you have a legal issue you have to
- 40:06navigate through, you already understand where they're coming
- 40:10from. You're not trying to figure it
- 40:11out in the heat of battle. And so it's a way to build
- 40:15respect. It's a way to to build those
- 40:18relationships. And I can't emphasize how
- 40:21important those relationships are.
- 40:23When things get tough, it's the relationships you will have to
- 40:27lean on. It's not going to be some other
- 40:30factor or technology whiz bang. You have to rely on the
- 40:34relationship. So don't just look at at the
- 40:38value of what you get out of that particular transaction of a
- 40:42relationship or relationship is something you build over time,
- 40:46you invest in over time. And when you think about it that
- 40:49way, it changes your paradigm in terms of why you invite these
- 40:53folks to your staff meetings or to present to your organization.
- 40:58On LinkedIn, Simone Joe Moore makes the point that applying AI
- 41:06on top of current processes is no different from automation
- 41:10questions. If your base isn't good, we'll
- 41:12just do the wrong thing. Faster and speed does kill until
- 41:16we slow down enough to make the speed sustainable.
- 41:20And Greg Walters comments that this is all classic needs
- 41:26assessment and system analysis stuff.
- 41:30When you do apply automation, whether it be RPARPO, whether it
- 41:36be a Gentek frameworks, on the other end of the spectrum, when
- 41:41you do, all you're doing is just automating an existing process.
- 41:46There's a danger with that, and this is something that people
- 41:49don't necessarily think about or understand, which is number one.
- 41:53Have you taken the time to understand how that process
- 41:55works? No #2 you're taking the human
- 41:59out of the loop, and we underestimate the value of the
- 42:02human. So when when you tell a human do
- 42:05a certain task, if there are situations that come up that are
- 42:10outliers to that task, the human has the cognitive capability to
- 42:14be able to say that doesn't pass the sniff test.
- 42:17I better ask someone because it's not my process, it's not my
- 42:21capability. When you automate those
- 42:24processes, all of those checks go away.
- 42:27And so that becomes incredibly dangerous where you start
- 42:30automating bad processes. And so that's something you have
- 42:35to think about and why it's always important understand what
- 42:38you're trying to work with first before you start layering this
- 42:42in. And frankly, is this the right
- 42:43place to start? I mean, that's a whole nother
- 42:45aspect, which is prioritization. And then Michael, the second
- 42:49question. He's saying this is classic,
- 42:53classic stuff. You need system analysis.
- 42:55Go ahead. Yeah.
- 42:56So, so I will sort of agree with that.
- 43:00I think you need more business analysis then you need system
- 43:04analysis. And the reason I say that is it
- 43:06all starts with how you operate, how your business functions, how
- 43:10the how the work gets done. Understand that first and then
- 43:14you can go into some system systems analysis to be able to
- 43:19ascertain how you should adapt that or change it accordingly
- 43:25when you start thinking about infusing AI into the mix.
- 43:29But you should, you should definitely start with business.
- 43:31And the reason why I make that what may seem overly subtle of a
- 43:36distinction is because if I were just to agree and say, Yep,
- 43:40system process, which you're kind of right on too many people
- 43:45would go, oh, OK, so I just apply my normal system process.
- 43:48Take this the way the system was built to understand how the
- 43:51system was built and refactor it for AI and that would be a
- 43:55mistake. Elizabeth Shaw says who should
- 43:58own the AIROY as an issue? And Arslan Khan says why is the
- 44:06onus always on technology folks to know about the business
- 44:11processes? And I think these are related
- 44:14issues. Who's responsible for AIROY and
- 44:18why is it always on IT? The easy answer for Elizabeth's
- 44:24question is there's no one person that's responsible for
- 44:27it. And This is why some
- 44:28organizations are putting together an AI council that can
- 44:32be driven maybe by the CIO, sometimes by the CEO of the
- 44:37company. But there's someone that leads
- 44:39the council that's a cross functional representation of the
- 44:42organization and they collectively, not individually,
- 44:46collectively are responsible for prioritization and ensuring that
- 44:52guardrails are put in place and prioritization and, and costs
- 44:56are, are being managed. And then I think the, the second
- 45:00piece of that is in terms of just kind of understanding how
- 45:04do you evolve that over time, right?
- 45:07And so it's an evolutionary process.
- 45:10We're still learning, but you have to be able to accelerate
- 45:14the decision making process in terms of what you work on first,
- 45:18second, third, and then evolve that right, Fail fast.
- 45:21We've heard these terms before, we're just applying them into
- 45:25AI. Let's just briefly talk about
- 45:28vendors, technology vendors, and services.
- 45:33Vendors. Historically, there was a real
- 45:36problem with services because companies would outsource that
- 45:41work and the expertise would lie outside the company with the
- 45:46third party vendor that the professional services vendor.
- 45:50And it created a cycle where IT never developed the expertise.
- 45:58And so they were, it was this cycle of dependence.
- 46:01What's your view when it comes to AI now, today?
- 46:05Number one, you cannot do it yourself.
- 46:09You can't do this yourself. You will need help.
- 46:12Wait, wait, wait. Sorry.
- 46:13So, so just so I'm clearly understanding you.
- 46:16So you're now advocating outsourcing knowledge elsewhere
- 46:21so that you don't have to worry? Is that what you're saying?
- 46:24No, no, that is absolutely not what I'm.
- 46:25Saying OK sorry I missed. What I'm saying is when when you
- 46:28go down the path of these these larger projects, so moving
- 46:32beyond efficiency and get into the more innovative projects,
- 46:35you won't be able to do them all internally.
- 46:37You will need help with that, but you have to understand where
- 46:41those partners play the most powerful role for your
- 46:45organization. To your point, if you are
- 46:48outsourcing your organization to a third party and you're
- 46:52basically paying for them to build the institutional
- 46:55knowledge and maintain the institutional knowledge and the
- 46:58learning, that's a mistake. That's a critical strategic
- 47:01mistake. But the way that you can
- 47:05leverage a third party is by using them to backfill some of
- 47:09the more mundane things and help your organization, help your
- 47:12team learn and augment some of that with that outside thinking
- 47:16so that you can help them accelerate their own thinking.
- 47:19But at the end of the day, you want that institutional
- 47:22knowledge inside your organization, not outside.
- 47:26So I'm not advocating that it be outside.
- 47:29What I'm advocating is you need to leverage outside resources to
- 47:34bring that expertise and help train and educate your team so
- 47:37that they can move more quickly through this learning process,
- 47:41right? They're not reinventing the
- 47:42wheel, but maybe they get a framework and then they can
- 47:45reinvent it for their own needs. But then you're also not
- 47:49building that that center of excellence outside of your
- 47:52organization. You're building that squarely
- 47:55within your company and that's not always the case,
- 47:59unfortunately. I ran into one organization
- 48:02where half the organization was outsourced.
- 48:06And the problem with that is a lot of that innovation was being
- 48:10done by those outsourced staff folks.
- 48:13And yeah, it's, it's an expensive venture and really
- 48:17detrimental to your organization, to the company and
- 48:20your customers. So you're not at all suggesting
- 48:23handing off responsibility for your institutional knowledge to
- 48:30a third party and then letting them run with it, but you're
- 48:35saying prioritize carefully, use them to help jump start.
- 48:39One of the commenters on LinkedIn and I, I forget which
- 48:44one, spoke about this prioritizing.
- 48:47So you're saying pick your battles very carefully?
- 48:50Yeah. And if you are abdicating
- 48:52responsibility to a third party, you need to go.
- 48:56You seriously need to go because that that puts the organization
- 48:59at risk. And I've seen it happen.
- 49:02And unfortunately, I've seen situations where the the senior
- 49:07most person in IT is essentially my words, they're holding the
- 49:12company hostage because of some of those strategic mishaps.
- 49:17And it never ends well. But it's going to force boards
- 49:21and CE OS to make some very tough decisions.
- 49:26And yeah, it's going to get rectified one way or another.
- 49:30But yeah, just don't go down that path.
- 49:33In one sentence, Chris Peterson on Twitter asks, do you see a
- 49:38specific C-Suite role for AI alongside the CEO like some
- 49:43organizations have done with the CDO, or would that stay within
- 49:48CIO? I'm thinking of like chief AI
- 49:50officer for example, so very quickly.
- 49:52I have long since said and I will continue to say, if you are
- 49:57looking at a Chief AI Officer or Chief Digital Officer, CDO or
- 50:01CCAIO, you probably don't have the right CIO.
- 50:05It's not to say that you don't need AVP of AI or VP of of data,
- 50:10but at a chief officer level, no, absolutely not.
- 50:15Why? Why do you say that?
- 50:17At the end of the day, is it really a role that is separate
- 50:21from the CIO? No.
- 50:25And then is it a role that is going to operate amongst the
- 50:29executive leadership team on par with the CEO, the
- 50:32CFOCMOCOOCROCHROCIO? And the answer is no.
- 50:38It's a very specific task. That's why I say VP of AI or VP
- 50:42of Data is a much more appropriate role.
- 50:45And then that would roll into the CIO, the strategic CIO cause
- 50:51again, the CIO, their closest relationship is going to be with
- 50:58the CEO and then secondarily with the rest of the ELT.
- 51:03You're not going to see those other roles kind of taking,
- 51:05taking a similar path. And like I said, it's a longer
- 51:08conversation as to more of the reasons why and how that plays
- 51:11out, but that's kind of the top line view on it.
- 51:15Tim Crawford, thank you so much for spending time and sharing
- 51:19your expertise with us today. I'm very grateful to you.
- 51:22Michael, thank you so much for the invitation to join and love
- 51:25the questions coming from the audience.
- 51:26This has been great. Everybody, now is the time for
- 51:30you to subscribe to the CXO Talk newsletter so that we can notify
- 51:35you about upcoming shows. You can be part of our community
- 51:39and you can participate. This show is all about the folks
- 51:43who are listening, so go to cxotalk.com right now, subscribe
- 51:47to the newsletter, and we'll see you again next time.
- 51:50We have really, as we always do, like incredible shows coming up.
- 51:54Two weeks from next week, there's no show.
- 51:56Two weeks from today. We have two members of the House
- 51:59of Lords in the UK talking about big tech and policy around AI
- 52:07and big tech. So that's going to be really
- 52:09interesting right from the horse's mouth.
- 52:11So join us. Thank you so much.
- 52:13Thanks to Tim Crawford. We'll see you again next time.