Silicon vs Carbon: Why Your Next Colleague Might Be an Algorithm

The technological revolution is here, and it's powered by artificial intelligence. In this riveting conversation with AI expert Kieran Gilmourie, we explore the urgent need for businesses to embrace AI technologies before competitors leave them behind.

TLDR:

• AI is being rapidly adopted across all business sectors, company sizes, and government departments worldwide
• European companies need strong AI policies to compete globally and maintain knowledge economy advantages
• When using public AI models like ChatGPT, protect sensitive data by turning off training features and avoiding sharing confidential information
• No profession is entirely immune to AI disruption - from therapy to art, law to healthcare
• Technology should handle automatable tasks, freeing humans to focus on areas where they uniquely add value
• Continuous learning and adaptability are essential skills for thriving in an AI-powered workplace
• Rather than diminishing human connection, AI can enhance it by removing mundane tasks and expanding creative possibilities

"For the love of goodness, can every Irish business please grasp this as quick as they can?" Kieran urges, highlighting that companies across every sector and level are rapidly investing in AI capabilities. This isn't a wait-and-see scenario – organisations implementing AI are experiencing transformative improvements in productivity, creativity and competitive positioning.

We take a global perspective, examining how different regions approach AI adoption and regulation. While the US leads in investment and China follows closely, smaller nations like Saudi Arabia are making remarkable strides by integrating AI across government services and business operations. Europe's regulatory approach aims to balance innovation with necessary guardrails – what Kieran describes as maintaining "infinite pace and infinite patience" simultaneously.

The conversation turns practical with essential tips for using public AI models safely. From turning off training features to exercising discipline with sensitive information, Kieran provides actionable guidance for protecting intellectual property while still leveraging powerful tools like ChatGPT. He explains why effective prompting techniques dramatically improve AI outputs, comparing these systems to brilliant interns who need clear direction to deliver their best work.

Perhaps most thought-provoking is our exploration of AI's impact on traditional roles. Kieran challenges the notion that certain professions are immune to disruption, suggesting that continuous adaptation is the only sustainable strategy. "Concentrate on the bits that are uniquely you," he advises, urging listeners to embrace lifelong learning while allowing technology to handle automatable tasks.

Far from diminishing human creativity or connection, Kieran believes AI enhances our humanity by removing the mundane: "Let the technology take the robot out of us so that we can be the most human we have been in the last 50 or 70 years."

Ready to transform how your organisation leverages AI? Subscribe today and join the conversation about leadership in the age of artificial intelligence.

Chapters

• 0:00 AI's Impact on Business Today
• 3:36 Global AI Investment Race
• 8:24 Privacy When Using ChatGPT
• 12:30 Effective Prompting Techniques
• 18:55 Will AI Replace Your Job?
• 28:49 AI and Human Connection

Transcript

Speaker 1: 0:00

Hello everyone and welcome to another edition of the IMI Talking Leadership Podcast. Today, my guest is Kieran Gilmourie. Kieran is known for his ability to provide actionable strategies that help organizations adapt and thrive in this digital world, and I must say we had a fantastic virtual event with Kieran a couple of weeks ago now for our IMI corporate members. We had about 650 people signing up to attend that event, which is probably the biggest audience we've ever had at one of our virtual mini masterclasses. So today I want to go a little bit deeper into some of the topics that we covered on the mini masterclass, and those are all about AI, and in fact, we have a whole lot of questions left over from that event that we didn't get to, so we'll run through those a little bit later. So, ciarán, to get us started, please introduce yourself through those a little bit later.

Speaker 2: 0:46

So, ciarán, to get us started, please introduce yourself. Oh gosh, well, look, I've been in business for over 30 years. One of the most exciting things I've seen for decades and was in the space a decade ago and came back into it is artificial intelligence Boy. Is this the place to be? So let's jump in, answer some questions and try and follow up the last 650 questions. I think we got with some more answers to try and satisfy everyone. Farah, let's see how we do.

Speaker 1: 1:09

Brilliant. Thanks, ciarán. So, just to set the scene a little bit, during our mini masterclass we had people from all different types of sectors, organizations, areas of the business. I think we covered every level, from people right at the start of their careers all the way up to CEOs. So, ciarán, can you give us a quick overview of some of the key AI trends at the moment and who you think should be using AI? Is it everyone?

Speaker 2: 1:36

Yes, it's the shortest answer I can humanly give. Look, I travel extensively. In the last week alone, I've been to Amsterdam, I've been to Boston, I've dived into phone calls in Dublin, I've been down to Cork and I've been in the Middle East. Believe it or not, all in seven days. There isn't a company, there isn't a level, there isn't a sector, there isn't a government who is not rapidly investing in AI, generative AI and agentic AI technology. Therefore, for the love of goodness, can every Irish business please grasp this as quick as they can? There is a sense of urgency, or there needs to be a sense of urgency around this.

Speaker 2: 2:13

It is not like the old days in that, oh, I think my competitors are doing this, therefore I better. And now everybody worries that the competitors aren't doing something. Trust me, your competitors are People you haven't even come across who will be. Your competitors are jumping in with ai and generative ai because the technology is so powerful, and that's everything from getting agentic labor to actually undertake the work and do it to simply having chat gbt on top of your desktop, allowing you to have an army of what I describe as really well-informed PhDs doing everything from research of your clients, research of your competitors, product research to selling, to answering questions day or night, to helping you design product strategies, portfolios of whatever. This is embedded in everything. It's only going to get more embedded and therefore, again, let me please scare, excite and delight you by saying jump in with both feet.

Speaker 1: 3:07

Thanks very much, ciarán. We're going to get into some of the questions that were left over from the event, and one person asked who wins the AI race USA or China? But I guess it's a little bit deeper than that, because I remember hearing a couple of years ago, when chat, gpt and generative AI first became so popular in the mainstream, that Europe was quite regulated when it came to AI and the US not so much. So is that still the case, or which country do you think is really grabbing the bull by the horns and going for it with AI?

Speaker 2: 3:36

Yeah, it's a simple question and a difficult one to answer at the same time. If you take pure spend on what's publicly declared as investment in ai, ai companies, data centers and everything in the us is leading china, as much as we hear the noise isn't even in the game. When it comes to pure billions or should I say trillions spent, remember faro we used to think a billion was a big number. Now we have to quote in the tens of trillions. Goodness knows what's next. But if you want a country that's actually investing in ai and every single component of it, from government to business and every single business, then you would not go too far wrong. If you looked at saudi arabia, where actually I spend quite a bit of time, the government heavily investing the society and generally heavily investing businesses are so the actual use of ai technology and it being embedded in everything from traffic to servicing you in a bank, to whatever. That's a country to watch out for. Now the eu has legislation. I actually quite like it, if truth be told, because you've a couple of methods of looking at this. Someone once described it as gloomers. In other words, you know ai is going to destroy the world. It won't, and you know, and therefore we need tons of regulation, and the corollary of tons of regulation means lack of innovation, but again, to a degree in Europe we're seeing that. I don't think it's a lack of innovation, I just think that the amount of funding that's available to do really interesting stuff is still concentrated in Silicon Valley. Let's see if our friend Trump has encouraged many a data scientist and many a university professor to come across to Europe. That may change, but again, I think it's down to that. I can't name too many European companies that have really excelled pre or post legislation, so I think it's down to funding the US has, you know, if we call it the boomers or bloomers which is absolute government regulation that wants business to flourish.

Speaker 2: 5:28

You have to ask where's the actual balance, is it at all costs? And the answer is it can't be, because, like everything, from a car to an airplane AI comes with risks. Therefore, we have to be careful that we don't run too far ahead to create problems that we only think about afterwards and then try and shut the door of the stable once the horse is bolted. I'm actually somewhere in between the two, which is I want us to run ahead with caution. So someone said infinite pace and infinite patience, both of those at the same time. You need to to marry. I think we can do it, you know. So I would like, uh, globally, everyone to do really well, because, again, you're running the instance. If if I try not stay away from politics at the best of times If we're looking at geopolitics, this technology so powerful in the commercial, military and so many other ways, we need Europe, in particular, to have its own large language models.

Speaker 2: 6:20

We need it to have its own very vigorous AI and generative AI and agentic AI space, and what we really need is Ireland to have an amazing AI policy, ai framework and investment in AI to allow us, as a knowledge economy, to really keep accelerating ahead and more than earning our place in Europe, never mind globally, when it comes to delivering really great business environment that's so much more productive, so much more innovative than any other competition. So viva Europe, viva the US, but viva viva Ireland.

Speaker 1: 6:55

Thanks so much, kieran. The next question is about ChatGPT. So I think the majority of us, especially in my profession, which is marketing, we all use Chat gpt to some extent. So whether that is just helping write some copy for social media or kind of an email, any kind of marketing collateral, a lot of us use chat gpt. A lot of people would use chat gpt from a sales perspective and operations perspective, hr, whatever it might be. But you mentioned that you turn off certain functions when you use chat gpt. For example, don't use this for training. Are there any other good ip or work privacy routines or tips that you'd have for people when they're using the kind of public model of chat gpt?

Speaker 2: 7:36

yeah. So just to be clear for everyone, in case you're not aware of all the public models we got chat gpt, we got gemini, we got bing bard, whatever it's called now. There's so many. Someone gave me a count about six months ago and said there's about 1900 variations, but a bit like now we don't search, we google. Chat gpt earned its stripes by coming out there, at least publicly, first, so that's the one we know about.

Speaker 2: 7:59

There should be no surprises to a degree in this. Ignore the tool for a moment, because we've had legislation for years. We need to make sure things are protected and you have a variety of versions of chat gpt, just like you have a variety of versions of co-pilot inside of your business. The most secure is usually co-pilot inside of your business or something locked down that it means that any data you're putting out into the wild or into the cloud or into the ether doesn't get seen by people you don't want it to get seen by now let's work off, for example, public chat gpt and there's about three or four variations of that the free, the plus and the premium version of the pro. One goes at twenty dollars a month, one goes at 200 and one is free but limited use. I would encourage anyone anyway always to adopt a cautious approach when it comes to not just you know, know, chat GBT, but email. If I don't want it appearing in a particular newspaper then I don't want to put it out in the wild and that's just good sense. And in addition, there's regulation and rules GDPR, the EU Act and there's usually industry regulations as well. So, for example, the Central Bank of Ireland will have particular requirements of companies not to do silly stuff with their client insurance, banking or whatever details.

Speaker 2: 9:07

But going off public look, without, without going too far, the shortest answer is ask chat gbt. Oddly enough, it will tell you. But as we're asking me, I would say look, turn off chat history and training. So you go to settings, data controls, switch off chat history and training and this means that your prompts aren't saved or used for public model training. Discipline is one that's just normal, natural, human discipline. So don't put in confidential information, your customer names, financial details, sensitive sensitive ip or job applicants. Or, as I discovered, one northern ireland government agency was doing letters to ministers or disciplinary letters, not the thing you would put into email.

Speaker 2: 9:48

Don't put it into chat GBT. You can, if you want anonymize information, so replace names, companies, project details with generic placeholders or topics like company A and project X. Perform regular platform hygiene, so clean out conversations by manually deleting anything that's sensitive from your conversation history. You can use something called incognito mode so you go into browser incognito private to reduce anything being stored locally. You can use one or two accounts, so you could have one account for work, one account for private use and then adopt strict privacy settings.

Speaker 2: 10:25

The only thing with that far is you tend to forget. You tend to forget that you're using one or both, or co-pilot and work and chat gbt. So do be careful with that one, unless you're really really conscious about what you're doing, be the was ever as educating the business or the teams or yourself. You know, put guidelines in place for chat, tpt, set explicit boundaries for sensitive information handling. Teach people the risks regularly, though, the risks change because companies like open ai constantly change the policy, so don't do it and forget.

Speaker 2: 10:56

You know, check back in at policy features, policy updates and everything else, and then the usual of the ones is like if, if something is really, really sensitive you might want to consider, you know, co-pilot in a secure tenancy or a model that's actually sitting in your own data center, or you know, strong data security, vpns on public networks, dual factor authentication a whole host of things all to make sure that your data doesn't leak. Now I don't want to frighten anyone. You think of the basic scenario. I can send any old email anywhere pretty much without thinking too much. I can do lots of things on these tools that allow me to be very productive and very creative, and I don't need to worry too much about settings. But if it's your financial details, your partner's details, client or whatever, like any other tool or technology tool, exercise the air of caution, because you would want your data handled discreetly and therefore handle other people's data discreetly as well.

Speaker 1: 11:53

Thanks, kieran. That all makes sense. I guess it really is just about staying conscious of the kind of data that you share and that you put into these models, especially if you wouldn't share it via other platforms like email. That makes a lot of sense. Chat gpt the majority of people that I know either use the free version or the kind of $20 one, but I know you mentioned using the $200 one, so does that improve the quality of the responses you get, or is it more about kind of unlimited amounts of data that you can get out of chat gpt?

Speaker 2: 12:29

I kind of mixture of both. You know, when you pay, you get earlier access to the better models, and some of the models these days, like zero, three mini as a research tool, are just extraordinary. You know, I, I, I just can never imagine barra going back to not having it. That that's, and it's an extra limb that I can't believe that I lived without, and it's also a case of usage. So therefore, in order to entice you, dare I say it, and encourage you to, to use the models, you get a certain amount of what you might describe as credits or units going against searches and then, if it's ChatGPT, it switches it off, if it's Claude, it stops you working for so many hours because behind the scenes, this is obviously costing the manufacturers a lot of money. Now, this isn't to say everyone needs the $200 a month one. I can't imagine living without it, to be honest, it's not to say the free one doesn't work. It's not to say the $20 one doesn't work, or Claude, or perplexity as a research tool or isn't your thing. Look, go in, start what I described as light. You know, use one of the free available models, practice with it, learn how to use it either. You know, for example, I might do courses there's lots on this summer or if you're in sales, marketing, finance, a small business owner or whatever else, you can come along and in one day you go from zero to hero. Jump onto YouTube, you know. Whatever it is, whatever is your, your desired learning, then you can quickly learn and as you quickly learn, if you find the tools more useful and more beneficial, then you pay your money.

Speaker 2: 13:55

For me, though, if I go back to mine, two hundred dollars a month works out at about what's that? About five, ten dollars a day. It is for free. When I look at my productivity and efficiency based on and I was on the earlier model as well you know I'm more creative, more innovative, more productive. I would need far more people in my team if I were not able to do all of the things that I need to do at gaps in my presentations. Adopt a contrarian view. Help me practice sales pitches or whatever else it is, or on stage presentations, all for the cost of $10 a day, and if I can't come up with something that you know is worth more than $10 a day, I shouldn't be using the tool. And, trust me, I do so again, pick your flavor and as you get better, find it more useful. Then spend more money. Don't dive into the $200 and wonder why it's not working for you at this stage. You're not getting the outcomes that you deserve based on the financial input you're. You're popping in there.

Speaker 1: 15:10

I think another way that you'll get the outcomes you're looking for is being able to prompt these AI tools really well. So that is something that you spoke about on the masterclass and there was a lot of interest in the chat in terms of how to build a prompt library, how to prompt ChatGPT or any of the other tools correctly. So how can people go about that? Because I do think most people are quite guilty of just throwing things into ChatGPT. You know the way you might throw it into Google, and that's probably not always the best way to get the results.

Speaker 2: 15:39

It's not the best way to get the results, not just probably it isn't Google, yeah, look, the way to get a better outcome out is to give it a better instruction on the way in. So let's imagine that Gemini, ChatGPT, co-pilot, whatever it is is a really really, really, really great intern, a really really great colleague, a PhD student who's just arrived inside of your business. Then what they need is clear direction. So write me an interview template. And then saying do that. And then, looking at the output, when they come back and have created an interview template for an artist, not the software engineer at level two that you wanted inside of your cybersecurity company, that's graciously on you, not on them. Whereas saying, look, you know, act as a CHRO. I want you to undertake the following task. You know and I want the following output looking like the following thing even something that's a very small framework to prompt and that's called you know, role action, outcome it, you get far better answers. Now again, not to try and sell the course, because I genuinely can't explain in five minutes you know all of the different models, all of the different ways to prompt this, all of the exercises that you can do, but it's almost like magic. Once you suddenly know how to use these tools, it's just boom. The things that you can get out are unreal.

Speaker 2: 17:02

I said I've been doing or using business technology for 30 years. I've never been genuinely as excited as I said on the panel or when we did the mini masterclass. Was I going to buy an eight pound headset that showed me a cartoon half body jumping around in the whatever. It was the metaverse, not a mission. It was never going to work. I was a chief AI officer 10 years ago.

Speaker 2: 17:27

The technology then was extraordinary. Once you used AI and great decision insight based around AI and predictive and prescriptive analytics. All of a sudden, your capability, your business decision making and everything else is just huge. Now the technology is practically free. You know it.

Speaker 2: 17:44

Yes, twenty dollars is not free, but intelligence has cost nothing. For you to actually spend time google researching, for you to hire a researcher, for you to hire someone who's going to create a marketing strategy, for you for you to hire someone who is going to design a marketing strategy, for you for you to hire someone who is going to design, you know, a HR policy or whatever else you're talking about tens of thousands to get good quality information out Be that a contractor or a fractional worker or hiring someone over a 12 month period as good as, if not better, outputs by prompting a $20 a month tool than you can out of. You know a huge swathe of the educated population and therefore education is or the outcomes or insight or knowledge is for free. When it's only $20 a month, that's nothing. I spent more in coffee in a week than that and I drink a lot of coffee.

Speaker 1: 18:36

I think it's slightly terrifying to think that you can get outcomes that are just as good with this $20 chat GPT model as you could by hiring a professional who's been through years and years of whether it's training or work experience. So let's flip things around a little bit. Are there any professions that won't be impacted by?

Speaker 2: 18:57

AI, I'm going to say no, but let me nuance that answer. So, classically, you will see a lot of literature on this and a lot of posts where they talk about roles resistant to AI replacements. We need to be cautious about the degree of impact that AI is going to have. So, for example, in this day and age there's a lot less saddlers than there are software engineers, you know and therefore you know job done there. I love saddlers, by the way, a horse rode for years. It's not to diss them, but the need just isn't there. So, classically, people will say well, look, what about a therapist or a coun? Now, a really great therapist or counselor? Absolutely that. That person who is huge emotional intelligence, who's massively well qualified, who treats you as a unique individual, who hopefully is non-biased when you're their patient fantastic. But let's take a great irish business or irish owned business, or the the ceo is robot. That's a therapy tool that I used for a month, because I won't even mention something that I don't think works. I will practice it. My reputation is too strong and I don't want to impact people as a therapeutic, but that gets me a huge quantity of the way and, yes, maybe I then need to go to a professional, but the number of therapists and counsellors that actually exist are far fewer than the actual need. So that has removed a whole. That has created an answer to a massive gap.

Speaker 2: 20:27

And if you look at, there's a recent report out from Harvard Business Review that said what are the top 100 uses of generative AI at the moment? Some of the top are actual therapists and counsellors. Now you have to be careful that you don't. You know, the classic word is transmogrification, where I now imbue ai with human trace. This is this is not carbon, this is silicon. So it's not a real person. Please don't get confused. But what it's doing is, uh, presenting a service that is emotionally intelligent, simulant and that sounds terrible. It's actually very good and but it does the job of a therapist and counselor. Now let's say social worker. That's a difficult one at the moment but, equally speaking, I start to see social workers at the moment bringing out AI that's got insight into patients records. I see them bringing along, you know, ai that's recording the notes and recording the actions and recording the transcripts of the meeting. Never mind, they don't get sued, but those notes automatically make their way up into, you know, the ether. I see online counselling. I see online doctors All of these things are happening at the same time. I see online vets and you imagine, as you're going on to the bed, the website, the whatever, the whatever, the virtual worker, you're answering a load of questions in the background.

Speaker 2: 21:39

Ai is doing that, classically again. Human-centric teachers and educators and I'm not suggesting replacing every teacher I'd value them massively but now you're getting individual education. You're getting career paths you know, indicated for you better than anyone else. Coaches and mentors the same thing. Authors and story writers or storytellers were the creative professions blocked by ai, but I can configure ai prompting at the moment that you will not tell that it is actually ai artists and performers. Well, it took 115 years to build 1.5 billion images. By that I mean pictures and whatever. In the last year and a half, ai has developed 115 billion images.

Speaker 2: 22:19

I'm not suggesting that michelangelo should be replaced by a robot, but there's a classic scene from I robot with will smith, where he talks to the robot and says can you paint a ceiling like michelangelo? Can you create a whatever, a symphony like mozart? And the robot turned around and says well, can you? You know? So again it's there. Film directors, producers, movie creators, creative strategists and innovators. Innovators, look at the latest advert for for uh coca-cola for Christmas. Don disappeared. Go on to YouTube and look, nine years ago, at the trailer for Morgan, the horror movie created by IBM Watson actors. Well, look at the Irishman you know, know fully. Ai generated. You wouldn't ever know.

Speaker 2: 23:01

Senior business leaders and executives Well, I can create an agentic board. I can create an agentic set of advisors. I can create policymakers that will answer policies far better than one, two, three year qualified individuals. Judges and legal arbitrators are increasingly using ai to make decisions, which is exciting and worrying. Worrying when there's ai bias built into it. And guess what the ai is not bias is the people who built it are bias. Therefore, we've got a problem. Emergency responders possibly, not.

Speaker 2: 23:29

I still need to drive out, you know, to touch people, but the amount of tech that's going into health care is just phenomenal. You know, I'm watching at the moment xr and vr, virtual reality and whatever where doctors are actually performing operations at distance and by that I mean they're in a different continent allowing them to perform precise surgery. They're using ai to take apart, you know, the body, the brain or whatever else. They're twisting it and turning it in real time. They're using precision instruments to do it. Skilled trades maybe you know plumbers and electricians. That's fiddly, you go down on your knees. But I can show you technology that will actually build the outer layer of a house. You know human skills.

Speaker 2: 24:09

This is the bit where we start getting into. You know well, humans are emotionally intelligent. Well, uh, sorry to disappoint, but only 36% of us are, statistically, and I can show you AI that will rewrite your emails, rewrite all your communications, in a manner that is emotionally intelligent, where the other 60 something percent of us are not Handling complex, problem solving queries. I can do that. Conflicting values or views, I can get the AI to argue. Now, anybody who's listening in and this is either going, oh my god, that's exciting or oh my god, we're all out of a job.

Speaker 2: 24:42

What I'm suggesting is AI should and can replace a lot of things that we do, and the question is is that exciting or not? Let's go back a little bit. When tractors went out on fields, everybody panicked because they lost jobs. No, better, better jobs were created. There was more people went into factories, there was more people went into offices and, at one stage, financial ledger work was considered the height of society. Now it's automated and digitized and it's not considered the height of society and there's no criticism of anyone who does that.

Speaker 2: 25:12

But the jobs that we consider to be intellectually rigorous, but the jobs that we consider to be intellectually rigorous that demand so much of us that you know require adaptive thinking, that require conflicting values or whatever else. You know, those you can automate because we are as humans. We're prompt, engineered. We go to college. Someone teaches us something regurgitated. We go to university, we regurgitated. These models are trained on trillions of bits of data that we couldn't possibly consume in our lifetime.

Speaker 2: 25:41

But what I argue is always let the technology replace the things that you know take up lots of time. That, if the technology can do it, fantastic. And there's a great thing out in harvard this month where they did a survey of hundreds of managers, gave them chat, gbt and other generative engines and they save seven, eight hours a week. If you can give me seven, eight hours a week, I'm throwing technology at it. Ask your future self in five years time, do I want to be doing the things that I'm doing, or could I get AI or automation or tech to do it? You'll probably say I wish I got tech to do it. Well, that's the good news.

Speaker 2: 26:14

So put the tech out, and then what I'm recommending to people is you need to have a lifelong learning uh attitude. You know what you now know won't be what you knew 10 years wasn't worthwhile, you know, unless you're into some professional doctor or something. Really you know dentistry, whatever. It is something really precision that's valuable for the next while you know. So, concentrate on the bits that are uniquely you.

Speaker 2: 26:36

Pass to technology the bits that you know you can automate and digitize it or do whatever else. Generate that lifelong learning perspective. Generate agility in your ability to work and to become valuable. Focus on the things that are uniquely human, where you won't get out competed by technology, because the warning's been out there for years. If companies can automate and digitize it, they will. Therefore, why wait for that to happen? Why not understand that there's so much that all of this technology can do and will do? It's not a case of the technology being the problem, it's our imagination is the only real limit at this moment of what we can do. Therefore, learn new skills, you know.

Speaker 2: 27:17

Go on again, dare I say, go to the IMI, go on courses, go on the things that keep you uh, you know up to date and skilled and valuable, and work and concentrate in the bits that you know, where you uniquely and truly add value, where technology will or won't currently replace you.

Speaker 2: 27:34

It may in the future, but that's for us to keep evolving as the technology evolves. So be in a constant state of change is probably my best advice, and a constant state of learning. So kind of exciting for me, farah, because I wouldn't go back to what I was doing five years ago and the things I'm doing with technology now. I wouldn't go back to doing those manually either. I'm now freed up to do more, deliver more, be more innovative and creative and, in all honesty, I haven't been happier in my work for years because the technology is taking the strain and I'm getting more time to do the bits that I adore and find fun, and that's making my career all of a sudden one of the most exciting time periods that I've had in in a decade. More, that's great Kieran.

Speaker 1: 28:20

I think a lot of people's fears with AI is just that it's going to replace the creativity you know. It's taking away things like writing or drawing or film production or any of those things where people have those creative outlets. But it's great that you're talking about how it's making you more creative because it's taking away some of, I guess, those mundane tasks. So thanks very much for that perspective. I want to end off with just one more question, and that is just about the changing nature of human relationships. So someone on the webinar brought up that a big risk is the fact that we used to have these relationships that were always human to human and now it's kind of human to machine. Maybe it will become machine to machine. What will that mean for human behavior in the workplace, relationships, mental health, employee engagement, all of that?

Speaker 2: 29:09

Yeah. So let me use a phrase a journalist friend of mine said, and she said if it, if it bleeds, it leads. So therefore, the worry is is you know, skynet, jobs are going to be taken over, human relations are going to be destroyed. Well, people said the same thing in 1956 about mainframe computers. Obama said it in 1990 when he was exiting, about automation was going to destroy jobs, remove the social fabric. Socrates, believe it or not, in 1500 and something, when the pencil came out, went on strike arguing that it would remove the ability to orally tell stories and stories would get lost. Instead of going actually now they're written down, they're captured forever. So again, the fears is not to denigrate anybody's fears or whatever else, but the fears have been around for years. You know, just because we're home working.

Speaker 2: 29:56

I hated COVID and what it did to society. I hated what it did to the negatives, but it made us reconnect with family members, it made us see each other as more human. Silos inside of organizations disappeared because once we seen inside of each other's kitchens and lives, all of a sudden, you know, that darned accounting department or that darned sales department became names, real people, and now forget technology. We've reverted back to being busy siloed organizations, political animals in many cases. So it isn't about the tool. We have human agency, which means we can do what we want we can be friendly, we can break down barriers or we can erect walls. So AI has not suddenly created this. What you're looking at and maybe some of the commentary coming through there is something called agentic AI, and agentic AI is, if 2025 was around, generative AI models 2025, 2026 or 24 was generative models 2025, 2026 is agentic labor, which is basically digital workers sitting on top of large language models that act autonomously, and people are getting slightly worried about this going. Oh, my goodness, they're going to replace my jobs. But look at existing supply chains. It isn't people moving bits of data around the place in the back of amazon. It's robots. It's people, it's connected machines, it is technology talking to technology using robotic process automation, apis, java, json, it's a whole host of things. This is just another tool that's allowing another layer of efficiency.

Speaker 2: 31:26

So the classic phrase and I heard it. I was in Boston last week and one of the heads of one of the biggest companies in the world and you see all these phrases that come out said this could be the last generation of workers that mind people only well, with respect, we've been putting technology into businesses for the last 100 years, so, as managers trust me, it's not the last generation minding people and technology. We've been doing that for decades, and so we should, and what I mean by so we should, far as we should be busting a gut to automate the automatable, we should. And what I mean by so we should, farah, is we should be busting a gut to automate the automatable. We should be busting a gut to take the robot out of the person, to allow us to be our most human, and therefore technology should be employed where technology can be, to allow it to do the data crunching, the ideation, the creation of stories who knows? Or go back to art. I can't draw for toffee, but I can prompt really well and therefore I have a creative outlet in AI because it can draw the things that I possibly can't, you know, but let the technology do the things that technology is great at, to free up us, up as humans, to do the things that we're amazing at.

Speaker 2: 32:38

And the more technology I've thrown at things in my life, while being very conscious of what I'm doing and the impact of it and the biases of it, and you know what I call cognitive offloading. In other words, throw it away and don't think about it. I do, I'm conscious of the tech, but I'm conscious of my own humanity. It isn't as someone said earlier on. But what if the technology tells me what to do? Great, don't do it. Who cares? We have total control. Let's choose to live the most amazing life by giving the tech the mundane stuff. Let the tech take the robot out of us so that we can be the most human that we have been in the last 50 or 70 years.

Speaker 1: 33:14

That's brilliant, ciarán. Thank you so much. We will leave it there for today. I know there's a whole lot more we could discuss. There's still extra questions from the mini masterclass that we haven't gotten to, so hopefully we'll get to those in another podcast. But thank you to everyone for listening and thank you, ciarán, for being on the IMI podcast today, and you can subscribe on SoundCloud or YouTube or your preferred podcast, provided, so ensure that you don't miss an episode.