Practical AI for Business

Patrick Daly:                     Hello, this is Patrick Daley and welcome to Interlinks. Interlinks is a program about connections, international business supply chains and globalization, and the effects on our life, our work, and our travel in recent times. Today on the show, we’ll be talking to Declan Foster, founder of Martel Change Consulting, and Martel Change Consulting is a niche management consulting practice that specializes in providing expert change management transformation and project delivery services. So welcome Declan, and thank you very much for being here with us today.
Declan Foster:                 Thanks, Patrick. It’s great to be here.
Patrick Daly:                     Delighted to have you. Maybe just to kick off and then overview, could you give us a brief synopsis, maybe a whistle-stop to our history of your career to date? What brought you to being the founder of Martel Change Consulting?
Declan Foster:                 Yes. Well, I’ve been working in the project delivery space for about 25 years now. Initially in project management roles, but more in recent years, I had a focus on organizational change management, which is the people side of these technology projects. And the majority of my experience was gained in Australia where I lived for about 18 years. And a few years back, I returned back home here to my hometown of Dublin. I have been ranked in the top 10 top leaders in project management by Thinkers 360. I’m also the co-author of a book called Humology: How To Put Humans Back at the Heart of Technology. I also believe in lifelong learning, and I’ve recently studied behavioral economics and received an honors degree in AI. And I’m also currently studying AI for business over at Oxford University.
Patrick Daly:                     Excellent. So that’s interesting. Technology projects and human beings. So what are the challenges and what are the pitfalls and where do companies maybe overlook the human element in technology projects?
Declan Foster:                 Yeah. Well, they are often overlooked and too often, and we focus on the dollars and the deliverables, and we forget that for any project to be successful, it has to actually be used by individuals and it has to be adopted by individuals, by employees in order to make it successful. And I think that’s really where change management comes in. And it’s handy to have somebody say quite separate from the project manager who’s looking after the cost, the budget, and the scope and the timeline, someone whose focus is just purely on the impacts on your employees and making sure that that’s done correctly. And I think one of the biggest mistakes maybe I find as well is that certainly, and I can relate to that from my project management experience as well, is that you kind of get to the finish line on a project and you maybe tick the box and you’ve met all your objectives as a project manager, but then sometimes the business units, business users still need to adapt to this new technology and this new way of working. And that’s where change management really comes to the fore.
Patrick Daly:                     Yeah. Okay. How would you encapsulate what change management is about? So when you have a project, and often technology projects and SMEs are driven maybe by finance and accounting. In larger companies, there might be an IT department and so on. A lot of those people are very kind of techy in their approach. So what does change management mean in the context of those kind of people, and what needs to happen that maybe doesn’t happen?
Declan Foster:                 Yeah, so I guess as I mentioned, first of all, it’s a focus on the people that are impacted. What are their concerns, what are their issues? What can we do to communicate what we’re doing a little bit better to those employees? And I usually kind of explain it to organizations to say there’s a return on investment from using change management, and it’s usually around these three items. It’s speed of adoption, it’s proficiency, and it’s ultimate utilization. For speed of adoption, you want people to get up to speed as quickly as possible with this new technology that they’re using.
                                           You want them to be as proficient as possible so that they’re using all of the features correctly so you’ve got a productivity improvement. And then you also want to look at ultimate utilization. So unless it’s a system that everybody just has to use and there’s no ifs or buts about it, sometimes you’ll have low ultimate utilization figures because maybe not everybody really wants to use it, or people go back to using a different system or they don’t really use the new system and they go back to using the old way on spreadsheets. So those three factors are really how I sell it to organizations and why change management is important. That’s the speed of adoption, the proficiency, and the ultimate utilization.
Patrick Daly:                     So part of it would be getting the right training, but perhaps also getting the right familiarization and the kind of buy-in, or even the appreciation for the need for the change benefits of it. I find myself, sometimes I get out of sorts because maybe LinkedIn have changed the feature and I say, well, I was happy with the features as it was, and now they’ve changed and now I’m upset. You get a lot of that kind of stuff, right?
Declan Foster:                 Yes. Yeah. Oh, absolutely. So you’ve got to communicate first of all why you’re doing the change. So any change you bring into an organization, whether it’s a large transformation or it’s just an implementation of a smaller back office system, it involves change and you’re disrupting how people go about things. So the most important thing is really to explain the why. Why are we doing this in the first place? And they usually are, most times we’re doing things for rational reasons.
Patrick Daly:                     We would hope there are good reasons. Yeah.
Declan Foster:                 Yeah, exactly. But sometimes we forget to communicate that. So it might be the project team or the sponsor. They’re well aware of why we’re implementing this new accounts payable system, but they forget to remind the employees. They’re asking them to do something different, to disrupt how they normally use the accounts payable process, but they have to remind them, what are the reasons for doing it, what’s the benefits to the organization, what’s the benefits to the individual? You have to communicate that.
Patrick Daly:                     Yeah. So in your current work at Martel, what are the main kind of services you’re providing to clients currently?
Declan Foster:                 Yeah, it’s focused on change management at the moment. As I discussed, I mentioned earlier, that’s really my focus on my career lately, although I still help some organizations out in project management roles. So we’re looking at change management primarily for technology projects. And I’m also, some of the work that I’m starting to do at the moment, as I mentioned, I’ve been a keen fan of AI and I’ve been studying and researching that for the last few years. Some of my emerging work at the moment is helping to demystify AI, particularly for leaders and C-suite executives
Patrick Daly:                     So those change management projects, is this maybe where a company is, say, is going through an ERP implementation and they need some independent help with change management? Is that a kind of a scenario?
Declan Foster:                 That’s a scenario, yes. So they’ve got a technology project or it could be a transformation project or a new ways of working, and they need some external help or some expertise to come in and guide them through that process and also upscale their existing employees and managers and how to handle change.
Patrick Daly:                     Okay. And are you generally then, are you working with or on behalf of the technology company or with and on behalf of the ultimate client or both?
Declan Foster:                 Usually with the client, usually with the business itself to help them and work alongside their vendor.
Patrick Daly:                     Okay. And how would you say clients are better off after having worked with you as opposed to if they hadn’t?
Declan Foster:                 Well, I think, first of all, the focus again is to making sure that the change is adopted by the employees and therefore ultimately that whatever change that brings to the organization, that that’s actually sustained within the organization. And I also aim to always develop and upskill the people that I work alongside within my team, and also to leave them with an organizational level, leave them with resources and toolkits that they can use in the future. So toolkits around practical things like stakeholder registers or change impact registers, surveys that they can use for, templates that they can use for further changes. So it’s about making sure that you’re passing on those skills to the people that you work with in the organization rather than making them sort of reliant on you.
Patrick Daly:                     Yeah. So they end up, I guess their business benefits then is they acquire a new capability, they adopt and can enjoy the benefits of the return on investment faster. So there’s speed, there’s [inaudible 00:10:18], there’s capability, and then there’s the skills transfer, I think, which you pointed. Would that be fair?
Declan Foster:                 Yes. Yeah, exactly. Yes.
Patrick Daly:                     And you and I met first as speakers at an event. We were both speaking on different topics, and you just mentioned it there. You were talking about AI, and as we know, there’s a lot of hype and not a little hysteria about the topic at the moment. So really, where are we at in your opinion, with artificial intelligence and the large language models like ChatGPT that I guess some of our listeners will have been playing around with versus AGI, which we don’t have yet, but people are maybe a little either frightened of or enthused by. So where are we on that journey, do you think?
Declan Foster:                 Yeah, well, I think that the advances that we’ve seen in large language models like ChatGPT, and there are several others, are really are a fantastic achievement. And I think they’re kind of a watershed moment for AI in two ways. First of all, it’s raised awareness over the past year or so of the potential of AI among the public, and it’s also democratized AI. It’s put AI into the hands of the masses. Prior to ChatGPT being released, these large language models were only really available to the technical folk, the programmers and developers via an API and application programming interface. So it’s easy to underestimate how significant an achievement things like ChatGPT are. I read somewhere where someone said it’s a bit like we found a talking dog, but now we’re critical about its grammar. So it’s a fantastic achievement. But on the flip side of that, there are issues with using large language models.
                                           For example, hallucinations. It can make things up. After all the large language model is just trying to predict the next word or the next paragraph or the next page. So sometimes it makes things up just because it wants to fill in a gap. So we have to be careful that ChatGPT doesn’t fact check itself really, at least yet so far. And then we have copyright issues and we also have bias and fairness issues with these machine learning models. And I’m also, I’m not necessarily convinced that the progress that we’re making with large language models will ultimately lead to AGI or artificial general intelligence. Although I think a while back, Elon Musk said that he expects we’ll reach AGI by the end of this decade. I ran a survey recently on LinkedIn asking people when they thought we would reach AGI or artificial general intelligence. And it’s interesting that the majority voted for before 2030, which really surprised me, but the next biggest boat was probably never. So you get sort of opposing opinions on it.
Patrick Daly:                     Yeah, it may be never. Who knows? Who knows?
Declan Foster:                 Exactly. And that’s a valid opinion. It absolutely is. And it might be that, as I said, we’re kind of almost going all in on large language models will lead to AGI, but there are experts out there that say, “That’s not the way to go.” So who knows if we’ll ever get there. The brain is such an amazing piece of machinery, if I can put it like that, that we don’t fully understand at all.
Patrick Daly:                     Yeah. Well, I mean the whole concept of what consciousness is, it’s not really definable, is it?
Declan Foster:                 That’s it. Absolutely. Yes.
Patrick Daly:                     So coming back maybe feet on the ground, what can we do right now with this technology? What can a business do right now with ChatGPT, for example?
Declan Foster:                 So there’s lots of things that you can do and it really depends on your role. So for me, applications and use cases of ChatGPT include brainstorming ideas. So you can ask it to give you 10 ideas for a blog or for an article or even for a workshop that you want to run. I think it’s very good for analyzing and summarizing text. I’ve used it to analyze and summarize survey results, for example. And I think a lot of folk, I’ve used it for this, and I think it’s a really good use case for most people. We can use it to compose texts for sales campaigns and for customer emails. So you might ask it to, say, improve my sales call email by making it appeal to the emotions of a particular audience. And then you can give it details about your product and details about your audience. And it would generally do quite a good job of being creative in terms of the text that it suggests. Because again, it’s great for that brainstorming, but where it falls over sometimes is in facts, but where it’s creativity and language, it’s fantastic for.
Patrick Daly:                     Yeah, I, for example, take some notes at a meeting and put the notes into it and tell it to do a situation appraisal, and it will do that. And then you can say, or I’ve presented it with the transcript of an interview where I’m interviewing somebody like you and we have a transcript and turn this transcript into an article for business people.
                                           So it’s taking our content and creating something else. Or I might write an article and then tell it to … No, sorry, I might write a script for a video, do the video and then tell it to write an article off the script that I’ve written. So it’s getting … Sometimes people say, “Well, that’s not your work.” When you say, if you just asked it to produce something that’s not your work, it just produces it. But if you give it content and ask it to do something else with it, to repurpose it, then the question arises, is this your work or is this not your work? It’s kind of a gray area, isn’t it?
Declan Foster:                 It is a gray area. And it’s also, I think that the quality, there’s a vast difference in the quality when you give it substantial information, for example, your script for your video, and then ask it to write an article. So that goes into this whole idea of prompt engineering, and if you give it enough information and proper instructions, you get quality output. Whereas if you ask it something really vague and generic, like write me an article for X, Y, Z, what comes out will be pretty bland. It wouldn’t be awful, but it’ll be pretty bland. So the better the input, the better the output.
Patrick Daly:                     So it’s almost like to get the best use out of, it’s not just to be kind of lazy and ask it to do stuff, it’s to use it as an enhancement of your own abilities. Isn’t that it?
Declan Foster:                 Yeah, absolutely. I mean, I’m all for the idea of AI augmenting humans rather than replacing them. I think that’s where we get the most value out of it. Maybe let’s say in 2050 we might be all replace, but for the moment it’s there to augment us, I think.
Patrick Daly:                     Yeah. So another thing that was going through my head is that I’ve heard some companies using it maybe to integrate into their customer service. How can you integrate this thing into your customer service and remain secure? How can you do that?
Declan Foster:                 Yeah, you’ve got to be really careful what information that you share and how you share it, certainly via the standard web interface and ChatGPT. So there’s a couple of things. I know that Open AI, the company behind ChatGPT, they’ve now released ChatGPT Enterprise Edition, specifically for enterprises or organizations, and that introduces additional controls. Then you’ve also got options, which I’m actually investigating at the moment where you can access ChatGPT and the underlying models from within cloud solutions like Microsoft Azure. So if you’ve got your data in a secure or hosted solution like Microsoft Azure, and there are others, and there are then ways with Microsoft to access Open AI and ChatGPT, but staying within that secure environment. And I think that’s the way organizations are moving to. I think in the early days of using ChatGPT, there’s certainly some famous cases of people submitting confidential information, company information to it. And I think the famous case was Samsung. I think two employees were submitting confidential information as part of a prompt or a part of a question, which they got slapped on the wrist for that, obviously.
Patrick Daly:                     What it has done as well, I was told this by actually a CEO of a Silicon Valley decision intelligence software company, actually interviewed him a few weeks ago. So he’s in the archives there, but he said that the advent of ChatGPT had kind changed a clip or changed a connection in people’s minds about what this stuff can do. So it has opened doors for him with regard to presenting his solutions and technology. Because some of the things that his solutions did, people had doubts that software could do that, and now that people can see what software can do through ChatGPT, they’re not questioning him anymore with regard to his claims about what his software can do. So it’s kind of opened people’s minds as to the capability of the talking dog is here. Right?
Declan Foster:                 Yeah. Yeah, that’s really interesting. Yes. Yeah.
Patrick Daly:                     So we might do, maybe as we come into the last few minutes, we might just change tack a little, then maybe I’ll ask you a few questions about your own attitudes and life experience and see what makes Declan Foster. So in terms of life lessons that you’ve learned along the way, what would you say is the most important life lesson that you’ve learned that stood to you throughout your career?
Declan Foster:                 I had probably say that resilience is the most important thing in life, and being able to build that resilience muscle is a really important thing to do. I mean, we all get knocked down in life and we all have setbacks, I think, but what makes the difference is how you respond. And it took me a while, I think, to learn that. And almost sometimes I think in our early years we expect things out of life when we expect everything to go smoothly for us. But I think I’ve learned then over the years that that resilience is the most important thing. And if you do get knocked back down, just dust yourself off and get back up again.
Patrick Daly:                     Yeah. Yeah. It’s interesting. Sometimes when things go wrong, we kind of get frustrated or whatnot or resent that it’s happened, but I’ve been struck watching the rugby World Cup, and obviously sport is not life, but it kind of sometimes is a distillation of certain lessons. And when they’re saying to Andy Farrell sometimes, like you guys were under the [inaudible 00:22:19] there for a long time, under pressure, we’re really lucky to survive that. And he answers, he’s done it a few times, he says, “That’s where we want to be.” So he’s almost happy in the adversity because the adversity he knows is making them more resilient. And I think some of the ways that the team stood up when they were in desperation on the line, but yet the other team didn’t score, and he’s saying, “That’s exactly where we want to be.” So as you said, training that resilience muscle, it’s going to take a different attitude to adversity, isn’t it?
Declan Foster:                 It is, yes. Yes. And then you bring up the point of also teamwork as well, and not being afraid to rely on those around us and ask for help, I guess, being part of a team, whatever team that might be, because you might be a solo entrepreneur or whatever that might be, there are always someone that you can ask for help or you can ask an opinion of.
Patrick Daly:                     And has the experience of the last two to three years, so we’ve lived through this COVID experience that kind of seems distant now, but it was so real for so long. The experience of COVID pandemic, has it changed, has it refined, has it reinforced any of your own personal views or beliefs about work and about life and about business?
Declan Foster:                 Yeah. Well, I think personally, I think we’ll never go back to a nine to five Monday to Friday sitting at a desk in an office, maybe in a city center somewhere. I think for me, the pandemic experience and the lockdowns has proved that it works. And I know some people say working from home does not make us more effective and efficient. But then I sometimes ask, well, the key question is, does it make us less efficient? And I think the answer is no, because then it certainly improves our work-life balance. I know it has for me, and it’s surely a positive thing for the environment with less traffic on the roads, less people commuting every day.
Patrick Daly:                     Yeah, yeah. I was wondering, well, a few kind of questions when COVID started was how the hell are we going to run the business at all? And then we realized we could run it online, and then we had to do it 100%. And then when things started to open up, you say, will it go back to the way it was? Will it go back partially? And I think it’s kind of settling now, maybe not 100%, but it’s settling now, and I’m finding at least personally, huge increase in efficiency because of a reduction in travel. So I’d say maybe eight out of 10 face-to-face meetings that I would’ve had pre-COVID, eight out of 10 of them are now online, and two out of 10 are face-to-face. And I’m finding with clients, so people who do have to go into work to an office some of the days, they also are not traveling as much because the meetings they have with their colleagues in the sister plant or with the colleagues in the plant in Cork or whatever, they’re not traveling to them now either, even though they might be back in the office.
Declan Foster:                 And I think what I’ve found as well is that it actually makes the times when we do get together physically in person more important when we’re face-to-face. So yes, the majority, absolutely the majority of my meetings are now virtual, but when we get together with a team or teams for brainstorming session or some type of workshop, we actually kind of appreciate that so much more, and that time sort of face-to-face. So who knew? Who could imagine that the day when we’d look forward to a workshop as a treat almost.
Patrick Daly:                     So outside of work now, what kind of things do you like to do when you’re not thinking about AI or change management or project management? What do you get up to?
Declan Foster:                 Yeah, I like to exercise, and particularly at the moment I’m getting back into running at the moment, but we’re talking about doing park runs rather than marathons, so enjoy doing the park runs on the Saturday mornings, gets you up out of bed and gets the exercise in first thing in the morning, so that’s good.
Patrick Daly:                     Yeah. That’s good. I do them as well. Although I sprained my knee, I got a ligament sprain in my knee, so I had to leave off for a couple of months. So just getting back now. Yeah.
Declan Foster:                 Yeah, yeah. You miss it when you can’t do it, don’t you?
Patrick Daly:                     Yeah, yeah, yeah. It’s frustrating.
Declan Foster:                 [inaudible 00:26:56].
Patrick Daly:                     Are you reading anything or listening to any interesting podcasts that are kind inspiring that you’d recommend?
Declan Foster:                 I’ve just finished reading a fantastic book called Ethical Machines from Reid Blackman, and it’s all about the importance of having an ethical approach to artificial intelligence, [inaudible 00:27:18] touched on earlier, and I can really highly recommend that. It’s a really fantastic book.
Patrick Daly:                     Who’s the author? Do you remember the author name?
Declan Foster:                 Yeah, sorry. It’s Reid Blackman, R-E-I-D, and Blackman
Patrick Daly:                     Reid Blackman.
Declan Foster:                 He’s a US commentator. You often see him on CNN and other places. He talks about artificial intelligence and also about specifically around the ethics around artificial intelligence. I particularly like the way there’s a lot of talk about the different principles and values that you need to have for when you’re implementing AI systems, but he broke it down into the three main values or principles when he talks about privacy, transparency, and bias. So you focus on those three things and you get them right for any AI project or any AI initiative, then you’re probably off to a good start. So I highly recommend that book.
Patrick Daly:                     Excellent. Excellent. Well, many thanks Declan. Time flies on here, as you will have realized. So thanks very much for being here with us today. It’s been an absolute pleasure to chat with you.
Declan Foster:                 Thanks, Patrick. Great chatting to you today.
Patrick Daly:                     And thanks also to our listeners for tuning in again today and be aware that if you enjoy this episode, you can find a full series of over 130 episodes of Interlinks on Spotify, Apple Podcasts, Acast, and other major podcast platforms. So until next time, keep well and stay safe.