Keith Shaw Hi everybody.

Welcome to Global Tech Tales, the show where we talk with editors from around the world about the latest technology and leadership topics to find out what buyers want. I'm Keith Shaw co hosting along with Matt Egan.

He is the global content and editorial directory director at Foundry, who is also representing the UK in this in this episode, also we have Chris Holmes. He is the foundry editorial director for the Asia Pacific region. He is in Singapore, and Qiraat Attar, Associate Editor for Foundry India.

Hello, everybody. Hello. Hi Keith. Hi Keith. All right, thank you for joining the show this time around. And this topic this month is around analytics.

But because everything is about generative AI these days, we're also going to be talking about AI and how to prepare your enterprise analytics for AI success.

So, Matt, can you just kind of frame the the discussion for, you know, why, why analytics and AI are so important, or, you know, as we, as we move forward? Yeah, sure. Matt Egan

I mean, I think analytics is extremely important to any organization, certainly any forward thinking organization that's looking to both on the inside, improve its existing products and services, and externally, build new products and services. Why AI? Well, to your point, because everything's AI at this point.

But I think when we think about AI, it is all about data as its analytics and on the one hand, to be successful in developing and training and deploying AI applications and projects. Your data needs to be in order, right?

You need to have a good way of accessing clean clean data, and your analytics tools and platforms need to be ready for the unique demands of AI.

At the same time, I think some it buyers will tell you that AI can potentially solve some of those issues with data management, although that's a somewhat contentious point in our editorial spotlight report this month, we're going to focus on how you can take your current analytics capabilities to the next level and position your organization to capture AI's potential, both, As I say, to boost productivity inside and to innovate externally.

So I think, like, yes, we're here to talk about analytics and the importance of it today and tomorrow. It's intrinsically linked with AI at this point. Keith, right? Keith Shaw

And so we usually start to show off with a couple of stats. So we've, we've gathered three different statistics from either research reports or surveys. And the first one was from IDC, and this was a quote from Stuart Bond.

He is the vice president of data intelligence and data integration software research at IDC, and he recently said this quote, your organization may be ready for AI, but there is a very good chance that your data and your data architecture is not ready for AI.

83% of organizations changed their data strategy because of generative AI making the delivery of high quality data and analytics products to AI processes within appropriate security and compliance guardrails as their top objective.

Second stat out there, Omnia Strategy Group reported that only 28% of respondents felt that their data centers were fully prepared to handle AI computational demands, and 46% said integrating with AI and machine learning workloads is one of their biggest challenges.

And Gartner has chimed in with this estimate, basically saying that preparing data for AI will improve business outcomes by 20% which means that data must be appropriate for use cases intended.

They said a key reason why 30% of internal projects are abandoned are because of poor data quality inputs. So what I want to ask the panel here is, as you're talking with CSOs, CIOs, IT leaders out there. Are they? Like, how ready are there?

How much readiness are they reporting? Again, because generative AI came in in 2023 like a whirlwind, and a lot of companies were still just trying to finish their digital transformation or other data projects. So what's the level of readiness out there in the world today? So Qiraat Attar

what a lot of CIOs have told me in our discussions is that the COVID 19 pandemic sort of fast tracked their adoption of, you know, digital sort of platforms and fast track the digital transformation projects, and that also led to a lot of data coming into their fold in a very organized way, because it was like asking your customers to log in.

So that gives you login information. And even if encrypted, you're getting sort of rows of data which may clearly outline who your customers are. And that's something that we witnessed with pharmaceutical businesses, healthcare businesses, a lot.

So I think when they got that, it almost seemed to them that it this would be a missed opportunity if they didn't start immediately drawing insights, etc, from it so early, what I have actually witnessed is that a lot of platforms are operating at a level of sophistication that I frankly did not anticipate, because they managed to leverage data right, and because the very first step of collecting data, they rolled out the kind of forms, etc, that ensured that they get slightly more cleaner data than not.

This is not true for all industries. Obviously, e commerce is probably ahead quick e commerce, that is the 10 minute 15 minute delivery. Businesses are probably doing a better job than, say, little more sort of legacy industries like manufacturing, mining.

You know, maybe conglomerates with some some departments are slower than the others, so maybe not everyone's on an equal footing. But data, being able to collect data going to the pandemic and digital transformation has helped them get a leg up.

So an unfortunate circumstance that actually led to transformation in this regard. And Keith Shaw

Chris, what are the IT leaders in the Asia Pacific region telling you around there, I'm around the idea of readiness. Christopher Holmes

So I think it's an interesting question, Keith, because again, we've seen a lot of organizations pivot now from talking around digital transformation to AI led transformation, which is almost sort of adding AI on top of their existing digital transformation projects.

And again, it's taking the new tools and new technology and being able to deploy those obviously underpinning that is, you know, data, and actually getting the data into the organization.

And as Qiraat said, we've as a result of the pandemic, we saw that digital trans well as before the pandemic, and it was accelerated by the pandemic, we saw that data come into the organizations.

And again, a lot of organizations were figuring what to what to actually do with this data. And as we've got this new tool around, sort of AI, I mean, you started to see people start to analyze that, start to look at it.

So it's almost like this transition from using the analytics now to sort of moving towards the towards AI. And it's really interesting. I mean, we're now starting to see companies add AI roles to organize, to their organization, but they haven't actually deployed the use cases yet.

So again, there's that anticipation that we're starting to see that.

I mean, I was talking to a tech leader at a wildlife park, and they're looking to deploy AI across the wildlife park so that as the visitor, visitor walks around, they can actually get information on the animals that they're actually looking at, you know, through the phone, through scanning barcodes, etc.

So again, there's a big transition there to actually making that customer experience, digitizing that and actually starting to, as I say, personalize that more, using the AI tools.

So again, there's a lot happening at the moment, and it's it does vary across the different countries, because, again, we do have different challenges in terms of maturity, even connectivity. So yeah, there's a lot going on. And actually, I mean, it was interesting.

Qiraat was talking about sort of mining and manufacturing. I mean, we've seen some really interesting adoptions there around visual inspection.

So again, starting to see camera technology actually being deployed in drones, and then being able to do some of the analysis of that, whether it be in inspection of products mine sites, plantation management, that was another one that was coming out through actually the use in agriculture.

So there's a lot happening out there. Keith Shaw

All right, in Matt in the UK, same kind of responses that you're getting, or is there anything? Again, because the stats are saying that a lot of companies are not ready.

But it feels like that, at least in the first two regions that they that they're recognizing that in and moving forward. I think that's Matt Egan right.

I think I would reckon that I would recognize a lot of what Qiraat and Chris were speaking about there. I think we would in the US as well, right? Keith, I think there's some real themes in terms of data management and analytics AI in general.

I think there are these three general areas where our it buys. They're thinking about AI and the first is this concept of productivity, bringing AI automation into data management in order to improve operational efficiency, automate repetitive tasks, natural language interactions.

The examples that the folks just gave, you know, stretching from mining through to, I presume it's the Singapore Zoo, but, you know, both use that kind of, those kind of principles. And then there's the second area. Is what we're focused on.

Might be kind of called responsible AI, properly managing intelligence about your data.

And there's a whole third area, right, which is agentic AI, next phase of AI maturity, another tool to bring in automation into data management now, in the form of autonomous action, real time and data becoming crucial in terms of the levels of preparation.

If I think specifically about the UK, if I extend that to EMEA and our US audience as well, it is a very mixed bag.

There are organizations who do have access to data, processes, platforms and infrastructure, which means they can just simply apply AI solutions to accelerate, I think more often we speak to it buyers who tell us that candidly that their organizations have challenges around these areas.

And even in that second group, they fall into two different groups, right? Some people need to fix their underlying infrastructure, process, data management in order to apply AI, right?

In order without they're unable to get to those analytics until they fix the base I am hearing more from it buyers who are prepared to kind of throw AI across.

Problem, and sort of fix those issues around data management, like they saying that their path to everything being perfect might be too long, so they're happy to add AI infused tools and platforms and generate actionable insights, even knowing that the input data may be a little flawed, like on the on the basis that you know insights today is better than building for perfection tomorrow.

I am trying to hear that more and more. Does Keith Shaw

that mean that companies might be able to leapfrog some of the the you know, using AI to jump ahead or at least catch up? I think that's Matt Egan

certainly a working theory within some organizations, and to the point that both Chris and Qiraat made that may depend, may differ, right?

If you're working in diagnostic healthcare, I think it's unlikely that you're going to be a cavalier about using an agentic AI solution to kind of more creatively promote solutions. But if your goal is to kind of refine an operational process within a big warehouse or something like that.

You might be prepared to take the risk that some element of the calculation isn't perfect, if it allows you to accelerate everywhere else, right, Keith Shaw right?

And that also comes to the debate of the creativity versus accuracy part of what are you using generative AI for?

And I think there's a little bit more leeway when you're talking about maybe creative or ideas or generating kind of that brainstorming type of idea versus something where it's like question and answer. You definitely need accuracy there. Chris, I think you wanted to jump in with something right.

Christopher Holmes

I mean, it was interesting. I mean, we were talking to a regional telco, and one of the things that this leader said was the expectation around small language models, specifically custom models. So instead of trying to use AI to solve the world's problems, actually getting very, very focused.

And I think that's probably something that we'll see accelerate.

And it comes back to sport Matt's point as well, where it's really around starting to sort of actually, let's be specific, let's target, let's actually go and use it to actually solve a problem and show and demonstrate that return on investment.

So I think that's, that's another thing that we're actually starting to hear Keith Shaw

all right, and so I want to jump ahead a little bit and go with the next question about the challenges that a lot of these companies might be experiencing when we talk about AI in any aspect.

You know, the, you know, there's all is, this is how it's going to change the world.

But then you look at and you go, okay, but we got to get over this hurdle, this hurdle, this hurdle, when, when it comes to preparing the data for analytics with, you know, in order to make that data better, what are some of the concerns that you're hearing about?

Chris, I'll start with you. There's Christopher Holmes

a couple of areas that I want to talk about around this. Keith, first one, I think, is around collection and connectivity.

Because again, if you look across the region, we still don't have great connectivity across large parts of the region, you know, if I look at Philippines, Indonesia, even parts of Malaysia, Thailand, you know, we're still not got full connectivity.

So again, when we start to talk around, you know, the mine sites, the plantation sites, even sort of agriculture remote villages, we still have challenges in actually collecting data from what's actually going on there.

And there are various initiatives that go on there, where you've got sort of mobile retail stores going in. They do the transactions actually within that mobile, that van, if you like, and then when it drives and gets back to connectivity, then it's all uploaded.

So you start to see that sort of data collection going on. So I think that that's one area and the other area. I think that what I want to mention is actually, is the organization actually ready for this?

I mean, we were doing a workshop with a payments provider recently. And again, the conversation went was, should we be actually looking at AI as just a bolt on tool, or should we actually be starting to really restructure the organization to take full advantage of this?

And I think that's another challenge that, again, it's one thing just to sort of do the bolt on bit, use the chat, GPT, the deep sea, whatever it is, and co pilot to drive some of that connectivity.

But is that really going to drive that next wave of business solution? And I think we're almost seeing, if you like two waves of two waves of adoption. The first being around the productivity improvements, the productivity enhancements. Again, you know, po matching those kind of things.

But then the second wave is some of those business critical, mission critical, industry critical solutions. So again, we start to talk around diagnostics in healthcare.

You know, I was talking to a CIO recently, and that's one of the things that they've actually deployed company wide, one of the few who have actually done something across the entire enterprise to actually make that difference. So yeah, there's still a lot of challenges going on.

And again, you can talk around the skills gap. There's still the security and privacy, you know, the garbage in, garbage out. It still hasn't changed.

It doesn't matter whether we're talking about analytics or whether we're talking about AI, you know, if the data is rubbish, rubbish in equals rubbish out. And, yeah we've still got a long way to go. Keith Shaw

Qiraat talk about some of the concerns that are going on in India. What are they most concerned about? Some Qiraat Attar

of the concerns, frankly, are developing on the go. For instance, we have an act which has been formulated, the bill has been tabled. It might get implemented.

Sometimes, this year, next year, digital, digital privacy and data protection act 2025, and that has led to industries sort of adopting a little bit of caution, because yes, they've been, you know, collecting data.

Yes, the customers have consented, but the act might probably change what you're allowed to do with the data, and will also supposed to empower your sort of end customer as to if they have the right to be forgotten, which they do, and if they do not want certain insights, etc, to take place.

So in the and the fines if you sort of violate an act like that is too huge, and the industry will of the the enterprise will suffer a lot of reputation damage as well.

So when caution steps in, I think certain things they might want to sort of scale back on, just so that they are, you know, in the clear above board. Don't make any mistakes in terms of, you know, just for to progress analytics wise and AI wise.

And the other thing that have, that I've been hearing about a lot is that the skill gap actually is causing more of a hindrance, I guess, than the imagine.

Because the enthusiasm is there, the enthusiasm is there to sort of have aI touch upon every, every aspect in the business.

But if you don't have the right talent, if they don't know and what the business needs is for their tech talent to think business first, to sort of ignore the FOMO that comes with, okay, if it's a tool, I have to try it and be like, is it even relevant to the business?

Because we're talking about even trying a use case is cost intensive exercise, and actually adopting it to implement it all wide is also cost intensive.

So I can just imagine, from an organizational perspective, to implement something today and maybe see three or four years down the line that it hasn't yielded the reward that it should have with the amount of cost invested.

So that is something that CIOs and all tech leaders are thinking about when they are at this stage, at this time. Keith Shaw

Matt, any any other concerns in the UK? Or is it all, all of the above? Basically, I mean thinking Matt Egan

specifically about analytics, right, which was what we're here to talk about today, as we've discussed quite a lot already, the underlying data quality infrastructure issue that's real. That's top of mind. I think that exists everywhere. Really interesting hearing Qiraat talk there about, in essence, ROI Right?

Return on Investment. I think that's a real challenge for a lot of organizations around these projects. I love the way keyra framed it. As you know, don't worry about the FOMO kind of thing, like, like, drill down to the value.

At the same time, organizations are worried about being left behind and from an innovation perspective.

But I do think that issue of ROI is increasingly becoming a space where some organizations can see AI helping right data preparation, in and of itself, can consume a huge amount of time and resources due to those difficulties in finding, accessing, cleaning, transforming, sharing data efficiently.

I think the increasing number and complexity of data sources coupled with the need to access them across distributed ecosystems, again, both the guys spoke about that that demands significant resources and expertise, and I'm starting to hear it buyers think maybe AI can help with some of that complexity, like applying AI to an imperfect system, as I mentioned earlier.

Um, other things I think IT teams are often overwhelmed by the rising requests for self serving data access and integration, varying data requirements from different users, complicating the process further.

So again, starting to see some opportunities for AI supported data platforms to help like reduce some of the challenges around data preparation and management, incompatible data types, formats, aging data. These things all pose obstacles to effective data access and collection.

Skills Gap, which Qiraat hit on, I think the data related skill gaps are further hindering the development of robust data management as well as AI related skills gaps.

It's another area where actually, I think organizations are starting to think about AI supported data platforms, or potentially agentic AI helping to winnow data into insights. Again, there's a risk involved, because you're not doing it through human insight, but potentially it could be helpful and work.

But with all of these, these pieces, all of these challenges, I think there is one underlying challenge that that we see and pretty much everywhere, which is, you know, you have the question of, can it be done?

And after you answer, and this is where Qiraat was coming in, there are the questions of, will it work and should we do it? Because I think.

Think the other questions that I'm hearing a lot, the other challenges that I'm hearing a lot are this year as opposed to last year. And Keith, you and I have spoken about this many times.

The other challenge with AI applied to analytics is, okay, we can generate insights, but will those insights help us? Can we trust them? And then there's the big question of, how are we managing the risks? Because some of those risks are unforeseen.

So it definitely is all of the above, from what what the guys were saying before, plus some others, Keith Shaw

I wanted to bring up the the fact that I think the skills gap is is something that everybody's seeing and facing around the world. It's this is going to be a big issue for addressing a lot of these problems.

But Matt, you brought up that, that issue of trust and explainability, it still feels like everybody knows that we're supposed to be addressing it, but I don't think anyone's ever happy with the answers that they get from the AI, you know, how do I know that that answer is correct?

It's almost like back when I was in school and we were taking, you know, I was learning geography, geometry, sorry, wow, math, basically. And you know, the answer was, show your work, like, show me that, you know the answer is correct.

And there was a lot of training of of like, Why do I have to show you I know it's right?

And, you know, they're like, No, you have to show how you got to that point, and it doesn't feel like that we're that we're we as humans are pushing that as much as we should. When Matt Egan

I speak to it buyers, specifically in the area of analytics, I do hear these kind of two competing voices that are in their heads. Right?

The one is we can really accelerate by by doing things that we might not fully understand the implications of but they can help us get there quicker.

And then there's what you're talking about there, Keith, which is, if we don't understand how the thing works, we don't understand how it gets into the insights A) Can we trust them?

And B) like, like, it's a little bit like, you know, I might be able to show my workings in geometry, but earlier on, the power in my house went out, and I don't know how to fix that, right, whereas my grandfather definitely could.

And I feel as that from the skills perspective, there is that, that risk here in that every time we're kind of replacing a human process with something that's driven by purely by technology, we are losing the capacity to manage that on a human level.

It helps us accelerate, but we lose some skill and some knowledge that we may need further down the line. Yeah. Keith Shaw

I mean, there's, there's the question of whether, you know, people will, will know how to code in the future, if, if AI is basically taking over all of that entry level coding. And that scares a lot of people, too.

Carrot, I wanted to ask you another question about, you know, especially this is around the term, this is just the term analytics that we're using.

Is this just a rebranding, and now we're just calling what used, we used to call it analytics, and now we're just calling it AI. Is it basically just another term, the same way that we refer to automation, robotic process, automation, all of that stuff.

We've all been around the IT industry long enough to know that, you know, sometimes they just need to rebrand something if it's not hitting and then maybe the new term will get everybody excited.

Is that the same case, or is this a really a new thing that we're all experiencing? I Qiraat Attar

think in this one, I'd honestly say it's on an organization versus organization perspective, because I do genuinely think that there are some organizations who would frankly be quite offended if you told them that their analytics is really, that their analytics is just, you know, rebranding, because they got on to the AI bandwagon before generative AI actually blew up the way it did in 2023 and they've been doing the Initial Case analytics using natural language processing, you know, like sort of the slower use, slower use cases.

They've been trying to implement them, and seeing when there was more room to fail, because people are not expecting as much. So they've been implementing it.

And there have been CIOs have told me that the conversations that they're hearing, they start talking about in their own organizations about a decade prior. So not for everyone.

Would I say that this is merely a rebranding, and for others, I would say that analytics and mathematical modeling and machine learning, all of them without AI, require a lot from a decision scientist or a data scientist, etc. That is, you need to know every case.

You need to adjust every variable that is expected from a person that is expected from a tech talent today, they have AI to run, I think, several 1000 use cases which drastically reduces time, drastically reduces instances of error, really factors in over correction, etc.

So there is AI which is really sort of helping you get that extra point 1% of precision.

So I think that is what it is being used for, which is, which is really minor thing, but if you're in an industry, if you're in an organization, you'll witness that AI is able to bring that additional improvement, which, I guess, what every new technology is supposed to do.

Keith Shaw

Maybe it's just an editor thing. Maybe it's just a journalist that irritates us. But you know, I obviously, if the people that you're talking with, the IT buyers.

Out there, if they're if they're generally excited about it, I probably think that they don't care what it's called, as long as it gets the job done right. I think Matt Egan

the right is right in my experience, in that it does vary from organization to organization. And we all live through the cloud era, where if anything was connected to the internet, like people slapped the word cloud on it, and that made it somehow cooler.

And I think certainly in 2024 Keith, you and I have spoken about this many times, people got projects away by calling them AI, even if they were machine learning, right?

But I'm sensing increasingly that there is a distinction that people are comfortable with, which is, there's, there's, there's fast processing, right? There's been able to process a ton of data quickly, and you can call that AI. You could call it machine learning.

It is what it is, kind of thing, I think, where Qiraat referred to increased precision, which is definitely one part of it.

But then there's also this idea that it's that I've heard a couple of times in this conversation, which is using AI to extract insights from data in a way that learns and develops and that's generating insights, even if you're not 1,000% sure that it's based 100% in truth, I should have, I should have done more math.

Keith, clearly, and I think what we're seeing and hearing in 2025 is that I feel like organizations it buyers, are getting a bit more comfortable with the latter, not wholly comfortable, but more comfortable with like using, using AI to generate insights, rather than to process things quickly.

And in part that is tied to this idea that return on investment needs to be demonstrated. Keith Shaw

And I want to get, I want to get back to the some of the stories that you're hearing from companies that you talk with, Chris, I want to ask you to see with especially with the speed of the innovation that's going on in the AI are you?

Are you hearing about anybody that feels like they're that they're behind, that they're not being able to catch up? I mean, we're starting to talk about agentic AI in the next phase.

And you know, you know that if you haven't gotten even to your first AI project, yet there might be a feeling that you're way behind. Or do you feel like companies are like, Oh no, no, we've got a handle on this. So Christopher Holmes yeah.

Again, I think it's a bit of a mixed bag piece. I mean, some of the interesting, interesting discussions we've been having recently.

I mean, it was a Malaysian University we were talking to, and again, they would one of the things we we sort of touched on was around the cost of actually deploying AI technologies.

And again, got to remember, we're dealing with some quite low cost countries, and one of the things that came out from there was they were very excited about seeing the low cost AI technology coming out from China, and being able to deploy that.

So again, there's also some decisions that need to be made around, who do you actually go with? And again, this is one of the hints, as we're seeing.

Because I think, as you said right at the beginning, it's evolving so fast, so quickly, there's still that challenge, is that okay, who are we going to go with? Who are we going to, you know, partner with? And I think people is, because it's evolving so fast.

And as you know, Matt was talking about agentic AI coming in, what's that going to enable? So again, there's a little bit of a wait and see. I think we're starting to see from a number of organizations. And again, particularly the skill shortage thing.

I mean, it's one thing saying we're going to have an AI project, but can we actually execute it? I mean, actually deliver it?

And again, there's people have been sort of bitten around some of the digital transformation projects we started on this right at the beginning, then they didn't really deliver everything they wanted to.

And so I think there's still a little bit of a concern around actually pushing out some of these AI projects, especially as most of the use cases people are talking about at the moment are customer facing.

You know, there's a lot of conversation around, sort of the chat box, the, I mean, I'm hearing a lot of work. I'm hearing a lot of discussions around the use of hyper hyper personalization, hyper automation, you know, all around, sort of driving that next wave of insight.

But you can't be personalizing and getting it wrong. You also can't be automating it wrong. So again, I think we're still waiting to sort of see some of these use cases, see some of these examples coming out. And again, I mean, just, just to, just to come back.

I mean, we're still seeing very few organizations actually adopt AI at scale across the enterprise. I mean, at the moment, if I sort of think around tech leaders, I talk to it, we're probably less than 10% at the moment, have actually deployed something enterprise wide.

So it's still quite nascent. Keith Shaw

Qiraat at anything, you know, are you hearing the same thing, where companies might feel like that they're behind or or are they, are they ready for this new world? And they just adapt as quickly as they can. Qiraat Attar

You know, in this one, I think I have to say that whether they are implementing at the rate at which everyone else is implementing, what they want to project is a lot of confidence that they've got this they understand what it takes. They have buy in.

Maybe there's caution behind the scenes. But. Certainly do want to say that when they're speaking to somebody, all the points that Chris mentioned is are very, very true.

They do have some considerations around do they even have the talent to execute the like I like to call the big castle dreams of, you know, implementing AI and have it all be perfect. They're working towards it.

And one thing in particular which I cannot remember, who said it to me, but a very prominent CIO said that the it's not all internal.

They're expecting that when they bring in a vendor or a partner and trying to get their software, they're expecting thought leadership from them that tell us what we probably don't know, what we're not looking at, that your your tools need to be that adaptable, because they're like, if the technology, if every one and a half year, if every 18 months, you're gonna hear a new buzzword, and then you're gonna it's already gonna look like you're behind, because the buzzword is gonna take over.

The softwares or technologies that they're bringing in should be adaptable, that more things you know, you should be able to build with it, rather than being like, oh, a new thing has come.

Now, I need to do away with this thing altogether and get a new technology that just doesn't make sense.

And honestly, CIOs, with a lot of experience, 2030, years plus in the business, they know that fads come and go, but like long lasting solutions as what is permanent, so they're actually looking at it from that perspective. Christopher Holmes

Okay, can I just add something to that? Because sure, he was talking to a chemical manufacturer, and they were actually talking about, they want to see these AI first solutions emerging.

So again, we're not just talking about the bolt ons that we're seeing from a lot of the vendors at the moment, but actually solutions actually designed in the era of AI.

And again, they were holding out for that as a sort of a way to actually look at how we embrace those and deploy those across the organization.

So I think there's also a little bit of wait and see, rather than the, you know, the bolt ons, have we been talking about the sort of, yeah, we've, i ai enabled something, you know, but have we really, so I think there's a little bit wait and see there.

There's Matt Egan

really interesting cultural aspects of that, right? I mean, I definitely recognize Qiraat at the CIO over large organization in India, they're never going to say that they're behind on a thing, right? It's really important for them.

Whereas, if you have a, if you have a conversation, certainly behind closed doors in the UK, like I've not yet met any it buyer who will tell us that they're done with AI, or that they're succeeding, right?

Like that is always a sense that there's something more to be done. Because, you know, arguably, we're, we're experiencing something akin to an industrial revolution, right? And I see smart leaders combining everything that's being discussed here, right? There's an element of wait and see.

There's an element of where you can, yeah, pull in those, those vendors and those products and services and plug them in, especially as it relates to internal processes, right in accelerating your internal access to analytics. But I also see smart leaders trying to build future proofed organizations.

So they're trying to build technology stacks that can grow and support innovation and developmentally on the horizon. Again, going back to a point Qiraat made beautifully there, which is kind of expecting the CTOs of their vendors to come in and help them solve their tech stack problem.

So I think it is about understanding that you can't wait for perfection to get started and to start getting wins, but we can't expect to be perfect at this stage and like maybe some of those big revolutionary changes.

It is a case of kind of waiting and seeing where the tech flows. Keith Shaw

I think, Matt, you answered my next question before I could even ask it, which was the, you know, what are companies doing to get these data, tools and the AI integrated in? Are they developing their own or, you know, it's that whole build or buy question.

Or are they grabbing platforms and packages? I think, Chris, you even answered that. You know, companies are waiting for an AI first package, rather than just some of these add ons. Or are they using their own internal teams?

It feels like even if you Yeah, you couldn't, you know if, especially if, there's a skills gap, you're probably not building it yourself. So again, it's probably all over the place. There's never good, there's never a good yes and no answer, no. Matt Egan

But I think, like, like I said, I do think there is a reasonably clear distinction between AI inside and AI outside. Right? Ai inside being we do a thing. Can we do it quicker, better, like, more more inexpensive, less expensively, more efficiently, by applying AI to the problem.

And like often, you can get an off the shelf platform, data management platform, for instance, that will maybe it's your existing vendor. They're adding functions that are helping you to do that.

But then there's the piece Chris was talking about there, which is the products and services, the customer journey, the thing that we do for our customers, we want AI baked into that. That's AI outside, that's much more creative, that's likely to be more agentic.

That is kind of, and Chris mentioned it before, you can't do it 99% right. You've got to do 100% right, right? So there is a real kind of difference, and you're probably building that technology. In house, because that's your IP versus what's happening inside. Keith Shaw

One final question before we get to the end of the show vote section, are companies generally, you know, the companies that do have a solid data and analytics process, are they very enthusiastic about what AI is going to be able to do for them?

It feels like the general sense is, is that, you know, yes, it is, or is it still too early? You know, in this case, like, I don't see anybody saying, Oh, that AI thing that's never going to work.

I think everyone's optimistic about what AI can achieve without falling for, you know, the bubble or the hype. It does feel like there's a lot of realism here, but there's nobody that's that's negative. On the on the technology is there. Christopher Holmes

I've not, I've not come across anyone that's negative. I mean, but coming back to your point, I think the interesting one is those that have already used Analytics, you know, they've had a strong analytics, they've got the clean data. They're certainly going to accelerate with their AI adoption.

I mean, again, I can think, around a beer company, they put in place a digital twin across their manufacturing process. And so with that, you've actually created a model of how that manufacturing process operates.

Now, with traditional analytics, you would then have to determine which sort of scenarios you run, which sort of parameters you change as soon as you actually start to bring AI into this that can actually optimize how that did, how you actually run the simulation models on that digital twin, which will enable you to get to a result much, much faster.

So again, those organizations have already spent a lot of time on their digital transformation. They're now ready for that AI led transformation, and I think they will actually accelerate much faster when they get those right tools. And that's the positive side, if there's not a negative one there.

I don't think I've heard anybody say no, but there's just a lot of lot of head scratching and going on at the moment. Keith Shaw

You know, it reminds me of another story that when I was, I was covering robotics and automation a few years ago, and I was talking with this brilliant like PhD guys. And Chris, I know you have a PhD as well. So Matt Egan

mention it actually he normally mentioned, yeah, Keith Shaw

one of your one of your peers, was talking to me about about an autonomous vehicle program that they were working on, and he says he told the story about he went to his data team and said, what do you what do you need to make this better?

And the answer, unequivocally, was more data, more data, more data. And they were never happy with the data that they had.

So I wanted to ask, like, it feels like, you know, are are companies that you're talking to happy with the day data they have, or are they always looking for more data to feed into the algorithm to get the better decisions and the better insights?

I think it Christopher Holmes

depends on the use case. Again. I mean, you know, autonomous vehicles. When I first arrived in Singapore, which was 25 years ago, I was working at a research institute. They had autonomous vehicle program 25 years ago.

They were talking about it then, and it was all around gathering the data, but you were restricted by the amount of data you could actually collect, just because of the bandwidth, the cameras, the sensors, everything else, and all that has actually improved dramatically.

So again, I don't think you can actually have too much data, but you need to make sure you've got the right data.

And I think that comment I made earlier about actually looking at those small language model, being able to focus on specific use cases, being able to actually deploy the AI there.

I think that's the way a lot of organizations are actually going to go just to actually test the technology, and they've got more control over that, rather than sort of trying to sort of hide set but change the world by sort of trying to deploy AI everywhere.

I don't think that's going to be the way that we're going to see organizations move forward, and certainly not with the tech leaders I'm talking with.

You know, there's still, they still want to keep a wall around it, keep control of it, and again, you don't want to be front page of the newspaper if it goes wrong. Keith Shaw Right?

I think that's that's the big fear, at least for a lot of us, companies, especially with the speed too. It feels like someone told me that an average enterprise, it takes them a year to deploy a project, just to get through the legal process.

And that just boggles my mind when you see how fast a lot of these these things are innovating. But sometimes companies are companies, and they just have to move it at their own pace. Qiraat, everyone, everyone enthusiastic over in India about this?

Or is there some hesitation because maybe of some of these fears? So Qiraat Attar

I don't think that there's a hesitation. You know, when we were talking about the some companies have a wait and see approach.

The wait part is actually the time that is taking place right now, which is when CIOs and tech leaders take these proposals about AI, implementation agent, take AI, what not to the business leadership to the C suite, and the C suite says.

Provide proof before we can go ahead. That's where the weight comes in.

So no, there is, there is no negativity, but I think that the enthusiasm needs to be supported by use cases that are actually working, and data that is actually usable, and returns that they're not expecting immediate returns.

You know, everyone agrees that room wasn't built in a day, and AI is not going to give you, it's not a magic wand, but they're like, at least an inkling of the right kind of results.

If can be shown to us, then we'd be enthusiastic about it too, and you put, like, the weight and the money behind it.

So I haven't witnessed negativity, but I have seen that CIOs are engaged in we did a wonderful feature on CIOs transforming into business strategists, and this actually the AI and the amount of it's almost like a CIO probably sees and the AI's potential first before a business leader, like a CEO or a founder does, and then you make an argument case on AI's behalf, being like, it'll work if we have these such and such, you know, things going for us.

So that's what I can see that CIOs are doing, which is why they're involved in a lot of business strategy decisions and making a lot of arguments in favor, while they're also realizing that, okay, these are the aspects that we still need to figure out.

This something that I have witnessed. I'm sort of on the periphery of hearing these conversations. Matt, Keith Shaw

Matt, any other, any other points before we get to Well, I got one more point that I want to ask or question. We can cut this out if we don't like it.

All right, so I've got one more thing around the explainability and the, you know, getting, getting the AI to answer the question correctly, or how do we know that that's correct?

I think probably most of us have been in the industry for a long time, and I remember in the early, you know, this was maybe like 30 years ago when data warehousing was coming out, and it was all this, like big analytics stuff that that was going on.

And I think we all heard the story of the whole diapers and beer scenario. Did you guys hear about all this?

All right, so, so the the pitch on data warehousing and why you need analytics in your company is that a grocery store, discovered that new fathers and new parents were coming into the grocery store, and they were buying diapers for their newborn baby, but then they also were also buying beer, so they had all of this data.

So then, because of that, they decided to move the diapers right next to the beer in their grocery store so that you could get the new fathers out quicker.

I'm just thinking now that now we're in the world of the AI, and if an AI came up with that analytics, I think most of the humans would go, that's crazy, or, like, explain why you're telling me, you know, to do this.

And so I don't know if we're there yet, or if we just assume that that's correct, I think, Matt Egan

I think it speaks to that's a really interesting example.

And and also, as, I mean, my children are older now, but it definitely resonates right from the days when I was a and I'd love to say that the beer aspects of it has diminished over the years, but it really hasn't.

It's a really interesting example, because it speaks to having the right amount of data which Chris hit on before right context is everything. So you can, you can extrapolate, extrapolate an insight from, you know, in the wrong amount of data, or the wrong cut of data.

And you may, you know, you may come to a conclusion that is incorrect, unlike that conclusion kind of thing.

So context is everything which is where the kind of which is where the art and science of this piece is it's, I didn't expect this conversation to to end in it's kind of like a remake of the movie short circuit, right?

We're all going around like Johnny Five saying, more data, more data, more data. But that kind of, you know, but it's not the worst example, because the data that Johnny Five was consuming was lots and lots of human written books right in order to get context around things.

So, yeah, I think, I think every organization is feeling the opportunity, but the threat is twofold.

Is that your rivals over innovate and go past you, or exactly what you're describing there, which is, which is, you rely too much on AI driven insight, and you don't have the right data input, and you reach the wrong conclusions.

You know, the other analogy is on the the story of the there was a plane that was flying from Canada to Korea, and one digit was put in wrong in the initial direction finder when it took off, and this was the plane that flew all the way across the Soviet USSR and nearly caused a nuclear war in the 80s, and it was one digit wrong, but because all the trust was put in the data rather than the context, the pilots didn't notice until it was way too late.

So I think that's that's the opportunity and the threat. And I feel like every buyer you speak to is feeling some version of that? Maybe the Keith Shaw

answer isn't, isn't more data, more data, more data, but good data, good data, good data in your clean data, right? Christopher Holmes

And I think it comes down to the use case. It's got to be, what are you going to use it for? What are you trying to actually solve here? Because again, if you say just give me everything, that's not going to help you.

You need to be to be focused, you know? I mean, if you talk about the retail example, if you think about it now, we'd have video cameras. We'd have sensors on the shelves. We'd be checking the, you know, we'd be checking the payment process.

We'd be looking at how long it took them to get much more focused data around that, to enable those decisions. And just finish that up. Keith, they're nappies now, so again, we'd have nappies, not diapers and garbage, okay, Keith Shaw

oh, you know, my US bias is just coming through on this show, and so I apologize. I should start every show, and I apologize for being an American, a US. All right, all right. So we're gonna, we're gonna end the show with our, with our, with our vote.

We, you know, we've done this on our previous shows it sometimes goes well, it sometimes doesn't. It's either a yes, no, and usually sometimes we go it depends.

But if you can keep the answer short, and the question that I have for this month is, is there enough time left for companies to get onto the AI train with their data and analytics offerings, or have they missed the opportunity?

And this might be a very, very small segment of the world, because it feels like everybody's already on board this train. But is there, you know, is it too late, or is there still enough time? Matt, I'm going to start with you.

I just love putting pressure on you to start us off. It's Matt Egan

definitely not too late, like and I think it's going to be an ongoing process for a good while. Keith Shaw okay?

Qiraat, Qiraat Attar

I think there's still time, because everyone's ramping up, and I'm sure that someone will make mistakes, or somebody's just starting out, playing field will get leveled. Chris, Christopher Holmes totally.

Still plenty of time out there, still plenty of time to sit there get your data ready. But when it does come, you need to deploy fast. Keith Shaw Okay?

And I'm gonna, I'll agree, I think there is enough time, because companies that might be behind can use the AI technologies to catch up, maybe not jump ahead, but at least catch up to where they might feel like they're falling behind.

And of course, you're never going to get anybody to admit that they're falling behind anyway, the whole question might be moot. Hey, thanks everybody this. This was a fantastic episode. Feel free to if you're watching this, to add any comments below.

You know, join us every week for episodes such as Tech Talk Today in Tech, CIO Leadership Live and DEMO.

If you are interested in seeing B to B product demonstrations, we're going to be back next month with more global editors talking about the topic of AI leadership in an AI world. So again, thanks again. I'm Keith Shaw, thanks for watching.