People don't want to talk to you because you think you're looking at ways to replace them. The most important thing I try and do is to say, I'm not coming in to replace you. I'm trying to help make your job easier.
SPEAKER_02What's your definition of a chief AI officer?
SPEAKER_00It sort of sits in between a chief technology officer and a chief people officer. You are not just the AI guy. You are a holistic person that looks at the whole business functions as a whole, looks at the processes, looks at the people. And you're the sort of jigsaw that pulls all that together and then adds the AI on top of it to make it more effective. Instead of a business taking 47 hours, which is the UK average, to respond to a client, we now can respond to clients within 60 to 90 seconds.
SPEAKER_01Chris Duffy is a chief AI officer and former British Army leader helping UK businesses use AI to solve real problems. He brings clarity, strategy, and a human touch to cut costs, save time, and drive growth.
SPEAKER_02Welcome to Using AI at work. I'm your host, Chris Dave. Each week we'll be learning how today's business owners, entrepreneurs, and ambitious professionals are getting more done with smart use of tomorrow's tech. Let's get started. Right now, every business leader is asking the same question. What are we going to do about AI? If this is you, ChiefAIOfficer.com has the answer. We give you a simple path forward where we provide executive and team training so your people know exactly how to safely use generative AI in their day-to-day. We also manage the deployment and implementation to make sure tools actively get adopted and deliver results. And we'll also guide company-wide transformation so AI becomes part of your operating system, not just another shiny object. The companies that act now will increase productivity, cut costs, and grow faster than their competitors. Those that wait will get left behind. So if you want to make AI work in your business, visit chiefaiofficer.com and see how we're helping companies of all sizes finally get results from AI. Hey everybody, welcome to the next episode of Using AI at Work. This is Chris Dagle. I'm your host, and today we have a chief AI officer named Chris Duffy calling in from the Bournemouth area, big member of our uh chief AI officer community. And I wanted to sit him down today because he's doing some pretty incredible stuff for companies, maybe companies just like yours if you're listening to this podcast. So, Chris, before we get started, um you've got an interesting history that's led you to becoming the AI expert. So if you don't mind, maybe share kind of how you arrived at um at being a chief AI officer.
SPEAKER_00Yeah, no problem. So um I finished 23 years in the British Army uh in communications, um, and then I left in September uh 2024. Um and in my last three years um of doing that, I was one when ChatGBT uh 3.5 came out in November 22, I was all in on cybersecurity and governance and sort of the transition I wanted to make. And then once I saw that position and that was going to be, and I was like, wow, this is a game changer, I sort of switched focus completely into the AI space, and then I went down a rabbit hole of trying to learn as much as possible about uh AI ethics, AI governance, AI strategy, and then started to move further and further down the line. Um and then as we come on to sort of the the September time uh in last year, and yeah, in last year, then I saw the chief officer course and thought, why not professionalize in the AI strategy side of it and get something good? Uh I was studying already for an MBA, uh, which I'm due to finish at the end of this year. But um, since I did the chief fire officer course, I seem to have got more credibility from that than I do from you know an MBA that quite a lot of people have got. And it's quite niche. You know, there's a lot of um a lot of people often contact me and say, What's a chief fire officer and how can I become one? So I mean that that's been really sort of a good talking point for people to engage in me.
SPEAKER_02What do you say when people ask, what is a chief, what's your definition of a chief AI officer?
SPEAKER_00Um, it sort of sits in between a chief technology officer and a chief people officer. You are not just the AI guy, you are uh a holistic person that looks at the whole business functions as a whole, looks at the processes, looks at the people, and you're the sort of jigsaw that pulls all that together and then adds the AR on top of it to make it more effective. Um, you know, people always look at uh the CTO as the guy that manages all of it and the you know, a change management officer or a people officer that will manage the people and their training, but nobody pieces the two together.
SPEAKER_02So what do you uh because I'm I'm sure that a lot of companies out there, maybe not the listeners of this podcast, but but most companies when they think about AI, they think, oh, that's a technology, therefore that should fall under the CTO. Um, do you experience that kind of uh belief when you're talking to companies?
SPEAKER_00Uh no, not at all. I mean that there's um the there's there's a number of things I've I've talked about it before, but the uh the CIO function and the CTO function, both of those roles are incredibly busy and they have you know really grounded in functions that have been there for years. And to add AI on top of a CTO's role and responsibility just puts an incredible amount of pressure on them that they already don't have capacity for. So it becomes a nice to have. Um, and then you see IO's already got you know the whole data problem to be managing and and ensuring that that's effective. So to then give him that role as well is is again pushing them completely out of capacity. The CI C AIO sits in the middle, and uh, you know, they know the technology side, but if they need more technological expertise, they can go to the CTO, and if they need more the data management, they can go to the CIO. But what they're looking at is the piecing of both of those functions together, and then how that can be driven forward to uh drive return on investment.
SPEAKER_02Nice. Okay. I I think that's that's helpful because for companies that that aren't necessarily like they don't know the difference between maybe analytical AI and like applied or generative or or operational AI, they kind of think it's it's all this one, you know, heavily technical thing and they need some PhD with data science to come in. But I I I agree with you, that's not necessarily the key. You might want that person to support the CTO, but for the most part, it is a it's a people thing.
SPEAKER_00Yeah, I mean, you know, every company doesn't need a CIO right now, but they would massively benefit from a fractional CIO to come in and give them some advisory, which is you know a lot of the work I'm doing. And um I always when I always go into a company, I always approach it with a problem first approach and a people first. It's always uh what's the biggest pain, what drives people crazy every day? You know, what what are the things that are costing you a serious amount of money or taking the most amount of your time? And then straight away the the border aligned in, oh, this we know that this has been driving us crazy for months. Is there anything you can do about that? And then you become the problem solver rather than the the AI guy, and that's the sort of balance that you need to strike. Nice. I don't I don't always I don't go in with an AI as an offer. The AI on it comes in, you know, the AI tools and the support and the the transformation type of it all comes in after we've identified the problem and everybody's aligned to that problem, and it's aligned with their business goals.
SPEAKER_02So that that's that's good. How where are companies like why are they reaching out to you? What what is the story going on in their heads that they need to talk to to Chris Duffy?
SPEAKER_00I think it's the uh there's it's the hype is the main thing. You know a lot of times I get we need we need AI in our company and and like why for for what purpose? And you know, a lot of companies are going, oh we already use AI, we we've got copilot. And you're like, okay, well what what problem are you trying to solve with copilot? And they go, Oh, well, people are writing emails with it. Yeah, that's great. But you know, if you gave if you gave all of your employees copilot, but you gave them a good training package and the problem that is driving you crazy that you need to solve, then the people are going to be a lot on board more on board with it, and the return on investment would be tenfold.
SPEAKER_02Yeah. Yeah, that's a good point. Writing using AI to write emails is certainly helpful, but it's capable of much more than that.
SPEAKER_00So Yeah, that's that that's like the one percent of what you can use it for compared to.
SPEAKER_02But I guess that's kind of like a gateway drug that's universal to every knowledge worker. I have to send emails, right? So I guess that's a good start.
SPEAKER_00I mean, people spend most of the days on emails, don't they? So it's the first thing they'll go to to try and make easier.
SPEAKER_02So the the clients that um you're working with, that they say, okay, we need AI. Tell us what that means. So you come in and the first conversation that you're gonna have is typically with who?
SPEAKER_00Ideally, a decision maker. You know, I d I don't want to go and speak to a CTO uh or a CIO. You know, in a big enterprise, it's understandable that you probably end up having to start there. But you know, the the smaller businesses, you have to speak to, you know, managing director, CEO, um like pretty much straight away. Because otherwise, you know, you can paint the world of a far better place with AI, but then they've got to then go and upsell it to their MD to make that decision. So, you know, you're already on an uphill battle and a back foot before you even start. And I'm not trying to sell AI. You know, if people want AI and they want to come to me for it, then you know I can offer advice on that. But I'm not walking around with a billboard and and uh a sale offer of come and get AI from me.
SPEAKER_02Not a bad idea. I should probably wear a shirt like that. Um so once you once you have that conversation, where are you finding that that it's starting to click for the leadership as far as like, okay, here's the way that I need to be looking at AI as a business leader? What are you telling them that that you think really helps them make that have that aha moment?
SPEAKER_00Um so I mean, when when we first meet them, uh normally the initial chat is is remote. And then uh myself and my business partner will uh both join the call, both with note takers, and straight away there, you know, we'll always say, We're gonna record this call. Is this okay with you? Or you know, they'll get an email prior to tell them that there'll be the call will be recorded. And then we can draft a proposal uh with a service offering one, two hours after the call, which will normally take somebody eight to ten hours, yeah, and it'll be around about 90% there within 10 minutes of being off the call. Yeah. Just because we can take that transcript, put it into the business brain that we've already built within the LLM, and then uh so I use Claude and I've built a business brain within there that knows everything about me, everything about uh Tim, my business partner, everything about what the business does, what our KPIs are, what our service offering is, our ACP, all that's built into there. So when I put my transcript into that, and then I ask it to draft the proposal, yeah. You know, it's it's drafted at 90% ready pretty much straight away.
SPEAKER_02Anyone listening to this, I hope you just caught that. Proposals are necessary, but they can be a big uh pain to produce. Very labor-intensive, no guarantee they're going to turn into business. The way that Chris just described this, with the way that he's using an AI-powered note-taker and the large language model Claude, which is his preference, and some of the pre-configuration he's done, he's able to take get off the phone call and have 90% of that proposal ready within minutes. So uh that's a fantastic use case. Do clients ever ask you to set that up for them?
SPEAKER_00Uh yeah. So once we start into an advisory, you know, that's one of the first things we say is you need to record your calls. You need to get your uh staff to have have calls recorded and and transcribed so that they can take that information and be able to do more with it. And I mean, even a great use case, you know, even from just transcripts and calls, is I can go back to a call that I had three months ago with a client that, you know, said, Oh, can you come back to me in a few months? We're not quite ready yet. And you go back to them and he's just had surgery on his knee when you spoke to him three months ago. And then three months later, you know, you quickly look at the call and go, Oh, he had a knee operation. Yeah, or you can even go in. I use Fathom as my note taker, and I can go in and say, ask Fathom and say, Oh, was there any personal information shared here that might I might want to bring up to build rapport? And then it would come back to me and say, Oh, ask him how his surgery is, and you'll go back into it, and straight away that rapport is there, and they're like, Wow, how do they remember that?
SPEAKER_02AI. Fantastic. So what once you've kind of they say, okay, you're you're the person that we'd like to work with, what are the next steps for you when you are working with a company?
SPEAKER_00Um, so firstly, to understand what their goals are, where do they want to be, what are their what are their current baseline and where do they want to be, you know, because uh my business partner's been in the growth advisory for 20, 30 years, you know, scale companies to hundreds of millions. So his expertise lies in there and in marketing, and he's really, really good at that. So, you know, we get the baseline of the business, understand all of that, and then we look at where the pain and problems are, and then we start to look at where um they can be improved, and then what AI systems and tools or processes can be eliminated or replaced to make them more efficient. So it's it's a good you know, it's it's a one to two day discovery, and then probably five days overall with the questions back and forth. And the the larger the company, you know, you might have to go around, speak to the staff, understand their processes, those sorts of things. So, you know, you you're building a picture of of exactly how the business operates and what the things are that are causing the business problems and the bottlenecks.
SPEAKER_02So if I was a business owner, one one of the things I would hope that my people are so busy that any disrupt, like they just don't have time for disruptions like answering questions from the AI consultant or any consultant, really. Does that come up like, hey, how how much time is this gonna take for my people? And if it does, how do you kind of mitigate or address that concern?
SPEAKER_00So it the the main thing is apprehension. People don't want to talk to you because you think you're looking at ways to replace them. You know, that fear and angst is is is rife. And yeah, the most important thing I try and do is is say, look, I'm not coming in to replace you. I'm here. I'm trying to help make your job easier. So tell me all the things that you hate doing, and I will try and remove them and make your job better and enable you to do higher things, you know, more important things that can drive business growth. So I mean if you're if you're a business owner, a small business owner, and you are fully embedded in the company and you're focusing on the day-to-day work of the company, you're never gonna grow because you're never stepping back holistically and looking at where the company is now and where it should be going. So until you can take some of that pressure off, then it's really a struggle to grow.
SPEAKER_02So some some of the people listening might say, okay, this is great. I'm gonna get, we're gonna get all of Chris's steps, and then I'm just gonna do it myself. Um what are your thoughts like as far as result or impact of because I'll give you a perfect example. There's a business um uh operating model out there called EOS. A lot of listeners may be familiar with entrepreneurial operating system. And the reason I'm using that example is because they've got a book called Contraction that outlines the process. I could get the book, maybe do those steps myself, or I can get an EOS implementer who has already been trained and certified and comes into my business and just kind of runs that for me. So do you have many businesses that say, hey, we've either tried it ourselves, we're good, or thanks for the information, we're gonna give it a shot doing it ourselves first?
SPEAKER_00Uh not that I've encountered, but what I normally get is uh, you know, you can send you can send a proposal to somebody and then they'll go, right, thanks. I'll I'll look over it. And then, you know, a couple of weeks go by and then you go back to them and go, oh, actually, it's not really a priority right now. I I'm I'm so busy with what I'm doing, then I I'll I'll look at it in a few months. And you know, that that's quite common, and it's it's not that they don't want AI, it's just that they are so busy in the day-to-day fighting fires that you know they they don't have the capacity to take on that next problem. Yeah, and what they don't realize is if they made the time for that, they would get all that other time back that they don't have.
SPEAKER_02They start opening up hours that first week. Yeah. Interesting. So Okay, now you've gone into the company, you have identified some of these pain points. Are they usually pretty clear on what those are, or do they need help kind of discovering, or do they end up saying, hey, we thought we knew the what the problem was, but after your audit, this is actually the problem.
SPEAKER_00Yeah, I mean, if you if you sit in a in a room with with six people and you ask them what the biggest problem in the business is, you'll get six different answers.
SPEAKER_02Yep.
SPEAKER_00So I I created frameworks when I started my own company. I use something called the engage framework and the ignite framework. Um now the engage framework is fully designed to align a leadership team to the common problem and then assign ownership to it. So you need to understand what your problem is. Everybody needs to agree that that is the biggest problem. You might come up with five or six problems in that in that discussion, but you know, pick one problem, you know, and everybody agrees that that is the problem that needs to be fixed. Then you align somebody to own that problem, and then once you've got that from there, then you need to buy in from the the people that you're going to implement it in. So then you need to start managing that holistic transformation. You also need to vet the tools, you need to know which tool fits the business. Quick, you know, you might have everybody uses Copilot, however, or you're on Microsoft, so you use Copilot, but half half the company might use Google. And they might use Google for a different reason. So you know, you you have to make the tool fit the company rather than the company fit the tool.
SPEAKER_02Yes.
SPEAKER_00So then in the ignite framework, following on from there, once we've got the problem, the solution, and the owner, then the ignite framework is the implementation team. So, you know, you find your AI champion in your company, somebody, you know, most of the IT guys are already passionate about AI and know quite a lot about it. Cool. Yeah. So if you can get somebody inside the business rather than bringing somebody in to do the tr the implementation, then you are a hundred percent better off.
SPEAKER_02Yeah. So tell me about that. Like we've identified the problems, we've now kind of like we've we've settled on some. And you mentioned finding some people inside the company who what is their role? Let's say you you've got a couple of the maybe from the tech team, maybe from the marketing or ops team, but people who are like, I'm already using it at home. I I'm I get it, I'm thinking in AI, as we like to say. How are you going to kind of structure their efforts to support the company's ideal outcome?
SPEAKER_00So the key thing I always do, I mean, I I teach master classes and I teach the same thing in this is um you've got to get buy-in from the people that you're implementing the tool because to them it's just another tool they have to get used to. But what you what you've always got to do is show them the wow moment. Yeah, once once somebody goes wow, and then they start thinking, you know, the mind starts running on the the multiple different things that they could potentially use this for in their day-to-day work. That's when you get the buy-in, and that's when people are interested in starting to adopt this tool. And the second thing is is the training package that goes with it. So you need to teach people how to use it, but how to use it in the way the company wants you to use it. Yes, not just all the different things that are available because then they get overwhelmed. And then the last thing is is feedback. A lot of companies don't give any feedback. So if you've I don't tell people to or companies to roll it out in one department, I'll say, you know, pilot it in three departments at once in small silos, and then get the feedback from each department constantly, and then sit them all down every week, and they can sit in a room together and discuss what works, what doesn't work, what they're having problems with, and they can share those problems and share those experiences and gain so much more from that, yeah, rather than trying to do it on their own.
SPEAKER_02Kind of that collective sharing and intelligence. Do you, as the the fractional chief AI officer, are you participating and facilitating those meetings?
SPEAKER_00Um, yeah, so we have we have weekly meetings, um, you know, um one to two hours a week, depending on the company size. Uh, but if they need if they need more help or any guidance on any of that, then you know, we can definitely get involved. So you mentioned this is the way to do it and see you later. Yeah.
SPEAKER_02Yeah. You know, here my experience has been if you if you train the people and they have the wow moment, they'll get some benefit. But if there's not somebody from the outside that's that's kind of continuing to to nudge or push or check in and not creating that tempo, that the adoption it'll it'll Wax and wane, let's say. So you mentioned this wow moment. And I love that idea because once somebody has that wow moment, I've seen it, you've seen it, like they're like, oh my God, didn't I didn't know it could do this? And then they start going, oh, I wonder if can it do this, this, this. How do you help facilitate that wow moment with these folks?
SPEAKER_00Um, so there's a there's a couple of different things. I've you know, what I try and do is demonstrate it, okay, demonstrate it to the to the the people I'm training. Uh the masterclass wow moment is, you know, we we build AI voice agent receptionists for people. So um I'll throw it up on the board, call me, and I'll just say, you know, you're approaching me. This is my uh this is my AI assistant. If you call this number and you want to book an appointment with me, then you know, you can come to me with a business problem, talk to this receptionist, tell it your problem, and then see what it can do. And then they come off the phone and everybody in the room's listening, they're just like, wow. And that'll book an appointment in my calendar for me. It'll send me a message afterwards and tell me that somebody's booked an appointment with the problems that they've got that they want to talk to me about. You know, that's one thing. And the second one I always get a wow moment is is um Gemini Studio, when you click um share my screen, and then you just say, Hey Gemini, uh, I've got a problem building out this automation. Uh, can you help me with this? This is the problem I'm seeing. Can you see my screen? And then it's like, yes, I can. And then they start talking to you back and explaining how to do it, and you just see people sat there going, that's beautiful.
SPEAKER_02So are you doing a lot of builds or are you doing mainly just teaching and strategy?
SPEAKER_00So the the intention when I started the company was to stay at the strategy side, but yeah, uh, I'm finding I'm building out automations more and more, uh, and I'm starting to build out custom solutions for companies as well. So there's one client that I'm working with, we're building a SKU management system that uh can turn um any product, uh, they can upload CSVs, PDFs, and they can put it into uh the system we've built, and then it will generate a uh web ready description in their own brand voice to be able to upload to the website. And that you know, normally takes any any company like that with a with a marketing um copywriter two two hours, two to three hours per product, and we can do five to ten in uh just under two minutes.
SPEAKER_02So just just for those listening to make sure that you're clear on this, you can take a CSV with SKUs and what else is in there? SKU and product name, I guess?
SPEAKER_00Yeah, SKU, product name, uh a quick description from the manufacturer in in what the product is, and then it will pull all of that information and then it will generate a brand, a brand voice, short description, long description, uh, and SEO, you know, meta description, all of all of that for you straight away.
SPEAKER_02And it can do uh a whole catalog of products at minutes as compared to having an entire copywriting team spending their weekend crunching this out. That's fantastic. So are you encountering you you mentioned it briefly, but I I would imagine that if I was an employee and I've heard all the hype about AI, specifically the kind of the fear-mongering about AI is going to take jobs and stuff like that, which uh there's no doubt it will, um, and somebody says, oh, here's the here's the AI consultant, they need to talk to you, I would be nervous. How and and you mentioned that you tell them um I'm I'm here to make your job easier, but is there any more to it to kind of like uh managing their expectations or concerns so that they're cooperating with the AI efforts as compared to uh introducing friction into the AI efforts?
SPEAKER_00And that that's a really difficult problem because it all depends on uh what the board what the board wants. If the board want to uh streamline the business, implement AI and streamline a business, then you know it it's understandably jobs are going to go, especially normally at the entry level. Like I said, it's it's understanding what the what problem they're trying to solve uh and where the business is going. And that's why we always focus on that side first. And then um I tend to not like taking as many of those jobs on because it's difficult, not because I don't enjoy the difficulty, but because it's you know, you can see where it's going. The emotional toll. I always focused on my business and vision in people first, yeah, and a people first organization. So it's meant to empower people and and help people and not replace them. And I know it's going to replace a lot of people, and you know, the the advice to people that uh are in those jobs where you know those those are going to go is you know, get get used to AI, it's not going anywhere, and it's only get more embedded in everything we do. So, you know, start to learn and understand it, and then make yourself more valuable in in what you do and show them that you know that that's not the only thing you can do. You're not there just as a data analyst, you can also do X, Y, Z, you know, and and people will adapt. You you had uh the Amazon warehouse um case study where you know they had so many people that were taking things from the shelves, packing boxes and all that, and then Amazon brought in all of the robots to do that, and all of those people, you know, not all of them stayed, but a lot of the people that had the skills were then trained in robotics and maintenance and interesting the the data management of that side of it. And it's just people people will adapt and and businesses will adapt, but it's just uh based on the speed and and the sort of uh capacity they've got to keep their employment.
SPEAKER_02So tell tell me what this means to you, this people first. What does that mean to you?
SPEAKER_00So you're going in to solve a problem that is peep that is pain, that is people's pain. It's something that you know has been driving the company mad for so long, especially you know, the smaller businesses. A great example for that is uh there was a lady who she was quite well paid in it in her business. She was a marketing uh consultant and she used to take in uh LinkedIn, Facebook, um all the different ads, put your meta ads, pull them all in. And then she used to sit there in an evening and read through and work out the outreach of each one. And I said to her, you know, you can put this into a dashboard and use an agent to search all of this for you, scrape everything, and then give you the outcome every morning at nine o'clock. And that was 68 hours a month. I saved it all. Wow. Just just from the initial conversation, and it's that and that's that's where I get the value, and that's where I go, wow, I've helped that person so much. And that's that's what you want, and it that's the job satisfaction, is is being able to do that for people. And then, you know, they say, Oh, go and speak to this guy, he's saved me so much time, and now I'm taking on three times as much work as I used to have. So again, it goes back to the growth. Aaron Powell I like that perspective.
SPEAKER_02Certainly, you know, as a as a AI professional, you start with the executive team and you ex you evaluate their strategy and you look for ways that AI can help them achieve that strategic goal faster or with less resources. Like that that's the obvious part. But this positioning of coming in and helping the people, I don't know that that's necessarily what a lot of executives associate with, hey, we've got an AI consultant coming in to talk to us. I don't know that that's that's the lens through which they're looking at it. Certain for sure. But now you mentioned earlier and and before we started the recording of this today, you were kind of briefing me on some of the uh the management of AI usage within the company. Let's say you've come in, you've the team is now, they're thinking in AI, they've had the wow moments, you've you're saving them time, um, and these employees and team members are starting to, they know enough to let's say be dangerous. They're like, ooh, I could use AI for this, ooh, I should probably be using AI for this. How are companies, how are you recommending that companies um uh manage and like kind of just uh uh control this?
SPEAKER_00Um so the control is the hardest part because you know, if you gave somebody a shiny new toy and then said don't touch it, yeah, you know, it's yeah, it is it is gonna be difficult. Um, but I mean that the hardest thing people can do. Yeah, I mean, I've I read the MIT report about um you know that the different uh consultancy companies that had gone in and did the survey and they said 90% of companies are embracing AI and 10% have said don't use it. You know, it's a complete ban uh on on the usage of AI in the workplace. Now I can understand that in you know if you're working in a top secret environment and that sort of area, but you know, if you're if you're in a business that is not working in highly sensitive material, less personal information, you know, then why not? And if you tell somebody they can't use something within the business, then they're gonna use it outside the business.
SPEAKER_03Yeah.
SPEAKER_00And you know, I post about this a lot on LinkedIn, shadow AI is is a key is a key factor and a key problem. Um for those that don't know, shadow AI is using uh your own personal AI tools in the workplace and putting company and proprietary information, you know, client personal information into open source for anybody. And you know, that's a real risk for companies. And if they don't put systems in place that allow people to use AI in the workplace to do what they need to do and to make their jobs easier, then they'll find another way. You know, huh humans are an inherently lazy race. If we can find an easier way of doing something, then we definitely will.
SPEAKER_02Agreed.
SPEAKER_00In terms of uh, you know, in internal LLMs is is a great one. Uh for you know, if you're in a more secure environment such as Pharma, you know, you can uh you've got your meta LLM, you've got um Mistral, which is French, which you can take completely off the net. It doesn't even have to be connected to the internet, you give it all your own information and then you host it internally, and then it can have access to all your files, and people can question that to find files or information within the company. And that is, you know, people spend 30 to 40% of their day searching for information.
SPEAKER_03Yeah.
SPEAKER_00And if you can find something that can find you all the information you need in minutes, then you know you've gave that person 35% of their day back.
SPEAKER_02Heck yeah. So what what are your thoughts about let's say a company says we we don't think we need to have a custom LLM, uh uh a meta or a mistral on site. Um, but we do still want to to be able to work with our IP. Would you say that's a no-no, or are there configurations that you recommend so that it is?
SPEAKER_00Yeah, there's there's there's lots of different ways of doing it. You know, you can use something called OneTrust. Uh you've got Service Now, which can set up uh asset management, or you can put in AI use cases into a form, and then that can be uploaded to ServiceNow, and then you can have dashboards to see um you know what what the risks are for each one of those use cases. Uh you can use a zero trust environment where you can set up a list of questions or information that can't pass outside of the company, but you can request and pull information in. So, you know, that there is lots of different ways of doing it, but it it's all down to the use case and it's down to what the company wants to use the tool for. And again, what problem are they trying to solve is is what it'll always come back to.
SPEAKER_02So let's say for a company that said we're, you know, we're we're using it, we probably need to, we just they just heard that. They're like, we probably need to think about how we're letting our people use it. What would be kind of the the entry level or at least the basics of making sure that there was some sort of secure usage framework for their employees?
SPEAKER_00So an AI policy is the first thing I'd say. You know, you you have to have an AI policy and it has to contain the guardrails. You know, tell them what they can and can't use AI for within the workplace, and then it creates accountability straight away. Because if they're made to sign a policy to say, we can't use we can't put personal information or company information into this LLM and then they're found to be doing it, then obviously, you know, the company's got more weight to come back to them with. But if you don't have any policy at all in in terms of AI, you know, how do you hold people to account for what they're doing? So that's that's the first advice I'd say is you you you have to have an AI policy in the workplace. If you've got uh employees you're responsible for, then you need an AI policy in place.
SPEAKER_02For somebody that's not necessarily even that aware of what's possible with AI, how do they how would they you recommend they start with a pot a use policy? Are you a use policy as in an AI policy or like they're hearing this and they're like, hey man, we we probably need to have one of these set up for like some frameworks for our employees that are using Chat GPT in the business. Where where would they start?
SPEAKER_00I mean there's this ChatGPT's got uh in ChatGPT you've got you know your custom GPTs and there's a custom GPT library, and you know, you type in AI use policy creator or AI use policy, and there's a hundred different templates in there, and they'll just ask you one question at a time. And as you go through the question set, eventually it'll draft you a policy. So, you know, there's really simple ways of doing it in that way. Or, you know, speak to a professional. If you're not not sure and you don't want to just make something that you might not think is fit for purpose, then just reach out to someone.
SPEAKER_02Yeah. So that's the hard part I find is that there's high demand for AI expertise, there's low supply, and there's not really like you can't go to upwork necessarily. I I mean, I guess you could, but I wouldn't be looking for a chief AI officer on upwork. Like, what would be some questions if if if a client called you and said, hey, we're thinking about we've enjoyed working with you in the past, we need to find somebody, let's say, full-time that's going to come in in-house. Uh, we'd like for you to give us some advice on the kind of questions we need to be asking to make sure that that person is really truly like a good fit or has the AI expertise that we need. What would be some advice you'd give them?
SPEAKER_00I mean, it one one thing I've found in, you know, in the in the US, you've got chief AI officers and they're quite uh extremely popular, and you know, that they are a flavor in government, in you know, yeah large large enterprises all over the US. In the UK, there's I can probably count on one hand how many chief officers there are, but what you do have is a lot of appointments of head of AI. Now, head of AI, the the prerequisites for that is normally PhD, machine learning, you know, really high-level scientific um intelligence and sitting at a board level, you know, as a head of AI, you are not, in my opinion, uh anywhere near as placed as a chief fire officer would be.
SPEAKER_02Yep, agreed.
SPEAKER_00You can manage a a highly sophisticated engineering team and deliver some unbelievable research and development. But um for actually delivering a project and uh being able to look at it in the full holistic view of the transformation of the company, the business goals, the alignment, the understanding of the implementation of it, you know, all that sort of stuff. It needs a full holistic view rather than the engineering focus. So the advice I'd probably say is you need somebody who's got business acumen, you need somebody who understands uh C-suite, understands how business works, and then on top of that, they've then applied AI knowledge over the top, and they've got experience in uh either working as a freelance consultant in going into businesses and then seeing what see case studies from them, what have they already done, what have they achieved, what what are the outcomes of that? Because again, you can say, and I know the government, the UK government, um I'll I'll go back to Copilot again, but they implement they rolled out copilot to 20,000 people, and then uh they ran it for six months, and then they came back and said, Oh, we've got uh a 791% return on investment, and the average person is going to save 26 minutes a day. Now that sounds great on paper, but if you look at that on what that could potentially save each person, yeah, you know, that that's two hours a day minimum. Yeah, and it just that instantly screamed out to me that that that rollout wasn't you know done in a in a way that people understood, and it was just uh probably write an email, you get on type thing, and uh so you know that that's that's where the the key information is is you know it has to be a full process, it's not just implement AI and then walk away. Yeah, it's not a consultant business, you know, consultants deliver a project and leave. And that's why we do an advisory and not a consultancy, because we deliver a project, but then we stay within in touch with them and and you know follow up and make sure that that's still delivering, fit for purpose, still in date. You know, because you might deliver a project in AI and three months later it's obsolete because it's moving forward so fast that you know it it doesn't work anymore, or there's something far better, far cheaper.
SPEAKER_02Yes. Yeah, that's the challenge with this is that you know, if if you have somebody come in, teach your people about a certain tool, in 90 days that information could be wrong. Not because the the the AI expert or the chief AI officer was was wrong, but because this technology, it advances so quickly that they could have said, oh, never use this tool for that that task. 90 days from now, they could have had some major updates, and now that's best in class. So it is important. I love this this idea of the advisory where there's a regular, you know, uh dialogue going on with some updates from somebody like you who's in the field, who's seeing it be used, and who's actually testing the stuff or building with the stuff. But I want to uh make a comment about that that government case study. I agree totally. They're probably saving the 26 minutes by writing an email, like we talked about earlier, right? Yeah, exactly. 26 minutes over 20,000 people, um, that's aggregated, that's a pretty significant. So I want the listeners out there to understand that even if you're just barely using it, like there's still big wins, although you're leaving hours and hours of productivity per person on the table for sure. So, Chris, I know that you um you take this role very seriously and that you have done a fantastic job of kind of putting your uh putting your thoughts together in in content that you release regularly. Um you've shared it within the chief AI officer community, you share it on LinkedIn regularly, but is there a place where if somebody wants to continue to uh consume your perspective on things, because you're out there, you're doing it, you're working with, you know, we didn't name the clients, but some major like brands that everybody would know. Um, so your perspective on this stuff is it's not theory, right? How do people find uh how do people plug into kind of your position on things? Where would you recommend they go?
SPEAKER_00Um, so I mean I'm I'm most active on LinkedIn. I am starting to outreach into the other the other socials, but you know, being ex-military, I wasn't really a social media person. So, you know, I'm I'm I'm hyper focused on LinkedIn. That's that's where you'll get me at any other point. Um, and you can see some of the things that we're doing for other businesses with things like uh hot lead response. So instead of a business taking 47 hours, which is the UK average, to respond to a client, uh, we now can respond to clients within 60 to 90 seconds from just an inquiry on a website. So they submit the inquiry, hit enter, and they're still sat at a laptop, and then all of a sudden they get contacted straight away, be that by an AI voice agent, be that by an SMS, you know, they're still in that position where they want to buy and they're not instantly onto another website looking at another contact form for more information in another company. And the pet you know, the the conversion rate from being able to respond within 60 to 90 seconds is 391%. And that's from the Harvard study. So, you know, speed to lead. Yeah, 100%. So speed to lead. Uh so that that's something that's become massively popular and it's it's ever growing. We we were quite early into it. Um, but it is I'm starting to see more and more people, you know, cold emails in my inbox offering me hot lead response or cold lead regeneration, which is another thing we do for you know clients that are service-based with thousands and thousands of of clients they no longer speak to, and we can regenerate those lists using AI and automation. Easy stuff.
SPEAKER_02Yeah, but very lucrative, very big wins for companies. Well, Chris, we're gonna put all of those um your the link to your LinkedIn profile and any of the tools that you've mentioned, they'll be in the show notes here. Um, but I I guess maybe any last bit of advice for I want two perspectives. I want you talking to the executive and you talking to the employee. Any last bits of advice?
SPEAKER_00Uh the the the first bit of advice is you know, anything that you want to implement in a business, you have to A understand and B use yourself, understand what what it is, uh what the tools can do, and then what they can do for your business. Yeah. And you know, that as a leadership team, you know, you can speak to a CEO and say, uh, we we we really need AI, and you're like, do you use AI? No. So yeah, they have to use it and understand it and and to be able to implement it. Otherwise, they're implementing something they have no idea about. Good point. And then for the people, for you know, for the people who aren't in the leadership roles, you know, embrace it and start small. Just pick one tool and and just play with it and learn about it. And you know, that the the better you get uh understanding what the capabilities are, then you can see what other tools are available and just expand from there.
SPEAKER_02Great advice. Chris, I know it's been a long time that we've been trying to do this episode. I'm glad we finally did it and for you late at night, which I appreciate you uh making the sacrifice to share this information with the audience. So, everybody, thank you so much again for tuning in to uh using AI at work. We promise to have uh fascinating guests just like Chris on every episode, where we're gonna be discussing exactly how youth should be thinking about and using AI in your business. And with that, we'll see you on the next episode. Thanks everybody. Thanks for tuning in to using AI.org. Don't forget to subscribe for more conversations about how to use AI at work. We'll thank you to our sponsors, AI Officer for Empowering Businesses, AI Education and Training. Visit their website for free AI reading and assessment and AI strategy to help you get started using AI at work.aiofficer.com. Follow us on Twitter at the handle usingAI at work. And visit www.usingaiwork.com for free resources to help you harness AI in your role.