 Welcome to the CIM Marketing Podcast. The contents and views expressed by individuals in the CIM Marketing Podcast are their own and do not necessarily represent the views of the companies they work for. We hope you enjoy the episode. Hello everybody, welcome to the second, as James said, live CIM Marketing Podcast, and the 81st CIM Podcast has started at the end of our fourth season. So this is our season closure. We were back in September for season number five, quite remarkable, and Ali and I, she sat over there, started it all those years ago. It's been going strong. Quick advert from me. We have now had 130,000 downloads, if you give me 800, so just short of 130,000 downloads. And we have 5,000 regular subscribers on Spotify alone. A similar number on Apple Podcasts, and around all of the platforms we go out on, around 10,000 subscribers. So it's been a really enjoyable project, exciting. We've had some fantastic guests, and none better than the two gentlemen we are joined by today. Mr Ryan Miles, who is from one of the biggest names in business, Microsoft, he's director of international integrated marketing there. How are you, sir? I'm good. Thank you for having me again. Great to have you. Great to have you on the show. And Mr Vincent Cider, who is CEO of Get Inference, and also a course director at CIM here, specialising in artificial intelligence, which, as James is saying, is what we're going to be focusing on today. It's something that I think everybody is interested in, so we'll find out more from the guys who know. Before I started, a lady asked me if I would get nervous about podcasts. I do them so frequently now. I don't get nervous as such, but I do get a little bit nervous about one thing when we've got a live podcast, is that when we get to the Q&A at the end, is I want to see a forest of hands, or people being enthusiastic to be the first person to ask a question. We're going to get 20 minutes at the end, so all of those things have been bugging you, or grating you, or eating away at your bad AI that you want to know. Try and get your question in. We'll try and get as many in as we can. We're going to move 20 minutes at the end of the show to take Q&A from the audience. But before we get there, we've got 40 minutes of discussion with these two gentlemen, and we're going to start with you, Ryan, about if you could take us through some of the emerging technologies that you're starting to see specifically or mainly in the marketing and business landscape from AI. Yeah, for sure. I mean, it's definitely a transformational moment. There's been lots of transformation in recent history. We've had the cloud, and before that, the web, the internet, and now AI is really on the level of that. And we're at the beginning of the journey. And I think probably the way that most people have experienced natural language models and AI is through chat. ChatGPT is something I think most in the room will be familiar with and hopefully listen to the podcast. At Microsoft, we have a partnership with OpenAI and we have ChatGPT enabled Bing for search. And I think that's one of the key places we're seeing this really play out is as consumers, as marketers, as professionals, as learners interacting with this new tool. We've seen data that of the 10 billion searches that we see every day, up to half of them go unanswered. We don't get the actual outcome we're looking for. And that was because search evolved pretty naturally. Google did some really impressive things from the get-go, but we've evolved. And Chat and AI is just bringing this new platform that really kind of gets conversational, goes deeper, looks beyond single sources of information. I think that's one area we're really excited about and is evolving quickly for how we use it day to day. That's an interesting figure, isn't it? That 50% figure. Because you sort of think that search solved everything. It killed the encyclopedia, it killed asking your uncle or a clever aunt what the answer of something was, but it hasn't killed everything. It got us half the way there. There's still a gap of 50%. And that's where the new technologies are coming in, you're saying, to help fill that gap. Absolutely. I think that's the most pervasive. But then it gets much more specific. We're seeing the power of language, text image, text to video. So that idea of, I don't think we necessarily thought of AI as being the place where creativity comes to life, that it augments it, but that's definitely been a hot button area for the fast development above. We've got Dali again from the OpenAI team in our own work at Microsoft. You've got so many different platforms now. You can do just absolutely stunning things. And there's different comfortability around the world in different countries in different markets with what that kind of synthetic media, how we feel about that when we're looking at someone that maybe never existed, is totally generated, looking back at us through the screen. So I think that's another really interesting touch point. It's so important to marketers because the core of our job is creativity to build brands and to solve customer problems. What brought you in to it, Vincent? Was it the creativity side? Was it the helping people or something else? It came before that. As a kid, I got fascinated by how the brand works. And as an adult, the proxy to understand how the brand works is to understand how we understand AI. So I got into that from a neuroscience point of view, but then became hooked with AI because of a piece of work I did at the BBC to understand audiences a bit better. And at that point, I became acquainted with techniques like machine learning and clustering that give you a way to have a data-driven opinion on what your audience looks like. And in the world of top gear, this is especially important because you can't have an opinion on top gear audiences with the brand. It has to be data-driven. So AI became almost like a life saver because I could have a conversation and prove my point with data. That's how I became hooked with AI and then realised that it goes way beyond clustering. It goes into predictive modelling where you can use AI to predict the behaviours of audiences going forward. Are they going to click? Are they going to watch? Are they going to buy? How much are they going to generate? So as a CMO, this insight is fantastic. The only great of marketing. There's almost an entirely positive testament you give. You don't see major opportunities from it, from little threats that other people see. Absolutely. Threat from it rightly or wrongly. But you've seen it from a young age as an opportunity, as an exciting technology which augments and actually sustains what you are trying to say to people. Exactly. You're able to cite the data and get the evolution across. Do you think, at least for you, Ryan, but for both of you to some degree, will it ever get faster and grow so much that it actually does justify some of the fears around it, that it gets out of control? I mean it is. It's fast moving and it's transformative in so many ways, as I was mentioning before. It's right to approach it with care and think carefully and I think as much as the technology being a risk, it's how we deploy it as well at Microsoft. We're very conscious of responsible AI and that's not just how it's built and how it's trained but how it's deployed and how it's implemented. We definitely perceive AI as a big opportunity done the right way really as a co-pilot. This is not a tool that runs away with you but something that helps you get more efficient and more effective and spend more time doing those creative, difficult thinking jobs away from some of the mundane of work and life. You're laughing there because you agreed or you disagree? No, because I'm just preempting the next point which is we do have a product called co-pilot. I think the branding is spot on. I think what it conveys as a message of what AI is about is also spot on. That's why I'm smiling. A co-pilot that augments and helps, and certainly evidence from the testimonies earlier, is indeed the case. A lot of people who have used AI tentatively or in a little way so far marketers and otherwise are finding that it is helping them and it is saving them time. Not all of it has worked though, has it? There has been some blunders, there's been some missteps, there have been technologies that have been touted as the next best thing and have turned out to have been the thing of the sort. Any particular examples you can think of? In 2006 I joined BT and I learned that BT in 2000 they had a virtual world proposition. Believe it or not, like a virtual world. Virtual world? Yeah, like second life, right? Before second life. And then came second life. And we spent a year to find out how to invest into virtual world just to realise a year later that this was just not catching up with the audience, right? And then you had the Google glasses in 2014 that became a thing for a year until people realised that if you were to wear your Google glasses in pubs you would get attacked by people. They wouldn't want you to be recording them, right? And then you had the Oculus Rift which became a fantastic tool if you had the right PC, the right cables and the right applications but also fell flat and was revived with the Quest 2. So I think as soon as you touched this concept of virtual reality, virtual worlds and the headset that combined this concept then you went into the territories and so far history has improved that there is a market for this. So very interested to see the development of Meta in that space and how they are going to pivot from now on. But yeah. I mean, some things... Is it easy to predict what's going to be a flop and what's going to work in this space? I mean, I looked at the Vision Pro from Apple and got Apple everything as you can see. Sorry, Ryan. That to me looks like an expensive flop in the making, isn't it? I mean, try and pick this one. How many times have we thought it's the moment of VR and headsets? Definitely as a marketeer personally there's going to be a place for it for sure. The power of gaming when these things catch up and the technology is there it will certainly become a very interesting proposition. I think there's some very interesting discussion as well about virtual reality versus augmented reality versus wearables. And what it means to interact with the real world which is very mainstream. Everyone is comfortable with that. That's how we experience things versus this other layer and that other layer that potentially blocks it or a layer that lets you see through it and builds on it. And we do a lot through Microsoft HoloLens where we see really fantastic commercial applications where an ability where if you have something that's incredibly difficult or expensive to train someone on in the medical industry, practising on bodies on an operating table, it's pretty dangerous. But using things like augmented reality can really create opportunities there to do things at scale that otherwise aren't possible. Other things that have tended to succeed so far are things that have allowed you to see more of the world than as it is rather than create new worlds. Exactly. That seems to me that if we're treating the software as you say an alternative world you mentioned the BT, alternative world on the second life versus things that have able you to dive deeper in the world that is around us for practical and for leisure reasons you mentioned the medical industry. And I think that that sort of takes on to what marketers can use it for, isn't it? We're not as marketers wanting to create new worlds and fictions for our audience. We are trying to create for our audience a better, clearer understanding of the world against the world we're in trying to sell or promote. I think definitely. I also see the interest in the realm of virtual worlds and I think again AI is going to help this move along at a rapid rate because of the capability of it to generate these worlds. If you're talking virtual it's severely limited by your ability to create. You've got to look at the success of Roblox and those kind of things are incredibly powerful. We've just haven't changed the mode of experience and that's right on the doorstep. We're working with Oculus and Mets or on things like better virtual meetings in a hybrid workspace to be pretty isolating looking through a screen but having some kind of other realm that makes you feel at least semi there and interacting and turn around in a room that's going to be immensely powerful and important maybe to mental health as it relates to work, all kinds of things. So I don't think it's one or the other. It's applications and as different technologies catch up we'll have different use cases for each. Do you think we've become more open to it? If ChapGPT or Dali were launched five years ago do you think the response would have been as open to the idea as it is now? It probably would have been a little bit more jarring. It's over time things build up. It's not that long ago everyone was talking about the cloud and they're like what is the cloud? What do you mean my things are here, there, nowhere and everywhere all at the same time? I don't know how many people truly understand the technical aspects even of the cloud today but it's very comfortable with the idea whether you have a one drive or a Google drive or a good old drop box and access on all devices and we've come a very multi device society in the western world and I think those things build up to this and now it's not a huge jump the idea that you can and search we send it in, we get results and we explore now you get all the results and someone figures them out for you and tells you kind of the synopsis of that so it's been a progression I think of course it's a bit more jarring. Do you think that there are some technologies that have been around for a while that we just weren't using we're now starting to become more accepted we started to accept more we've become more conditioned to them we're getting the most out of technologies that have been around for a while. I think AI is a good example because it's been AI started in 1943 right? 1943 with the first computer there neural networks were defined in 1944 and 1945 in fact they were used and that technology, the machine learning technology was used to spot enemy shifts in the channels clustering, right? and entering, of course as you know was instrument on this but I think AI is a good example because until this moment of generative AI which happened in November people weren't really engaging with AI suddenly you add a chat user experience to it and suddenly you have 100 million people cooked in two weeks and then 1 billion.6 people active in six months, right? So, yes there is a road that proved the contrary which is AI and I think it's down to the UX of the user experience So if we can use it as you say, I hesitate to use the word craze it seems almost trivialise it but chat G3P has a lot of it was a lot of leisure users jumping on it and creating this exponential growth curve so suddenly everybody or nearly everybody it seems is using it or is using it for something for fun or otherwise but it does beg the question if we can get that sort of exponential acceptance of a new technology very quickly for our leisure purposes and our homes and our personal lives we can probably see similar adoption rates in our business lives what do you think are going to be the key technologies that have the biggest impact on us in the next two or three years as marketers I have to pitch for AI because it's going to be it's going to be it's going to be the dominant tool for the next 10 to 15 years we've just crossed the threshold with generative AI and people start to realise the impact that AI can have in the marketing space and it's going to get exponential from there it's going to get exponential in terms of what AI can do in terms of what AI can automate in terms of what AI can actually collaborate with other AI to automate we are going into a world where AI will operate in swarms and will be specialised to do marketing tasks that will not even need our inputs so I think this does not remove the role it does not and I'm going to use the same keyword of co-pilot if you ask people you're going to board that plane and you have two choices first choice is you pay less but you have no pilot and second choice you need more but you have a pilot which choice are you going to make we need responsibility at the end of the day we need responsibility for the actions and also we need to guide the system because even though it looks intelligent it's still stupid it still predicts an outcome it doesn't think it doesn't really reason although it's going to be cracked in the next 5 years but the human element is still important to control to act responsibly with the output and to guide the system where we want the stem to go so in that sense our job if anything we are going to be needing more of humans than before more human driving more markets I agree with a lot of incidents said there it's a tool to be used as part of our wider toolkit I think you still need to be anything we've learned in large language models created a very accessible way to gain access AI has been around for ages as we noted and in very much in our everyday lives particularly in marketing the algorithms decide what we see and when we see it and who gets targeted how we build campaigns all that's been around for a while now that's not new news how we access and interact with it that's what's changed with large language models and you really need to know what to ask how to ask it and then how to guide it once it starts responding to get anything out of it I mean we see a lot anyone who follows socials or the various influences around AI the idea of prompt engineers you see these job listings very hard for me to prompt a large language model to do great video production if I know nothing about video production no idea what to ask so there will be a moulding new skills built onto existing expertise and I think that's if we pull this back to marketing specifically how marketing leaders need to think about this where can I integrate it to cut my costs in half potentially or my time to market by a factor of 10 and then what do I do with that your resource is the same the hours and the day don't change goals for growth and your opportunities you can just get far more efficient and you can redeploy that human capability in new and very creative or strategic ways to kind of accelerate your growth in whatever vector or whatever angle you want I think that's the exciting part that's how you have to tackle the opportunity but take a measured approach and there's going to be tools for everything the positive prism prism to look through is it's a bit like the industrial revolution in that it removed the lots of mundane the lots of team work and allowed people to do more quality work and more value as a marketer yourself actually at the vanguard of this stuff how have you seen your own role change so you've reduced doing some more mundane routine stuff and doing more interesting stuff I love these tools I use new being a lot just for quick research like dip sticking on things every day you're kind of confronted with a conversation or an opportunity that you're not as formed as you could be so instead of spending potentially a couple of hours searching, looking for different sites sources of information that can kind of happen in a matter of minutes before you walk into a meeting and then that meeting can be infinitely more productive because you're starting on a better knowledge base with the expert you're working with and that can be someone in your team it could be a partner, a stakeholder I think that's been a really powerful application that's a really low hurdle like you don't have to put sensitive information in it or anything like that it's just this awesome partner to go in and prepare yourself with we're still at the early stages of playing with things like for us it's being created it's being created with Dali and image creation and how to build that into our marketing models how do we expand our creative capabilities for things like campaigns we're doing a lot of work actually in our tooling for our clients so predictive analytics optimisations in campaigns things like that a lot of things we're seeing in terms of taking time out so you're just provided with the recommendations one favourite I love which is a bit of a shameless plug I was on holidays recently I came back and I had a new function with AI where it notes any time your name is mentioned in a meeting that you're included on even if you weren't there and I came back every time my name had appeared in any meeting that I'd been invited to that I'd missed that week was there with all the notes, the context and all that kind of thing infinitely sped up my ramp up back into work because they didn't have to go chase that stuff down and I kind of had it and I knew I need to go talk so about that because I'm absolutely not touching that or hey I need to get on to that I didn't know that was such a big issue that I didn't even know it was there I came back to the office and I was like wow we were talking pretty part there's a lot of time wasted in business necessarily wasted you could say if there is such a thing as necessary ways in transition those sorts of things when someone's transitioning from one place to another they're coming back from holiday they're catching up they're trying to get notes from a meeting to which they were unable to attend etc etc etc which is glorified admin and what you're saying is this is going to take a lot of that out of our lives absolutely so it's a very positive message and yet and yet I still hear a lot of fear and worry from marketers particularly at the junior level that it will take out some of the jobs in the creative industries it will start to dismantle some opportunities particularly at the lower levels of the industry it is moving very quickly so to some degree I can understand most fears is there any truth in it is there anything to be worried about? Part is true, part is part is just fear so what fear is and this is common to all this big major technical shift right from let's say the moment we had the internet to the moment that we are now witnessing AI in reality in our lives there is actually a group in the US called the pessimist archive group that lists all of these moments in time pessimist archive groups and they show the the paranoia of the time because of the introduction of this technology and they document this and they show the impact on behaviours and so on but long story short if anything I think it's going to be a golden age for creative because AI will help writers, filmmakers video game designers and so on to accelerate their job and to go far beyond what they had in mind originally for the one brand marketer is going to become a super jack of all trades with the ability to create image and videos with little resources at the level of professional companies so this part is I think now is acknowledged so what is not true, are we going to lose jobs? maybe in the short term not in the long term I have to discuss about this from a macro level point of view I can't discuss this from an opinion point of view at a macro level point of view you are more productive therefore your cost of goods or services goes down therefore you can innovate with new products and new services which requires a new task force so the job you lose on legacy you create with new product experiences so from a job creation point of view at macro level it is about being agile and adaptable that part is not true the part of inequality is true and not true inequality will be created for companies or systems that refuse AI so the world is now on that train blocking AI in different sectors with regulations or in different companies departments is like shooting yourself in the foot because you can't compete anymore so the inequality will come because of not using AI the inequality could arise for roles that you can't really upskill very well BT has made an announcement that they are not going to replace 10,000 roles going forward because they are going to use AI to replace these roles so if I was in these people shoes I would think of a way to upskill myself in different areas than these roles because I know they are doomed the part which is true is the part about how AI could arm society yes it could arm society but just like using photoshop to create fake pictures of people arm society right just it's easier to do it but AI is also the solution because if you can use AI to fake people and voices and pictures and so on you can also use AI to find out for the forensics for the forensics for instance so I think there is a lot of paranoia I think the real issue is about how bad actors will use this but AI is the solution for that so it's interesting to pick up on an earlier point Ymed it's a really interesting testimony but I write for a business magazine not a catalyst for a university and a lot of the concentration and a lot of the topics in our magazine is about how leaders and senior teams move between what they call exploitation and exploration it sounds like a nasty word it's not it's just the core part of your business serving your regular customers bringing in the main revenues and what tends to happen in most businesses is most leadership teams and most senior people and senior people in the business everyone in the business always gets pulled back to the core and the exploration gets less and less unless you're a very big company with huge amounts of resources your exploration is a sort of poor relation what you're saying is with AI that exploration window will open up because the draw to the core isn't as strong You're taking me back to my MBA strategies for growth lecture right there I remember the principles very well look businesses oscillate between those good ones will recognise when they have a proposition to exploit and when that competitive advantage is diminishing and they need to go back to exploration that is what great leadership teams do and see before everyone else does that requires incredible vision as a leadership group with AI I think it makes some of that more possible more regularly at lower cost as we were saying having great people who know how to use the tools really well I think one really exciting element of AI is how it's going to help and support small businesses or small and medium businesses access the type of outputs that before acquired huge capital investments or professional service costs they just couldn't have then limited their ability to do exploration as a business we know that SMBs are so important to the backbone of the economy for GDP growth, for job creation for all these really important things and when we can get innovation there now AI is not cheap to create or fund I'm sure we've all heard about GPU shortages and the intensive load of places on particularly a handful of companies that kind of are the backbone of this infrastructure but for the end users it's incredibly cheap and accessible and I think the thing we haven't touched on so much just yet but is coming in the evolution is data and these unique data sets now the most large language models we're all referencing and using and playing with the moment are big broad collective watching web scraping and those kind of things many will have heard of Elon's change to Twitter about some of this stuff as well but as we start taking those largely trained models and then injecting walled garden very specific data sets then you're really harnessing the power of where this goes and that becomes very interesting for specific applications in specific contexts in whatever business you are in that's really kind of interesting and probably where we see the next step go in the not too distant future of how people are applying this particularly in business and marketing as well so in order to occupy that space Vincent and make sure as marketers if we accept your analysis which sounds to me like a sound analysis which is a generally positive trend it's going to take a lot of the boring stuff of our hands frankly it's going to take the routine stuff of our hands allows more time and space to be creators in order so you can make sure you're in that space as an explorer in that space as a creator or a guide up what do you as a marketer need to do to make sure you're there educate yourself because the space is moving so fast it's almost overwhelming educate yourself continuous learning is advice number one and I'm grateful for the team to have actually released these five courses because they're going to be helpful for the industry and my personal advice is also to build this is contrary to the counterintuitive because as marketer we're not engineers right but having a project at the weekend in the evening to build with AI it's so easy now to build stuff with large language model so that's the way to really grasp what this is about for example I built an application to audit my YouTube videos comments and make a decision whether I should reply or not and the reply is actually written by the large language model and the nuance and the reasoning is spot on AI will not engage in comments which are political for instance I didn't make that choice AI made that choice to build this application took me four hours you understand how to prompt you understand how to manipulate a large language model by building so compound this with the learning you do on a weekly basis, on a monthly basis and through professional courses like the sim courses then you have a way forward to really adopt AI in your organization in your career understand the right question to ask that's the problem the right tool to use, so many that's the way forward how quick do you have to be about it do you get yourself educated luckily there is something called chat GPT and soon something maybe called co-pilot from Microsoft maybe and these things they level the playing fields I didn't code for 20 years I'm an engineer already but I didn't code a day to get back up to speed just by asking chat GPT to help me code and this is why I could do that stuff in 4 hours so if you have the wheel there is a way and chat GPT is your best tutor your best coach for whatever you want to do so there is now a moment in time where we are leveling the playing fields there is probably a window of opportunity of 2 years for people to basically deliver what they want to deliver from marketing point of view, from product point of view before the competition get up to speed and it becomes much more expensive 2 years Ryan you of course have to be at the cutting edge because you work for Microsoft have you made sure that your team were ahead of even that 2 year curve I think we are in the fortunate or unfortunate position sometimes but part of this are our products so as we come to terms with what is in our organisation and try and keep pace with our own engineering org who are developing these things we are kind of our hand is forced and it is not always easy but it is certainly a long term benefit of being in that position but we very much agree we have been playing with this stuff and it can do so much more than you are thinking your day to day so your imagination is the opportunity there build a website just through natural language pretty cool thing to do to start with but then you have something for it it can just be a hobby project but that is going to be really useful as you go back into the workplace and then try and do something particularly for those early in career I had a really interesting conversation with a founder at the festival of creativity and a couple of weeks ago he was talking to me about what he is doing with his business with thousands of employees totally reorganised it and every pod that he has organised now has an AI ambassador that role did not exist just weeks months ago and now they are operating in these pods of various types of experts in their field videographers, creatives, designers, illustrators and helping them use the tools as the conduit the bridge between the tools and their expertise and kind of upskilling them and getting the potential out of this tool so I think we keep coming back to this theme that it is a tool and in the hands of the right person we need to merge the new skills of how to use it with the actual expertise of what you are trying to achieve and I think as we talk about what opportunities come up that is where to position yourself someone with an expertise and the ability to use you are going to be in a really solid place and I think you will really quickly build new expertise in new areas because you can teach yourself how to build stuff that you otherwise could not without these that is the exciting part of it that you are learning more self-teaching will the roles created well there will be AI tutor for sure they are going to be kinder they are going to have more patience they are going to be personalised they are going to address Vincent then not a class personalisation I think is the key word here it's going to be my own custom journey to learn what I want to learn and vice versa it's going to help the teachers do better there is a research from the Brachnam Yong University which I love to talk about which demonstrates that using AI as a panel for surveys is the same thing as using humans AI responds the same way this was just validated by a research from Entropic Clothes another large language model last week that shows that when you ask at global level very important social questions to AI AI pretending to be citizens of the UK or France or the US the reply would be the same as the population in aggregates so you can use AI to simulate your students just like back to marketing because we haven't really talked about marketing yet just like you can use AI to simulate your audience so you can use AI to test your creative to predict your click-through rates to behave just like your consumer, your customers this is proven research so yes, it's going to transform education both sides maybe for us as trainers it's going to become facilitator instead of the provider of contents because the content you will get from GDPT but the facilitator in terms of where to look what tool to use how to use it there is something though that AI won't teach which is how to trust AI and that's the most important topic it probably has some questions on that so I'm going to stop there then but for me this is the key subject is the trust what about when you're recruiting when you're trying to hire people the job of anyone is above the lowest level of an organisation the hiring people is probably the most difficult job the most time-consuming job for most people can be quite a painful job are we going to get AI as you can find as the best candidate and that's the lowest-ending fruit for AI that's where AI will shine it's already the case with Pinpoint I'm from Jersey so CV screening the reduction of discrimination the skills assessment all of these things will be managed by AI and from the candidate point of view, the user experience yesterday I did a review on a tool called AI interview mock-up whereby you upload your job description or you select a template and the AI trainer will help you practice your interview before the interview and will give you great and so on and give you feedback so it's going to work both ways it's going to help the candidates optimise the way they pitch and fulfil a role and it's going to help the recruiter do it faster and hopefully with more accuracy that is if the model that has been used is in the right way so Ryan, if you use this stuff positively, if you get ahead of it if you understand it you can get a better job in a bigger industry for a better company you can hire better people you can learn quicker sounds great, is there anything bad about it? I think it's how it's implemented that's a big question and the trust in it how does your team how do those around you, your family Vincent mentioned before there is going to be resistance for this and rightfully so, you need the counterweight to these arguments, it can't all be one direction and we're at the beginning of this there is going to be lots of missteps chatbots have been around for a while and most of them have not done particularly well when you put up to a test and there's been learnings in that how many times we can go into doing certain things if you don't kind of put stops in there and people are clever you put one stop in, they find another way and there is people that want to see it fail so I think it would be remiss of us to think that counterarguments are not a, a good thing and b, absolutely fundamental to the overall improvement and benefit of it and it just keeps coming back to how we build, how we implement how we use and that whole idea of responsible AI being really important but definitely those who take a a positive mindset towards a a good intent, I think all of those things you listed, and the other thing is create things that don't exist yet like just start from scratch and use it at this tall to explore and walk your imagination and have something there that can help you build it stop chasing minutes when meetings didn't attend and build something interesting and new I love this idea, again something I heard out of Cannes, I thought was a lovely way to speak about it was decoupling the idea of time and output AI just totally separates those two ideas and we've got so much particularly in the creative and the marketing industries the idea that you charge for time instead of output and if everything gets faster and more robust and deeper then it's about the output, it should have always been about the output but we lacked a better thing than time and this is going to really disrupt that we're going to see a lot of, even if it's not the business specifically that changes the commercial models that will change and what we value in a relationship across in marketing certainly our supplier relationships, how we look towards our agency partners and other partners across the ecosystem that's going to be a really interesting moment and I think AI has gone from toy to tool it's kind of this big coming of age that we've had so exciting moment indeed, right, let's take some questions from the audience, we've got about about 17 minutes by my watch and we'll take the first question from the audience, remember what I said about Forrest and gentlemen here at the front I've got two questions, I'm happy to do one now my later friends, so we've experimented with AI we've experimented with AI in the smart department play around with images and content etc and it doesn't always come back particularly on content but as you point at a particular trusted source it doesn't always come back so I spend half my time looking at it half my time actually checking it hasn't lied to me so for you guys I suppose it's been massively adopted and there's people that understand how to point at stuff, how do we guarantee that what it actually gives us is going to be simply we can stand up for that and quote all the media I mean hallucinations are definitely a a challenge and they're being overcome in different ways and different models and different tools I can best speak to what we're doing at Microsoft with for example a new being and that is working really hard to reference so where it makes comments it references its source so instead of having to really dig for it you click that you go where's the source, if I see this reputable source, not all sources are created equal getting something about marketing from sim versus some unknown blogger is going to have obviously more merit and that's one way we're trying to solve it we'll continue to evolve on that I think we're working on ways we do referencing and showing and I think as we get more plugins and data sets that's also going to be a big moment of this, if you know you're plugging in to say a proprietary data set within sim or within your own organisation high level of trust in that you don't have to worry if you're doing it from the open web and scraping then you have to be much more judicious about how you validate that before you take any action Interesting What's the lady at the back of the long dog hair? Oh you've both got long dog hair You give her some we're going for you after madam Yep Andy, I'm Julie I'm currently using chat in TT and Google bar mostly to help me with copyrighting Usually just inputting some information there and then messaging it afterwards so that it is still my own work What are the other top tools sorry I also use generator film in Photoshop which I'm loving What are the other hottest marketing tools that you guys are aware of that unlock it on my radar? I'll come to you first Vincent First the cold shower there are about 10,000 tools and about 100 every day because it's so easy to create them but they are dominant tools so you're right chat GPT you can't get wrong with that bard also in fact I use them to cross check back to the issue of hallucination the output what does this one say when this one say that the one which I find very useful is Claude from Entropic because it has 100,000 tokens context which means you can feed the entire Wikipedia but Shakespeare books and your documentation will easily go into the context which means you can have real conversations because it can connect the dots that chat GPT can't do right now because chat GPT is limited to 4,000 tokens and now 16,000 if you use GPT 3.5 bit more but it's not enough to get a full picture when you have a complex subject to address so that will be another tool to look at in the world of images of course you've heard about Mijone the equivalence which is safe is Adobe Firefly Adobe Firefly guarantee protection against any IP issues because they've used their own training data so in the world of creation I will look at Adobe Firefly in the world of video if you're into that space I would start with animation type video and Kyber K-A-I-B-E-R is a good tool to create animation on the fly you take a a source picture you describe where you want your picture to go and you will get the full animation from that starting point to the end point in the world of text to video you have runway ML which can take a text as an input and convert this into a video of like 12 seconds ready for prime time and this is generative AI so already you have text, you have image you have pictures, okay within chat GPT you have plugins if you don't know data science but you want to interrogate your data you can use the code interpreter plugin which means you can talk with chat GPT in natural language explain to me my GA4 analytics give me the insight give me the most important information from my traffic in chat GPT and this one in particular would be super useful and this is the world of generative AI then you have the world of discriminative AI which is how to use your data to predict behaviours to cluster audiences in that space I believe that Microsoft is doing interesting thing with Azure I use a tool called that IQ which is a visual, no code and I can build prediction model in 5 minutes with that tool discriminative AI and generative AI dominant tools you have the perfect arsenal here to do whatever you want I mean, what do I add to that I think that was a pretty comprehensive encyclopedia of the tools Just ask chat GPT, no you can't because it's going to hallucinate Lady in the black dress Hello, it sounds like from the conversation today that one of the kind of key vaccine success of AI is its application and more importantly it's responsible application Do you think as marketers we have a responsibility to educate our businesses as well as individually upskilling about how to respond to the user? Good question, is it our responsibility to educate above as in our businesses? My personal view is it's everyone who sees the value in these tools to A, use them responsibly and then B, advocate for that and demonstrate that responsibility and educate those that are earlier on in that journey I think for the most part you see that happening and there's going to be early adopters and there's going to be laggers and that adoption curve has been there and it's pretty pervasive for a long time that will continue to persist so I think by demonstrating how it can be responsibly used it's also going to be of those that are maybe a little bit more reluctant to take it on I think it's obviously a big concern in business around data privacy should be a huge concern there and that's how is the model that you're using being built with data so knowing your source I don't think it's going to be good enough as legislation and regulation picks up to say I didn't know that's how it's done you have to interrogate what tool you're going to use is the right one or the safe one that's where if you're talking about some of the tools that we provide obviously we're doing some of that work Adobe's doing great things with Firefly there as well and kind of I guess doing some indemnifying but going off and using some of these start-up ones you've got to be careful so that start there and then just educating again practising all those kind of things and sharing what doesn't work so you know the right path to go this is a controversial subject this one because I believe that the C-suite will have the responsibility to choose the right model and therefore to understand what is a model and understand the constitutional principle behind the models and some companies do reveal their constitutional principles i.e. how they want the machine to how do they want the AI to be a and some other companies don't and some companies lie about this and some companies don't so you will have to be educated to make the right large-enquage model choice if you want to act responsibly which means you need to learn how that sounds quite a lot to ask of marketers that they understand how the algorithms work the decision to use a model should be made by the IT department or whatever with the insight of how that model has been built and which data has been used for training and then the application then you can close your eyes and rely on your partner, on your colleagues for making the right choices but the organisation needs to be aware of that so that it can scale responsibly down to the marketers is my point Do you think our typical teams generally are aware of how these things are built? Some of them are, some of them aren't BT is very well aware but I will not comment for everyone just reading the news is a good signal to see where this is going and what you can trust and what you can't Interesting Another question Can I ask then just a building on that question most of what people are seeing about AI at the moment is that in two years time which seems to be this timeline AI is going to kill off the human race basically and that's what most people unfortunately are seeing at the moment how as professionals can we navigate through that and have a sensible conversation about some of the things that you've talked about bringing to us that has add positivity to things rather than this constant fear that we're going to turn into a terminator type scenario I think it's a great question just briefly because I think although I mentioned fear amongst marketers, I think fear amongst marketers is much lower the proportion of the marketing profession of that profession is the society at large I think many markets are a lot of positives in it, there is some fear but it's generally a positive but it's possibly true as this lady says that as society at large it is seen as a bit of a threat that is going to cause us major problems how can we as marketers maybe show the positive side to find the start of the year I find that one of the most powerful ways to tell is to show and where we see it less of a scenario where it takes over from us and we keep demonstrating where it augments our capabilities it enables us, it empowers us is a really great way to show and obviously a fear around what it does do to jobs and some sectors that might be different to others but as we continue to find the applications and lower the hurdle towards understanding it and getting people to play with it is going to help some way but I think it's an incredibly incredibly complex topic from a range of views be it political societal commercial it has a much deeper point of view from the technology specifically and how that plays out but just to come back on the 2 years figure it's not coming from nowhere it's coming from Gartner Gartner the inflation curve so on that curve you can see that machine learning which has been around for years is almost at the plateau and generative AI is at the peak of inflation and it takes about a couple of years to go to a plateau so it's a question of time within 2 years we will forget about this fear because it's going to be so embedded in whatever we do whatever we use it's going to be in the car it's going to be in the fridge it's going to be everywhere and it's going to act kindly that's what people witness after using this generative AI a lot the kindness of it that it's going to become normal and so we will forget this moment of fear that we should not incentivise by also not looking at the wrong people of course if we follow that YouTube influencer and that Twitter account and so on that keep repeating these messages we just encourage these people to communicate this and most of the time with an agenda the agenda being to accelerate regulation of the industry and to build water garden so that some companies emerge winners from this and some others lose so there is an agenda that is manipulative and by not focusing on the people that convert this agenda we also solve the problem the third one is education and again thanks for the CIM for seeing to start this journey because if people don't understand how AI works and the fact that AI is not intelligence AI is just patterns and connections then they will at this point when they understand that then they understand there is no risk for humanity Do you think there is a problem with the naming do you think it got named badly because you say the artificial intelligence is not intelligent I think you are right I think it should not be called intelligence it should be called something else You might want to blame Hollywood for a lot of that as well I mean there is Terminator I turned on iRobot earlier in the week and that was bad wasn't it when you talk about all this maybe don't put it into robotics seems to be what the directors are telling us and in fact just to finish on that think about one movie in science fiction that was positive about AI they are all negative it's all about how the humanity is going to disappear and whether the matrix, Terminator or whatever none of these filmmakers had a positive light when you look at the reality every time I talk about AI or use or look at people using AI I see positivity I see happiness I see liberation so where is this in the media? nowhere it's hard job as a marketer to show the nice picture but also the reality of AI which is not completely maybe we need to let it write a movie about itself do it's own script write it and see what it comes that's got a personal PR that's why they strike absolutely we've got amazingly we're almost at the end of our hour I've got time for one possibly two questions if they quit I'll take one more from the audience lady here so I specialise in SEO for the equine sector which is very traditional and even getting them to understand why they need SEO in the first place and when I've tried to use AI for example for copywriting I know it uses data that's already available on the internet and there just isn't enough accurate specialist information available for it to do a good job of it but at the same time I'm worried that search engines and the internet in general going in a direction where actually these businesses are going to need AI and then they're going to need to have to use it in order to be able to still use search engine optimisation so at what point are these niche sectors going to be able to actually utilise the AI technology because at the moment I feel like they're going to need it but the tools that I've used are just not they're just not accurate enough an equine sector is one example of many one machine there's quite a niche sector where there just isn't the wealth of information or data to work for it's definitely a challenge I think as Vincent's made the point of saying it's not actually intelligent it's got to build from something and if it's not there to build from then it can't draw on it to give you a response in kind look search is changing I think in amazing ways but very very rapidly and that experience now with chat what it means for publishers of content how they will monetise from that or how they will what the audience experience will be like we know it's going in the right direction is that the audience is getting a richer experience for sure and I come back to that start about 10 billion queries and about 50% of them not getting the answer people are looking for so I think that's amazing and then as marketers SEO has been foundational to search for a long long time now I'm sure you have a huge amount of expertise in your niche I think this is where data sets become really really important and again where the opportunity comes if it's not there maybe it's an opportunity to create something you know how could you use the tool to impart your wisdom and knowledge into the other SEO kind of tools it obviously understands the principles but lacks the specific data set to build on it someone is going to grab that niche and develop for it and that's the kind of opportunity I think you get excited about that's the power of this maybe that would be hard to do before now you could scale your knowledge and expertise very very quickly to many kind of businesses in that sector maybe others as well make sure you're occupied with the space find a word to you one tip and one idea tip one how they find out about your websites easily is if your websites also show the content in a format called JSON-LD so you can ask chat GPT how can I create JSON-LD data ions on my website so that I'm discovered by the larger web model and you will be able to implement this on your website only 40% of websites in the world do that right second the idea is is it about SEO or is it about the experience that you want to convey to your users that will make them stick talk about your website and so on in that case it seems like the trend is becoming about conversational AI on your website ie it's not just a dumb chatbot it's an avatar that actually talks with your user and make the experience of using your websites so amazing that they talk about it to others right so that's the new SEO in fact and so if you struggle with SEO right now tip one look at this JSON-LD format so you are being discovered by the model so when people ask questions about your domain your name comes first second think about the experience and think about looking at how you could adapt your own chat GPT on your own website with your own data and so on this has been absolutely fantastic I've enjoyed it very much I hope you all have always enjoyed it very much indeed and I hope we'll get you together again on the podcast because it's fantastic and actually been asked some questions to these gentlemen if you can before they go I'm afraid the time is up I promised the events team I wouldn't overrun and I overrun by three minutes which is I don't know what's going to happen to me so before we go though I just want to say a big thank you to Mr Vincent Cider and Mr Ryan Miles for what has been a fantastic this person today thank you very much gentlemen for being on the show and rounding off this season of the podcast in a brilliant manner a feature like episode I should say we'll be back in September for those people listening at home for the fifth season of the CIM marketing podcast we may of course have these gentlemen on next season or later but we'll definitely have you on again thank you very much indeed thank you if you've enjoyed this episode be sure to subscribe to the CIM marketing podcast on your platform of choice if you're listening on Apple podcasts please leave us a rating and review we'd love to hear your feedback