Audio transcript:
The Hidden Factors That Determine AI Success
Jeremy Hodge:
You've spoken before about the importance of being client zero. Yeah. When it comes to ai. Right. And what does that mean to make insight that proving ground for ai, and why is that a non-negotiable, uh, today?
Adrian Gregory:
Yeah. I, I think we've gotta use, uh, our own technology. And through that, through that use, you know, we've got to, uh, really bring that expertise, uh, you know, to our, to our customers. Uh, you know, you've heard the phrase Cobbler shoes, uh, I'm sure, which is, you know, the cobbler's children have, uh, the worst, uh, you know, shoes in, in, in the whole town. We, we just cannot be like that. Um, I mean, aside from that, there's so much benefit, uh, to be gained from it. Uh, there's so much efficiency, there's so much, uh, value creation for customers that it's not something that we can leave alone. Um, but it also helps us to just, you know, create those use cases, talk to customers about it. Um, and if you, if you think back to the days of outsourcing, and that was a big thing. It was quite typical to have your own CIO, uh, go talk to customers. Say, here's what we're doing. Here's how we're dealing with technology. Here's the roadmaps I've got, here's the architectural frameworks. So this is a, this is a similar thing. You know, we need to bring that internal expertise to our customers.
Jeremy:
Welcome to Insight on the podcast for leaders who need technology to deliver real results. No fluff, no filler, just the insight you need before your next big decision. Hi, I'm your host, Jeremy Hodge, and today we're getting insight on becoming an AI leader with Adrian Gregory, president of Insight emea. Let's go. Tell me about a story where we've taken, uh, you know, a high impact AI use case. What was the problem you were trying to solve, and what were the results that you saw from that?
Adrian:
Yeah, so one, um, is in the solution space. So, um, and it's actually a homegrown tool that, uh, that we've used, um, which is, uh, which was originally called Sales Coach. It's now, now called, uh, insight AI Coach. 'cause we've added, um, lots of different functionality to it. But the kind of the original use case, um, was that, you know, when we were building proposals for, for customers answering RFIs, RFPs, how do you really get across and understand what have we done for other customers? How to structure a statement of work, uh, correctly, how to produce a proposal correctly, how to think about pricing correctly. And really the problem statement was, we, we got an output that was really dependent on the individual team, uh, that was involved. Uh, so none of them incorrect, but just all had their different flavors from, you know, the different, the different teams.
Adrian:
And they would draw from their own knowledge, not necessarily from the wider corporate knowledge. Mm-hmm . So with, um, insight AI coach, it managed to, um, a pull together consistent quality responses. It could also toggle things like risk. Uh, so, so you can look at, you know, do I, do I wanna take on a deal of risk, uh, with this? Or I'm in an environment where I wanna take zero risk. Uh, how does the pricing work? What does the competitive pricing look like? So do a, a market scan, uh, you know, based on information that, uh, that would fed it and, you know, make sure you were competitive, um, in the marketplace as well. And then just, you know, produce a really, you know, quality standard response, um, that helps us to, you know, win more customers. Um, so that, that's an example of, you know, what, what we've used, uh, internally. It, it also helps to accelerate production of these, um, proposals. And, um, it also helps us to get more throughput with fewer people.
Jeremy:
And what has the team's reaction been to that?
Adrian:
Yeah, I, I think initially skeptical, you know, like all these things you go through the change curve. Yeah. Uh, so that initial, wow, this is great. This looks, uh, this looks absolutely fantastic. Um, and then, oh, what you want me to use this day to day and everything, uh, really, uh, but you, you know, that's kind of what I, what I did for a, you know, you know, for a, uh, for a job. And it was a key part of what I do day to day, to then a realization that actually this is getting better quality. I'm starting to get, you know, different references, different use cases, more specific, uh, references for the RFI, uh, or RFP that I'm responding to, or the statement of work I want to make. And, uh, I can do more of these. So, so it, it, but it, so it kind of went through that change curve, and I would say there was this initial euphoria mm-hmm . Then a period of why is the uptake not really happening? Uh, and then it sort of kicks in and really takes off and embeds into, you know, day-to-day operations.
Jeremy:
So it sounds like to overcome that skepticism, just having teams dive in and adopting is the best way so that they see the benefit and then can make it part of their workflow.
Adrian:
Yeah. I, I, I think so. I, you know, I just really emphasize this, um, management of change. You know, it's, it's super, super important. And sometimes you think that's obvious. You, you think, look, I, you know, just put in, uh, something that's fantastic and it'll work and people will pick it up and run with it. Uh, not So you, you need, you need the change management, um, around it. And, uh, you, you know, to really help people figure out how to, how to use this effectively. And, um, you know, in that role. And that's the same externally as well, you know, you come onto that. If, if, if we don't address this as part of our customer engagements, then uh, you know, it's, it's, it's not gonna go down as well as, uh, if we've got a really comprehensive change management approach. Right.
Jeremy:
So, ma making these proposals creation more streamlined. Yeah. Any other use cases that you think are really promising that you'd wanna highlight?
Adrian:
So another one is in, um, operations. So, uh, you know, we, we deal with a, a large volume of, uh, let's say more transactional quotations as well for, you know, product, be that hardware, software, pulling together bills of materials and, and, and things like that. And then, you know, how does that connect back into the logistics organization for, um, you know, for warehousing, shipping, you know, freight, all that sort of stuff. Um, and we have got, you know, quite a, uh, uh, a back office operation that, you know, deals with all that stuff. So we've, we've introduced, um, you know, uh, again, steadily, uh, you know, technology that will take on, uh, more of the mundane, uh, tasks, uh, that are involved in that in terms of, uh, you know, scraping together bills of materials, doing quality checks on them, uh, to make sure they're right, things aren't missed, or, you know, you've got wrong versions.
Adrian:
You know, you've had a new, new product, uh, you know, upgrade or things like that. So it kind of catches all those things. Um, and then make sure the pricing is correct. It's, you know, you've got the right skews number, uh, all, all of that quality check. And then it produces the kind of paperwork and out outputs to logistics. So, again, not something that's, you know, uh, you know, really taken away, uh, the need for the back office function, but just made some of the mundane tasks of, you know, producing checking, you know, which can be quite laborious. 'cause you're comparing, you know, one list against another, and then you, you, you know, just trying to, um, uh, you know, quality check in that way. It, it's kind of automated a lot of that, uh, and just made that job much easier and much easier to, uh, uh, you know, get closer to a hundred percent accuracy, um, in terms of those, uh, bills of materials quotations. And then, you know, what goes on into the, into the legit logistics function,
Jeremy:
I imagine to pull that up, it requires a lot of integration with existing systems, with existing data sources. Yeah. What did it look like to deploy that? How much time was spent having to do, preparing the data, making sure things could integrate?
Adrian:
Yeah, I, I mean, honestly, on the technology side, relatively straightforward. Um, and back to my point earlier where, you know, often times technology is not really the, uh, the longest poll in the tent on, on, on, on this stuff. Um, so honestly, getting that done is, um, is now a matter of weeks, you know, to build that, that type of thing. Um, and to deploy it, more of it is about the operational processes and making sure that it's part of the workflow. Um, and back to that, you know, change management, um, you know, that I spoke about, that's where it takes the time and to make sure that that's integrated effectively.
Jeremy:
Excellent. So it sounds like learning a lot from being client zero, you know, sort of the, the wins, the failures, the unexpected hurdles. Mm-hmm . How does that shape the advice that you're giving to clients?
Adrian:
Yeah, I, so back to, back to change management on, on, on what we found. So typically in a lot of client conversations, there's also a misconception about what constitutes, um, either a problem or an opportunity that needs AI to solve it. So what we find, um, talking to clients is in, in a lot of cases, uh, it's not actually an AI shaped problem, right? It's a classic data and analytics, um,
Jeremy:
Or automation. Or
Adrian:
Automation, some, you know, machine learning, some something like that, you know, that, that kind of a problem or opportunity to solve. Um, so the first thing, you know, that we've learned is look, you know, try and sift out here what's really AI and what's and what's not ai. Um, and then think about outcomes. Uh, so outcome focused. So the way we approach it with clients is talk about the business, talk about the, you know, the business outcomes that you want. Talk about the business operations, uh, that get you there. Uh, so there's, there's a lot of, a lot of talk about the business, um, upfront, rather, rather than the tech. Um, you, you know, because otherwise, if you start with technology and, you know, kind of, you know, work forwards from there, it will work backwards from there. Um, it's not gonna work. I, there's a famous Steve Jobs quote that says, you know, start with the customer experience and, um, and absolutely work backwards to the technology, not the other way around.
Adrian:
And, you know, he said that, uh, nearly, nearly 30 years ago. Um, and it's, it's still very relevant today. So that's kind of how we approach it. And then once you've got, um, you know, use case candidates identified, it's then the expertise about, well, okay, what's, um, you know, in a four box matrix matrix, what's, what's gonna deliver high value and what's easy to implement? Uh, you know, versus what's gonna deliver lower value and be quite hard to implement. You cross all those out. You definitely go for what's easy to implement, what's high value, uh, and then you've got some, you know, question marks, which might be more tricky to implement, deliver high value, but you might wanna think more carefully about those. And then you get into the tech, uh, you know, solutions go, go deploy. And, uh, you know, back to what's, I guess, become my favorite, uh, buzz phrase of the day, uh, change management, you know? Yeah. Back, back to that is really how to embed it in operations.
Jeremy:
I wanna go back to something you said about, you know, client comes to you with a problem, and sometimes it's just automation or data management. Yeah. When you eventually get to the technology discussion around that problem, and maybe there really isn't much AI involved, are they concerned about that? Is it more they just need whatever technology or whatever tools is gonna help them solve the problem? How do they feel about that?
Adrian:
No, I, I, I don't think that's particularly, um, an issue in, in many cases, uh, you know, organizations have got their own expertise around this. Uh, but perhaps they just, you know, because, um, their kind of focus group, if you like, is themselves, yeah. They don't necessarily have all of the experience of doing this with a whole bunch of other clients or, uh, you know, different, different cases. So, um, so no, they, they, they do bring, um, expertise in there. Um, they're not really offended that it's not ai and there's something else. It's, uh, not really an issue for us. No.
Jeremy:
So you talked about how, you know, so much of the processes aligning on the business problem you're trying to solve, making sure you're gonna drive business value and then getting to technology. But when you get to technology, there's this classic buy versus build dilemma mm-hmm . And that's not unique to ai. I think every technology shift we've seen that happens when you're evaluating that, is there a single question leaders can ask to figure out what's the right way to go? Or how, how do you advise them on that?
Adrian:
So it depends. It depends what you're trying to do. Um, you know, with, with the business, and I still think, um, we're in the problem solving, um, or opportunity capture phase with, with ai, which is, you know, discrete problems, discrete opportunities, and how do you use AI to, um, you know, uh, help to automate, drive intelligence, um, and drive different outcomes and, and do that much more effectively and efficiently than, than, than you could before? Where we've not got to is, you know, how do you run the fabric of the business, uh, on, on ai? You know, that's, um, maybe that's an emerging question. Um, but you've still got, you know, large systems of records, uh, that exist. Um, and, and largely business processes are based around those systems of record. Uh, and then you've just got a whole set of other things, other problems and opportunities to solve for, you know, around that data, uh, around external sources of data and around, you know, client journeys that you bring in, that you bring into that mix. So, so I still think we're quite, um, kind of discreet problem or opportunity solving stage at the moment.
Jeremy:
Do you think most companies underestimate the work they need to do on their data and their systems before they jump into one of these projects?
Adrian:
Yeah. I, I think it's becoming more and more, uh, accepted wisdom now. Um, and I'd also counter it a little bit by saying, you know, also don't overthink it. I, I, I mean, look, it's, um, it, it is of course much, much better to have all your data in the right shape format, um, in the right, you know, kind of data lake, data warehouse, um, you know, be able to in ingest external sources of data. Now, of course, of course, that's much, much, uh, preferable, but also depending on the problem you need, you know, you need to solve all the opportunity, uh, to try and capture, sometimes you don't need all of your data in the right place, in the right format. You know, you, you just need a, a subset of data. So, so yes, it, it, it is, I, I don't want to, you know, kind of pretend it's not fundamentally important, um, but it, it, it doesn't have to be there as a prerequisite for all AI use cases, I think is what I'm saying. And, and the market is now becoming much more knowledgeable about how important data is, um, to more sy systematic use of ai. I, I, I would say.
Jeremy:
Okay. So, so companies don't need to be scared. They don't need to think, oh, I've gotta get everything in order before I can get started. There's pathways if you wanna go the more complex route versus say more out of the box almost.
Adrian:
Yeah, absolutely. Um, that, that's, there's, there's, there's all sorts of different options. Now, if you, um, if you want to say, you know, I want to look at all of my business processes left to right, and I've got, you know, 250 business processes in the, in, in the organization as an example, and I wanna systematically work through all of those and improve them and want to use AI as, as, as part of it, and think about my whole end-to-end business operation as a result, then yeah, of course you need your data in the right place, uh, in the right structure, in the right format, with the right roadmap, um, against it. But, um, let's say, you know, the examples I was using of, um, proposal writing, price comparison, you know, making sure you get quality into, into those documents, contracts, things like that. You, you don't need, uh, all of your data in the right structure warehouse where, you know, you, you, you can work with a subset of data, uh, quite happily to, you know, get real benefit.
Jeremy:
So if I'm a business that's overwhelmed by just how AI can transform my business, and there's so many use cases out there, and you just had to give me three starting points, let's say mm-hmm . What would be a few use cases you might recommend, um, to the average business in terms of how they might find value from ai?
Adrian:
Yeah, I would, I would start with, uh, clients. You know, I'd, I would start with their clients and say, well, okay, what are the things that, um, on the client journeys, they're either finding, uh, most difficult, um, or you are thinking about, um, you know, areas where you, where you, where you can improve and, and provide, uh, better value. So it depends a little bit on the business, right? And it depends on are they consumer facing? Are they B2B, um, you know, how do they engage with, you know, with, with their clients? Um, but typically, you know, you get, um, a lot of client touch points where, you know, let's say they need to interact to buy something, or they need to interact to, you know, raise a complaint, uh, or they need to interact, uh, you know, because they wanna find out more information, uh, about your offering, or they want to, you know, there's some kind of buying signals of what they're searching for on your, um, you know, on, on your, on your websites, various other things like that. So it's those points that I would think about capturing and using AI to provide a better experience to clients and help 'em to get to the right answer faster, um, and feel like they're being engaged. So, you know, if, if you've got a business where, you know, clients wanna get in touch with you, and there's a certain experience around that, and the struggling, then, um, you, you're gonna fall behind pretty quickly. Um, I would suggest, so I, I would start with clients.
Jeremy:
Yeah. So it seems like that common thread is just go back to your business problem, don't get so caught up in the technology, all the changing news every day about models or new applications. Just like, go back to the fundamentals. Yeah,
Adrian:
Yeah. A hundred percent. Don't, don't make it technology focused. Make it business focused. Yep.
Jeremy:
You know, on that, I wanna talk about sort of timeless principles. 'cause it, we hear a lot, oh, so much is changing. It's a whole new world mm-hmm . But I have to believe at both a technology level and a business level, some things hold true, right? Yep. That were lessons from 20, even 50 years ago. And, you know, we've talked a lot about how you have to ground back in the business, ground back in what your customer's problems are. What are a couple things that stand out to you in terms of those timeless principles that even with AI and all this change still remain true?
Adrian:
Yeah, you've got to, um, you've gotta think about things, um, in the right way. Um, for, I, I mean, look, I'll, I'll g give you an example of, um, how we're engaging, uh, a lot of clients on, uh, what we call rapid ai. Um, in, uh, in, in emir, we, we've got a number of customers, uh, looking at this. So it kind of starts with, um, really breaking down what the business is trying to achieve that that's never changed, right? Uh, I mean, I've been in the tech business 30 years that, that's never changed. Then trying to, trying to weed out which use cases are gonna be easy to achieve, um, you know, versus those that are gonna be harder to achieve, that's not changed, right? You, you, you know, effort versus versus output. Then, um, maybe what has changed, um, uh, a little bit along that journey is then what you do with that information.
Adrian:
So we're now moving much, much more to, you know, kind of rapid proof of value, rapid prototyping and making sure that, um, you know, there is, um, demonstration, uh, of that value, uh, early. And I would say the technology is such now that you can produce those things extremely quickly. And so going into, uh, you know, what's, what's changed, what's, what hasn't changed. So what hasn't changed then in all of that is, um, is, is is having the right team, um, you know, you still need, you know, humans in this to, uh, interpret, um, interpret outputs, interpret requirements, um, match the technology to it. Now, maybe the shape of that team has, has changed. So, in, in, you know, in this particular example, uh, you know, we have a lead consultant, by the way, that hasn't changed, that's been around for, um, as long as I've been in the industry.
Adrian:
Uh, but then you've got the concept of four deployed engineers, uh, who are doing this rapid prototyping, you know, um, uh, approach, uh, that has changed, right? You know, they're, they're relatively new roles that have come through. But having the right team who can translate client, you know, needs or outcomes, uh, and marry technology with it, that's, that's not changed then into methodology. Um, so methodology technology enables methodology changes, um, you know, for sure. But having a methodology that you follow, uh, which cycles around, you know, business need, uh, prototype, um, understand the prototype, go deploy into production, understand what we've got, iterate, you know, that whole, you know, kind of agile, iterative way of working, you know, that's been around for, you know, I dunno, 20, 25 years, uh, now and probably beyond, uh, that hasn't changed. So that kind of way of working and that methodology, yeah, tech has brought a different acceleration into that. Uh, but that hasn't changed, uh, either or has changed, uh, quite markedly is the reference architecture. And so, uh, so with AI in particular, you've now got a different reference architecture. You've got a different set of technologies, you've got a different set of possibilities. Um, and going through that, you know, kinda lifecycle, uh, with different tech, you know, that, that, that has changed. But, but fundamentally, the team engagement, uh, the methodology that you take them through, uh, understanding the outcomes and iterating, those things are gonna remain consistent no matter what the tech is.
Jeremy:
I'm glad you mentioned the team. 'cause I wanna talk about the people element of this, right? And especially skills. And when we think about skills and ai, I, people talk about things like prompt engineering, which is a given mm-hmm . But it, it feels like there's some soft skills that still matter now. Yeah. So can you talk about things like, you know, critical reasoning systems, thinking, you mentioned a few of those, but yeah. When you're hiring and you're building your teams now, what are some things you're looking for?
Adrian:
Yeah, so it's that you, I mentioned the role for deployed engineer, uh, for this rapid prototyping. Um, it's a marriage of re really deep technology understanding and understand the art of the possible, of the reference architecture, uh, that, that, that I spoke about marriage with an ability to work with, um, a lead consultant who can help to, you know, decipher a little bit the, you know, the, the business requirements or, or the outcomes, uh, you know, the out the, the outcomes, uh, you know, that are wanted. And, you know, that role is, is definitely changing because then what you can do is work with the tech in real time against, uh, client's business needs to, to very quickly start to produce, uh, outcomes, uh, which are visible, uh, that, that the client can see. So, so you've got this mashup, um, of some of those roles. Whereas typically, um, you know, not many years ago you'd have, uh, you know, a kind of a, a lead consultant. You would have a product manager, you would have an architect, you would have developers, you know, you'd have all these kind of discreet roles that would work together in, in, in a team to produce this. A lot of those roles are, are getting squashed now into this kind of forward deployed engineer, uh, concept. Um, which is, uh, which, which is a very interesting role.
Jeremy:
Hmm. I wanna go back to, you know, the client zero notion. And you're obviously being pitched on a lot of projects, I'm sure, for AI and a lot of different ideas. When you hear a proposal for an AI project, what are some sort of green flags when you're hearing that, that say, oh, this will probably be successful? And what are some other ones where you're like, Hmm, I'm not sure if that's the right way to go?
Adrian:
Yeah, I, I think back to what, you know, what is, what is really the outcome. So, um, another thing that hasn't changed, uh, in business is return on investment. Um, and that is absolutely critical. So, uh, you know, is this, is this really a need that, you know, the business has? Is it really something that people are asking for? Um, is it something that is gonna, you know, provide a fast, um, you know, uh, value creation, uh, solve a problem, or enable us to capture an opportunity? Um, how long will it take to implement? What's the cost? Uh, what's, you know, what's the return, uh, on that investment? So, so I really look at things from, from that perspective. So if it's, uh, a great idea for use of technology, could be amazing, could be interesting, uh, could be like, yeah, wow, that's, uh, that looks fantastic, but in reality, there's not gonna be an ROI. Then those types of things, you know, end up going to the back of the queue, um, are the ones that actually may be more straightforward and less, you know, kind of wow factor may go to the top of the queue just 'cause they'll deliver an outcome faster. So, so that's, that's i's kind of, I, I tend to put a business lens on these things. Yeah.
Jeremy:
Keeps things simpler. You're not getting caught up in the model or the complexity of the architecture that makes it sort of look cool. Yeah. But really, yeah, it goes back to what's the value it's driving.
Adrian:
Exactly. Yeah.
Jeremy:
You know, Adrian, you've got a, uh, oversight of a, a lot of different countries oversee ea and, you know, I'd love to hear your perspective on what's an AI adoption trend that's happening. And I know it's a lot of different countries in ea uh, a lot of different continents, but what's something that especially American leaders might be underestimating or missing that's happening in these other regions?
Adrian:
Yeah, I, I, I mean, I'd say, um, a lot of these things are happening globally, right? So I, I don't see, you know, massive differences, uh, you know, between, uh, you know, at least the desire of a number of, uh, you know, different countries and markets, um, uh, to progress. Maybe there's, there's an ability. So, so it comes down to, I think, uh, access to capital, um, is really, really important and how much, uh, affordability a market has got. So if you're looking in, you know, UAE and you know, Saudi as, as an example, they've, they've truly stated, you know, both countries have stated that, uh, they want a percentage of their economy to be based around AI in the future, and they're investing heavily into it, you know, like really, really heavily. Um, so, you know, for them it's almost a national mission, uh, to go drive it.
Adrian:
And they've got capital, you know, they've got a lot of capital to spend. So, so you're seeing a lot of acceleration there. Um, a lot more, uh, kind of experimentation, a lot longer tolerance on ROIs, uh, as well, uh, which is coming through, which is probably a little bit different. I mean, I'd say in the US market, there's, there's, there's still a huge amount of investment, you know, going on. And probably in ai there's a higher tolerance on, uh, return on investment than there is for, you know, many other things in, in the US economy as well. But I'd say, uh, yeah, there in, in, in Saudi and, and UAE, that that is absolutely the case. I think one of the difference, um, that I do see, particularly in the European market is, uh, kind of the focus on ethics. Um, so there's a lot of concern about, um, you know, harmful AI and making sure that there isn't, um, you know, discrimination Yeah.
Adrian:
Or there isn't, you know, kind of issues with unintended consequences. Um, and, you know, really thinking about, um, how to build that in as, as part of the solutions. And we've already seen some regulation come in, and I think it will just strengthen and strengthen. So, so I think that's, that's one difference that I see. Um, another thing that may not be familiar to, uh, US audiences, you know, 'cause it, it's kind of just been built in, in the us but that is sovereignty. Um, so with obviously a lot of the, you know, big, um, hyperscaler organizations, you know, being US based, it's not something that the us um, you know, thinks about, which is, okay, well, where is my data and what are the restrictions on that data? And, you know, what needs to happen with it? And what if, um, you know, there's, there's, there's, there's an issue.
Adrian:
What does, what does that mean? What's my risk profile? So, uh, in, in a lot of European countries, there's a lot of thought about data sovereignty, uh, and making sure that that's built in as, as part of the solution. And again, you know, working with a lot of us hyperscalers who have got, uh, you know, uh, data centers and facilities in all of the European countries, um, but making sure, but there's also an extra thing about making sure it's not, you know, kind of in the scope of US law about, you know, being able to pull data, you know, back to the US if, if, if required. So, so that's probably an extra thought process as well that I see.
Jeremy:
Do you think that with the data sovereignty issue, that there's gonna be a whole new slew of providers popping up throughout Europe? Yeah,
Adrian:
We're very much seeing that. So we're, we're actually working with, uh, one organization who's based in the, in the DAC region. Um, and, you know, they're, they're building, uh, their own AI data centers, um, very much based on a, on a sovereign theme. Uh, they think there's a, a market for it. They're also, uh, building very sustainable data centers as well, because again, that's probably more of a topic that's, you know, still, uh, you know, very much at the forefront of thinking in Europe, uh, you know, versus the US at the moment. So, so we are, we are starting to see some of these, um, yeah, kind of smaller niche providers crop up as well. Um, and, and yet it, it is driven by the sovereignty angle.
Jeremy:
Y you mentioned that, you know, in, in some ways regulation can actually strengthen innovation. I'm, I'm wondering if you could kind, you know, unpack that a little bit more and talk about how, how you see that happening and what the perception is in Europe right now around things like the EU AI act and how that's actually strengthening Yeah. AI innovation.
Adrian:
Yeah. I mean, the EU AI Act, um, you know, is, is coming into force in, in stages. Um, and I would say that, you know, you know, for, for organizations like ours, and I think for, you know, general business to business AI solutions, um, it, it's a very straightforward thing. It's just how, um, you know, think of it as a set of rules to make sure that, um, AI isn't, isn't harmful, uh, isn't discriminatory, uh, doesn't provide unintended consequences. Um, and then over the next few years, it will just, you know, kind of ramp up and tighten on those, on those rules, um, where it probably causes a problem. Um, uh, well problem, but you know, more of a thought processes around social media. Um, and, and then if you apply the EU AI Act to a lot of social media and say, well, who's accountable for, um, you know, you know, lack of discrimination and making sure there's not any unintended consequences when yeah.
Adrian:
How much is that to do with the social media platform, and how much is it to do with the users of the platform? And where do you draw the line on, you know, what those social media organizations are accountable for and not accountable for? Then you're probably getting into, um, more debatable areas. Uh, but, but I think for, for general, uh, solutions, a lot of these get, get built in now in terms of innovation, what it will drive, I think is better, um, kind of quality checking, uh, on the way through. Um, and you know, already you, you know, we've got, um, ag agentic solutions now where you get quality agents that are built in, you know, checking. So, so these things, because it's rule based, it's actually pretty straightforward for AI to update itself on those rules. You know, get a quality agent and check that, you know, how you're designing, um, either your AI implementation or your AI product, that it's got these things, you know, built into it and it's, and it's safe and, uh, you know, compliant with, uh, with the geographies that it needs to be.
Jeremy:
And you mentioned sustainability as well, uh, earlier. Are there efforts that you're seeing, um, that companies are starting to adopt in Europe that are, um, unique, and what are some of the, the things that are top of mind for leaders right now?
Adrian:
Yeah, I, I think there's probably still a, um, a bit of a fight for the sustainability agenda. So I, I think we wound back, um, you know, let's say pre, uh, you know, pre pandemic, uh, then sustainability was, you know, right up there. And if you asked any European based organization, you know, at least, uh, then sustainability would be one of the top three, you know, kind of business, uh, considerations. Um, you even had, uh, you know, those oil producing companies, uh, countries in the, in the, in the Middle East hosting, you know, coal, um, sustainability forums, things like that. So it, it was right up there in terms of, um, you know, uh, a, a clear business issue, less so now, right. I, I mean, it's still there, uh, and I think there's a little bit of a fight, um, for that agenda, but, but as I say, you are, you are starting to see organizations, uh, you know, come up with offerings that, you know, play to, that build in sustainability as, as, as part of, uh, as part of what they've, um, part of what they're offering.
Adrian:
I do think it's starting to come to the fore again now with AI and, and, uh, power usage. You know, it's been well documented, um, the massive increase in, you know, power needs, some power consumption with, with ai, even to the point where, uh, you know, uh, countries, electricity grids, uh, cannot handle it, um, with many of these, uh, data centers coming online. So there's, you know, things like, you know, nuclear pods, uh, uh, you know, and, and different nuclear technology that, uh, is coming through as well. And I think sustainability is getting, you know, more and more, um, you know, back in the spotlight again. I mean, if, if nothing else, the cost of power alone
Jeremy:
Yeah.
Adrian:
Is driving, you know, if you can be more sustainable and use less power, uh, then, you know, that's gonna, you know, help the environment, but it's also gonna help the business. And I think that's, if you can get the cross hairs of those two, then, then that's very powerful because, you know, it benefits everybody.
Jeremy:
Absolutely. Yeah. A lot of water usage too.
Adrian:
Yes. Yeah. Especially, yeah. Yeah, exactly. Yep.
Jeremy:
Um, I want you to think five years in the future, you know, what is gonna be that differentiator between the companies that, you know, just sort of hobble through this AI era and those that truly harness it to transform their business? What, what's that gonna be?
Adrian:
So, I, I think history is littered with examples, um, where organizations face tough decisions, and ultimately they'll, they will face decisions about, um, what do they do with their existing revenue streams. And, and I think, you know, for, let, let's just say most organizations, uh, they'll face decisions about, you know, you know, should they, do they cannibalize their own revenue, uh, you know, to drive faster growing, you know, sources of, of, of, of revenue and high growth, uh, clients, uh, or do they try and hang on, uh, to what they've got and, you know, kind of EE that out. So, yeah. Famous examples in the past, uh, uh, you know, organizations like Kodak who, you know, invented, uh, digital photography and then, you know, decided not to really go for it. Uh, you know, blockbuster is a, is another very, very famous one who, you know, decided that, you know, rentals were where it's at and, and, and not streaming.
Adrian:
So you're gonna have similar examples. So I, I think the winners are gonna be those organizations that are brave, uh, and, you know, go after their own sources of existing revenue, um, and cannibalize it. And it's gonna be different business models, it's gonna be different charging models, it's gonna be, uh, you know, different business models that, uh, they need to bring to their clients. And they're gonna have to be brave about going after those. The, the ones that will lose, um, are the ones that just don't get there fast enough, and they try and protect their quarterly numbers, um, as much as they can, as long as they can. And ultimately, they'll, they'll end up, you know, being the next code act, the next blockbuster. Um, and I think we'll see a few of those examples.
Jeremy:
So speaking of transformation, I wanna talk a little bit about the IT channel mm-hmm . Which insight's a part of, you know, for decades we've had value added resellers and systems integrators, but it feels like AI has some new requirements for companies to be successful. And I'm curious from where you sit, does the traditional channel model work, what are some of the new requirements, uh, that companies should look for in a partner?
Adrian:
Yeah. I, so if you, if you take the traditional channel, I think the traditional channel's role was to help, uh, customers to, uh, select buy and supply ch uh, technology, right? You know, that, that, that, that was pretty much it. And then there was an opportunity to perhaps add some services around it, you know, uh, but it was mainly things like, um, support, maintenance, uh, you know, break fix those, those types of things. And, um, you know, as we found with, you know, our own transformation to a solution integrator, uh, many, many more customers need help with, uh, designing, uh, you know, building, implementing, and managing that technology as well. Um, so I think that's been a trend for a number of years that, um, you know, there's, there's kind of less value in, you know, the transactional, uh, I wouldn't say zero value. There's, there's, there's still absolute value in it, and I think every organization, uh, needs help to understand the technology, select the right technology, make sure it's the right configuration, you know, make sure, um, you know, they've got roadmaps sorted.
Adrian:
They, they understand life cycles, they understand how it all fits together. So, so I think there's still value in that, but I think what AI is starting to drive is a real focus now on, um, you know, you really need to think about your business, your business outcomes, uh, how to frame the implementation of, of this technology, and how to, how to really get the outcome, uh, from it. Uh, so I think how it's changing, um, you know, the channel is, if your an emotion is to help with, um, you know, access to this technology and, and transaction, then, you know, maybe it's reasonably good times at the moment because everybody's, everybody's building out their AI infrastructure, and obviously there's some very big players doing that. But, you know, I, I think now and onwards, there's gonna be much, much more focus on how to get the most out of this technology, um, and how to deliver, you know, business outcomes for clients. So, so you don't really have that motion, then I think you're just gonna go up and down with, uh, uh, with the technology supply and, you know, you, you're gonna become less valuable to, uh, to clients. So, so for me, that ability to, uh, you know, have, you know, consulting or advisory capability, uh, to have professional services capability, uh, and to have managed services capability is gonna be absolutely vital. Um, you know, to really deliver value to clients, maintain clients, and, and acquire new clients in the channel.
Jeremy:
You know, throughout our conversation so much has come back to understand your business problem, understand your customers as much as, as important as AI literacy is, which I'd argue is, you know, every employee, um, should be thinking about. Is it fair to say that the core just business understanding and business literacy is still either number one or equal with AI literacy? Yeah,
Adrian:
I think so. And you, you, you've gotta have the right mix of team. Yeah. Uh, you know, this, this is not, you know, even though I describe, you know, a forward deployed engineer of being some kind of, uh, you know, tech, uh, superhuman who, you know, you can pull, so, you know, lots of different strands together. Uh, whilst you are seeing that yet the, this, you know, kind of tech consulting or tech advisory role is, is very, very, uh, relevant. And I would say pretty much everybody needs to understand, uh, you know, businesses, business models, but also how then to engage the technology discussion into those business models. And if you think about, look, I've, I've been in this industry 30 years. I think when I started, you still had organizations where technology reported to finance. Um, it was a room in a, in an organization, um, in the building.
Adrian:
It was usually in the basement. Um, and it, you know, it kind of had a bunch of requests to, you know, provide things to the organization, mainly technology. And since then, I've just seen a, um, a steady progression of technology becoming more and more important to every single business. And I think we're at the point now where every single business's strategy relies on technology in some way, shape, or form. I cannot think of one business, um, that doesn't have tech somewhere at its heart, um, as a strategy. Um, you, you know, you can even say very manual jobs. I, I, I make a joke that even my window cleaner, uh, has tech at the heart of his business because, you know, he has tech for bookings. He has, um, you know, tech for, uh, you know, for client satisfaction. He has advertising, you know, he has ratings, he has all sorts of things that, you know, drive his business. You know, I, I mean, it's, it's just every single business. Um, so, you know, that's why it's so important.
Jeremy:
Yeah. Permeates everything. Yeah. All right. Final question. What's one thing you learned recently?
Adrian:
Oh, one thing. Um, so a couple of fun ones. I, I, I learned never to take the, uh, the red eye from New York to London. That'll do it. I, I, I'd, I'd, I'd not done it for a while, but wow. Uh, that, that is a hard flight. Um, I also learned that, uh, when beavers are chewing through trees, they stop to listen, uh, to the tree to figure out which way it's gonna, it's gonna fall. Uh, so there's some great footage, uh,
Jeremy:
Talk about getting feedback on
Adrian:
Yeah, yeah, yeah, yeah, exactly. And, and thinking, you know, make sure they're not gonna get crushed by the tree. That was, uh, quite an interesting, uh, piece of learning, but no, bring it, bringing it back to tech. Uh, there was a recent survey, um, with open AI and Anthropic, uh, who looked at, um, uh, AI usage, um, and actually there's, there's quite a growing swell of, uh, AI usage for, uh, you know, things like, you know, relationship, personal reflection, uh, greetings, um, and general chat. And actually 4% of all use AI usage is for that which are broadly the same as computer programming. Wow. So, so there's a little,
Jeremy:
It's the same as coding. It's, I did
Adrian:
Not realize that's, it's the same usage a according to, you know, open AI and an philanthropic, they did a survey of all their, uh, AI usage. So, so it probably, you know, kind of leads us to think, um, you know, what are the social impacts of, uh, ai and there's a lot of people, you know, using it for, you know, to help with, uh, those aspects of life. And I think that's just gonna grow and grow. Oh,
Jeremy:
Yeah. Fascinating topic. I'll, we can almost do another podcast on that. Yeah,
Adrian:
Absolutely.
Jeremy:
All right, well, we'll leave it there. Adrian, thanks so much for coming on.
Adrian:
Thanks so much, Jeremy. Yeah. Appreciate it. Thank you.
Speaker 3:
Thanks for listening to this episode of Insight on If today's conversation sparked an idea or raised a challenge, you're facing head to insight.com. You'll find the resources, case studies, and real world solutions to help you lead with clarity. If you found this episode to be helpful, be sure to follow insight on, leave a review and share it with colleague. It's how we grow the conversation and help more leaders make better tech decisions. Discover more@insight.com. The views and opinions expressed in this podcast are of those of the hosts and the guests, and do not necessarily reflect on the official policy or position of insight or its affiliates. This content is for informational purposes only, should not be considered as professional or legal advice.
