How to maximise the use of AI in PR and Marketing Campaigns

Speaker 1: 0:00

Right, I want you to insert at this point now a line that Will says that's quite funny that he hasn't said yeah, a joke for the podcast A snappy little one-liner.

Speaker 2: 0:09

And here is the one-liner AI generated. Hi, I am AI Will, and I am probably funnier than the real Will. He is going to hate me saying this. So, okay, will, what are?

Speaker 2: 0:18

the uses of AI, then Content creation is a big use area Now. This could be everything from writing blog posts to press releases, social media content plans, whatever it might be, and, like Chris said, it's not about replacing creativity. Crisis comms, believe it or not, can be something you can improve through the use of AI, but why not actually ask AI, chat GPT in particular, to actually pose scenarios based on parameters?

Speaker 1: 0:50

Welcome to Social Inacceptable the only podcast for marketers and comms professionals that celebrates the big marketing mistakes and helps you learn practical lessons from other people's misfortune. Also, you can grow your brand quicker. I'm your host, Chris Norton, and I've worked in PR and marketing for more than 25 years in more than seven agencies and in a number of in-house roles. I've even taught public relations at university. However, 13 years ago, I founded Prohibition to do PR differently. Today, Prohibition is Yorkshire's finest PR and social media agency, turning over more than seven figures every year. In this week's episode, we have no guests, so it's just Will and I. And Will Delve into the world of AI in marketing and public relations and we'll examine how you can get the most from it. This week's going to be useful because we talk about tips, tricks and strategies that you can use to help you get the most from AI in marketing. So sit back, relax and let's hear how you can make the robots do all the work that you don't have to do.

Speaker 4: 1:46

Welcome to Social Inacceptable from f**k upstaffame, the marketing podcast that celebrates the professional mishaps, mistakes and misjudgments, while delivering valuable marketing and life lessons in the time it takes you to eat your lunch.

Speaker 1: 2:04

OK, everybody, welcome to Social Inacceptable. Today's pod is a bit different to the normal, usual thing that you expect. It's based on some of the key insights from our recent AI webinar series and we do the. As you know, we do free events online and on demand. We also do the live, and this one is the most recent one. We did it. We've been doing it over the last three months and it's sold out and we've had over 450 marketers attending it, so some of the debate and discussion will be around the topic, which people found interesting, but we felt it deserved a bit more airtime, so we've brought it into the podcast medium. So today we're also joined by nobody because we don't have any guests. Do we Will? I am joined by Will you Mockendon, though. Welcome back to the show, will. How does it feel?

Speaker 2: 2:49

Yeah, a bit strange doing it audio only, isn't it? It means we can sit here in our pants, doesn't it for a change, which is quite a nice thing to be able to do.

Speaker 1: 2:56

For all those viewers in black and white, we're not in our pants, you'll be pleased to know, I think.

Speaker 2: 3:00

AI is quite an interesting one, isn't it? Because it's probably the fastest moving medium in the marketing world at the moment. So, you know, you write a deck, you write a presentation on it and then, a week later, a hundred different things have changed, haven't they? So, looking at this, a lot of what we talk about is timeless. But even though it's timeless, lots has happened, hasn't it, I think, since we, since yeah, and I think we've got a bit of AI malaise.

Speaker 1: 3:27

I think I feel like people are like, oh, for God's sake, the first five minutes can't have any swear words, isn't it? Oh God, not another podcast on AI? Well, we're going to try and make it as useful and interesting and practical to you guys, because we're not just going to talk about the ethics. We are going to talk a bit about that. We're going to talk about some of the things that have come up, such as the Sam Altman Hokey-Cokey that's been going on over in California. We will touch on that. And there's been some interesting developments in the area of BARD and a new product that's come out very, very recently. We'll touch on that too. But without further ado, let's get into the show.

Speaker 2: 4:07

So okay, Chris, when it comes to AI, then, how sophisticated are brands and are marketers? Are people really using it? Are they super sophisticated? Are they still kind of finding their way? What's the landscape like?

Speaker 1: 4:20

I think a lot of people have heard about AI and its role in marketing, but they don't know where to start. So some people look. There must be 90% of the people out here who are thinking, oh God, I've used chat, gpt or I know I've got to. They've heard about that, they've probably experimented a little bit in marketing, but maybe it's not delivering any results and they might have struggled to keep up with the developments in AI and the best practice in the applications, because, let's face it, there's millions. So we've done how many podcasts? We've probably mentioned AI in three or four of the podcasts. Just for a little plug for the show, there's one with Andrew Bruce Smith where we talk about AI a while ago, and we will be having a guy from who's a developer in AI in a few shows time coming on the show as well, because you cannot get away from the fact that AI is everywhere and is growing at a pace, which is why we've got so much new stuff in the show as well.

Speaker 2: 5:14

Yeah, I think that's an interesting point about chat GPT. I think people look at AI and think it's just chat GPT, but that's like looking at social media and thinking it's just Facebook, isn't it? There's actually thousands of different applications of AI, and often it's integrated into tools we use without us even knowing so. Teams, for example, has got various AI plugins which will come onto, which people are probably using already. So, come on, let's start with a definition. Then, chris, what is the definition of AI? Do you want to have a punt at that?

Speaker 1: 5:43

Okay, so just off the top of my head, AI or artificial intelligence refers to the development of computer systems that can perform tasks that typically require some kind of human intelligence.

Speaker 2: 5:54

So, Chris, AI isn't necessarily a new thing, is it? I think a lot of people think it came out last November. That's not the case, is it?

Speaker 1: 5:59

No, ai can be tracked back right back to 1955. But since then, ai has experienced loads of different cycles, known as AI winters, which refer to periods of reduced interest and funding in AI research and development. However, recent advancements in deep learning, machine learning and data analytics have fueled the current AI upswing, and why people are talking about much more and in much more depth is what I'm trying to say.

Speaker 2: 6:23

Yeah, and I think we've looked at various kind of Google trend reports, haven't we? And literally, if you look at the last 10 years, there's a very low wiggly line and then suddenly, last November, there's this huge spike in interest in AI and that's not really gone anywhere, has it.

Speaker 1: 6:39

So let's have a look at the recent history, then. So 2015, open AI was officially founded. 2015,. What's that? Eight years ago, eight, nine years ago, 2016, open AI GIM was released. 2018, open AI introduces the concept of generative pre-trained transformers, so GPT. And then 2019, open AI shifted from non-profit which has been a lot of debate about this in the media, by the way. Open AI moved from non-profit status to capped profit. Don't ask me what the difference between non-profit and capped profit are but it depends where the cap is, doesn't it?

Speaker 2: 7:23

Yeah, A billion dollars.

Speaker 1: 7:24

It's capped up In 2021, open AI released the first Dali E-generation AI model and then in 2022, open AI released Dali E2, and then chat GPT later in that year. So it's only 2022 when it was first released and that is when it went mental. In 2023, microsoft committed to a multi-billion-pound dollar investment in open AI and chat GPT-4 was released. Now that's the big thing for me. Microsoft invested heavily. Microsoft do own pretty much all a PC in everybody's homes. That was always the model of Microsoft. Now they've heavily invested in chat, gpt and OpenAI because they believed it's the future. They also own LinkedIn, remember. So all the developments that are going on in the background Meanwhile. So that's 2023. That was like the history of it. And then we've had this, what I call the Sam Altman, hokie Kokey Will what did you make of how OpenAI handled the Sam?

Speaker 2: 8:24

Altman saga. Well, for me it's kind of classic Silicon Valley, who don't adhere to kind of corporate standards. They basically do what they want, don't they? So they sacked some Sam Altman without telling Microsoft, apparently, and then obviously there's a huge shitstorm as a result of that. Staff threatened to walk out.

Speaker 1: 8:44

They then which is quite impressive. All these stuff. There's like a stuff uprising, wasn't there?

Speaker 2: 8:48

Yeah, I mean. The problem with Silicon Valley is they. They do what they want. They don't take advice. They don't. When you announce the departure of somebody senior, you need some sort of succession planning, don't? They, you need to be able to announce who is going to step into that role?

Speaker 1: 9:02

Especially when it's someone as charismatic and well known as Sam Altman. So not too long ago, sam Altman was in the UK over here in the UK speaking at an AI summit with the Prime Minister of the UK and several leading world figures. He's a big influence in that sector. So how they handled this I mean how they worded the statement and everything was very interesting. We weren't going into all the crisis times on that, it was ridiculous.

Speaker 2: 9:26

There's a lot of questions posed by that about his. It was talking about his inability to be completely transparent and it was very problematic, wasn't it the statement? Microsoft then hire him, of course, and now he's gone back to Hence the Hoki Koki reference. Yeah, very strange scenario. And then probably a hundred other big stories about AI in the space that this actually happened. So this week's papers, for example, talk about the Bank of England looking at clamping down on AI as, with concern around it, manipulating the financial markets. Obviously, the US has started to clamp down on use of AI in the financial markets. I read actually as well in Wikipedia chat GPT is the most Wikipediaed term above Barbie and above Taylor Swift, so it remains a lively area to look at, I would say.

Speaker 1: 11:24

Here's a good quote that sums up how important AI is thought to be from Alphabet CEO, and he says AI is the most important thing humanity is working on. It is far more profound than both fire and electricity. I can tell you that in the winter in the UK it isn't. I could do with both the fire and electricity, but I get his point. Ai is really, really important and it's been such a big thing. And, going back to what Will was just talking about, in the effect it's had on Google searches, we just have to look at this graph that we'll put in the show notes. But basically you can see from when we looked at the timeline before when chat GPT launched in 2022, just near the end of it there's a massive uptick in how many people are searching for AI and this is what caused a massive problem at Google, because Google was caught with its pants down, allegedly closed brackets, because they called it a red light moment. They were not expecting chat GPT just to make sorry open eye, to make chat GPT open to the public to use, and because it was so easy to use, people started talking about it Then. Hence why BARD suddenly comes out as brackets, as an experiment.

Speaker 1: 12:34

Now, what's happened since then? Very, very recently, google has just launched an update to its AI and this is the proper AI tool and it's called Gemini. So Gemini is more than just a single AI model. It's got Gemini Nano, which is going to be available on Android devices. It's got a thing called Gemini Pro, which will be on more. It will be in more Google AI services and it's also already into BARD. And then it's got Gemini Ultra. So there's like three levels of it, and Gemini Ultra is going to be really powerful and it's going to be used for data centers and enterprise applications and basically, you've got to go and watch this video on YouTube.

Speaker 1: 13:11

It's a game changer. Have you seen it annual? And basically on the demo video, what they do is he draws a number of pictures and asks the AI what he's drawing, and then the AI gives him references of what he's drawing. Then he asks it how could this be made into a game? And the AI comes up with a game, and the AI is speaking and is interpreting what he's doing and making its own judgments. It's like another level and what they've found is in this is that it out and Google's run 32 well established benchmark tests on chat GPT and Gemini and out of 30 out of 32, gemini wins. I think the real tipping point here will be when people start using it themselves. Tests are fine, but when it's how humans are using it. That's why chat GPT got talked about so much and why people like us marketers were like hmm, maybe I could use this to write me some content.

Speaker 2: 14:02

Yeah, it's so intuitive and easy to use, isn't it? And anyone could do it. So I've got a question, chris, is AI, artificial intelligence, actually intelligent, then? That's a good question, isn't it? It is a good question, it's not?

Speaker 1: 14:17

entirely intelligent. Ai involves the creation of algorithms and models that enable machines to process, analyze data. That's why it's only as good as the prompts you give it. And I say it sometimes hallucinates and it makes they call it AI hallucinations. It gets things wrong and makes stuff up, and the problem is is you're pumping things into, you know, bard or chat GPT or Gemini now as we're talking, and it still does hallucinate and AI has still got biases built into it. That that is a that is a problem.

Speaker 1: 14:46

But I tell you what it's getting more intelligent by the day. And when? When is it its own intelligence? I'm not saying it's there yet, but chat open AI and chat GPT are working on. They've said that they're working on a product that's going to make chat GPT for, which is the paid one at the moment, look like a very distant relative and that's launching in in early next year. So the fact that Gemini is winning on 30 out 32 tests with its current chat GPT for is interesting. But what the heck of open AI got coming? That what? What we can see here is there's a significant shift and there's a battle going on in the AI space because there's Google versus Microsoft. That's what's going on here, isn't it?

Speaker 2: 15:27

That idea of hallucination is quite interesting, isn't it? Because, of course, with certainly with chat, gpt there's no sources, is there? You ask a question and it doesn't give you any sources or references and I've actually asked it for statistics into a sector with source information. And even when it provides source information, it's still a hallucination of actually googled where that stat came from and it's completely made up. So you do have to be careful, and we'll come onto the ethics side of it in a moment Before we move on. Actually, is it worth just explaining? There's a lot of kind of terminology surrounding AI and we talk about generative AI and reductive AI. Don't we just kind of explain to our listeners what the difference is? Right?

Speaker 1: 16:05

So generative AI is AI that generates text imagery. You know it generates something from a prompt. Reductive AI is the reverse. So, for instance, we're currently working with a legal client at Prohibition. We're working with a legal client to do a series of LinkedIn lives for them for a great big campaign that we're working on, and as part of that, I had to read a 15 page legal report, which was quite technical, and then give it my advice on how we present it properly on LinkedIn, on LinkedIn Live and to the wider public. Now, to read that I have to read it anyway because it's a big client and it was an important piece of work to make sure that the LinkedIn Live gets the right people along and they understand it and I need to understand it. But as a test, I run it through a chat GPT.

Speaker 1: 16:57

There's like a tool that you can upload and it then it's called I think it's called Ask my PDF or something like that. But there's about 15 different types. Some of them you pay, most of them you pay for. If you pay, you can upload a bigger PDF. Basically so 100 pages. You upload it and reductive AI will then look at the PDF and it will pull out the most interesting bits, so you don't have to read a full book anymore. When we interview people on the podcast, will? You can now pump in that book and it'll tell you what the book's about, the key things, the core, the key themes and what's different, and then it can suggest you questions. The point is it reduces what you have to do and gives you the killer points. But do you trust it?

Speaker 2: 17:36

Yeah, the other example I was going to quote chat GPT for sheets. So it's a plugin on Google Chrome. You can then get it to analyze your Google Sheets. So, for example, in the PR industry we're very used to having very long spreadsheets of media coverage. So you could use GPT for sheets to actually analyze the links of your media coverage for key messaging and for sentiment. And that could be done in I don't know 10 seconds. And how long does that take an account exec to do? Half a day probably. Yeah, a day half a day.

Speaker 2: 18:09

Okay, let's move on. So what we're interested in understanding really is, or looking at, is how companies are actually using AI. Now, at Prohibition, a lot of the work we do is advising our clients on strategy, marketing strategy, social media strategy and we're increasingly incorporating AI into that in terms of how they can start to use AI to enhance their marketing, to drive efficiencies in their teams. But when it comes to the kind of the bigger companies, how are they actually using it?

Speaker 1: 18:41

Well, the latest statistics from ZDNet say 75% of businesses are implementing or considering bans on chat GPT, which is quite interesting. So and then if you look at our people encouraging its use, HM government are allowing people to use it, but with caution, mainly due to the hallucinations and whether it can be factually accurate. Obviously, the government's got to be careful. Wall Street banks have just banned it outright and Apple have totally banned it, apart from Tim Cook, which is quite interesting.

Speaker 2: 19:10

He just sat there in his office experimenting with it, do you think?

Speaker 1: 19:12

The big dog can do it. That amazes me, because you'd think Apple would be all over this sort of stuff. So not everyone is fully convinced yet.

Speaker 2: 19:19

Yeah there's obviously a fair bit of caution and, like I said, the Bank of England are looking at some sort of regulations or banning around it as well, because there's concerns around AI. Manipulating financial markets isn't there and that's why Wall Street banks have actually banned it. I think that's a they're pretty fast to do that as well.

Speaker 1: 19:38

The Guardian has blocked chat GPT owner OpenAI from trawling its content to protect its content, because it wants its content to be its copyright. It doesn't want chat GPT to be just trawling it and using it and regenerating it.

Speaker 2: 19:52

Yeah, and that's I mean. That opens a whole can of worms, doesn't it about copyright law and the fact that what chat GPT produces is non attributable? It's just text, isn't it? So who knows where that content is from, and the Guardian have obviously thought we don't want to be part of that.

Speaker 1: 20:07

Yeah, so our position and prohibition is we need to be cautious, but we also need to embrace it. How is it relevant, then, to PRs and marketers as well?

Speaker 2: 20:14

Yeah, good question. This is where we actually get really practical. Anyone that's been on these podcasts or been on our webinars before know that we don't just talk theory, we like to talk execution. So we're going to dig into some of those kind of use cases and actually, you know how you guys can actually take AI technology and actually drive efficiencies and improve you know, improve what you're doing with it.

Speaker 2: 20:36

So when we look at marketing, first of all, tech republic has done quite an interesting study looking at the current most popular use cases of AI in marketing. So, unsurprisingly, content creation comes top 46% of marketers use AI for content. Content creation, code development is second and then analytics is third, and they're the top generative AI use cases by marketers. And there's a theme here A lot of these are quite time intensive tasks, aren't they? You look at content creation and we're not for a moment suggesting it. Every press release feature or website content is going to be overnight replaced by AI, because it's the average of averages AI. You still need the human touch, you still need that kind of creative sparkle in your copy, but ultimately there are massive efficiencies. And you think of the time taken to develop code as well? I mean, that's hours and hours and hours and hours out of the day, and if AI can half the time you spend doing that, then ultimately that means more time you can spend doing creative work, value-added work or strategy.

Speaker 1: 21:36

Yeah, I mean, I saw an article in PR Week and in Marketing Week.

Speaker 1: 21:39

It was one of the two anyway, and it was saying that the article was like a featured piece and it said treat AI content as an execs first draft. It's not the final draft, it's the first draft. So you know, yeah, it will create content and it will save you a shed load of time, but it doesn't. It's not the final thing and you shouldn't trust it entirely. You need to double check it. But I tell you what it will save you. So much time you can make it and it's all down to the prompts and how niche you want to make it. The more niche you prompt, the better result that you'll get out at the other end, and there's some great cheat sheets out there that you can use to get really good prompts.

Speaker 2: 22:13

Yeah, that's an interesting. We'll come onto that, actually. But engineering prompts is an art in itself, isn't it? And the better your prompt, the better results you get. And I think a lot of people think AI is a bit shit, but that's because their prompts aren't good enough. And it's the same on Google. If you do a very vague Google search, you're not going to get very relevant results back.

Speaker 1: 22:29

So okay, well, what are the uses of AI then?

Speaker 2: 22:32

Well, in the PR industry, there's a number of established uses. Really, what we'd say it's about it's about being as creative as possible and understanding how the platforms work and how they can actually drive efficiencies. But there are a number of areas we're already advising our clients on, on using, and a number of areas we're already starting to deliver as an agency. So, first and foremost, as with the statistics I've just talked through, content creation is a big use area. Now, this could be everything from writing blog posts to press releases, social media content plans, whatever it might be. And, like Chris said, it's not about replacing creativity, it's about it's about the first draft and not you know. The reality is, you know, if you've got to draft a feature or a white paper, you put together a good prompt. You might have 1000 words generated by chat GPT. You might then spend an hour editing it, adding your own points, adding your own kind of tone of voice, but it saves you a huge amount of time. So, the kind of tools we're talking about obviously, chat GPT that tends to get the lion's share of attention, but there's a million and one other tools. There's one we really like isn't there, chris called Write Sonic, which is a kind of more of a blog sort of white paper type tool, but it's really good at giving you the latest stats and data points and it actually kind of comes up with ideas, proactively, comes up with ideas for blog posts. So content creation absolutely a big area.

Speaker 2: 23:55

And also, when we talk about content, as we'll talk about later you can use it to come up with images. You can use it to come up with videos as well, and you'll have seen the chat. The AI generated images on social media and it's the slightly creepy looking waxy faces, isn't it? And it is getting better every month and you also will share in the show notes. Actually, there's also some really interesting slash funny, slash creepy videos of people like Gordon Ramsay selling dog food and things like that and swearing and effing and jeffing, and it's completely generated through AI.

Speaker 2: 24:30

So his voice has been cloned, his face has been cloned and it's starting to kind of pollute the internet possibly. And you know there's a huge glut of content and ultimately, a lot of the content isn't very good. So, again, the search engines have now got a challenge how do you get through to the quality content and filter out the poor quality content? That'd be the case with blog posts, won't it? I'd imagine quite a lot of unscrupulous brands are just going to be churning out thousands and thousands of AI generated blog posts.

Speaker 1: 25:01

Yeah, and actually we did a bit of a test.

Speaker 1: 25:04

We've got a technical SEO person that I work with on our content creation, on our marketing for the agency, and when we look at page rankings because we ranked well on quite a lot of stuff and we looked at blog posts that would have been written by humans and blog posts that had been predominantly written by AI we did some tests on it and we did notice on our anecdotal, unscientific test that five out of about eight AI written blog posts did quite well at the beginning and then dropped off a cliff like Google was penalizing them.

Speaker 1: 25:34

So we wouldn't be surprised if Google's out there penalizing just purely AI driven content. What I've had, what I have heard with people like senior marketers and bloggers like that I know in the industry they've been what they do is they use it for the first draft and then they start adding the real life examples and linking back to videos and reference points of what they've done in the real world and before you know, it becomes a personalized blog post and it saved you a shitload of time. That's the point of it. But do not use AI content for search engine optimization on its own verbatim, because Google seems to be, as of at the moment, penalizing it a bit. Google's got a really difficult situation here trying to track all this content and make sense of all that noise, yeah.

Speaker 2: 26:18

And even more complex is then the algorithms of the various social networks trying to clamp down on AI generated posts.

Speaker 2: 26:24

So they're going to have a big challenge on their hands for sure. We talked about being creative with how you can actually use AI to kind of improve what you're doing from a PR marketing perspective, and these are some examples, actually, of the more creative uses. So crisis comms, believe it or not, can be something you can actually improve through the use of AI. So traditional crisis comms model is you have your crisis plan, you come up with some scenarios and you test the crisis plan, but why not actually ask AI chat, gpt in particular to actually pose scenarios based on parameters? So, for example, you could upload your existing crisis plan, you could upload some information about your company and you could actually then ask AI to pose you some crisis scenarios which you then have to respond to. And that's that's a combination of reductive and generative AI and can be hugely useful for you.

Speaker 2: 27:21

Same goes for media training, actually, as well. Again, the traditional model for media training is we will you know, this is something we do an awful lot of we will work with a particular client or a particular brand. We will media train them, we will look at potential risk points of the business and we will conduct kind of mock interviews and scenario testing and things like that.

Speaker 1: 27:42

We did a whole podcast on media training with Guy Clapperton. Check that one out.

Speaker 2: 27:45

Yeah, very, very good point. But we can actually start to use AI to enhance the way we tackle media training. So, for example, if we're releasing our financial results, if we're releasing a new product to a service, we can actually upload our press release or our results or whatever it is, and then ask chat, gpt or whatever tool we're using to actually ask us tough questions that a journalist might ask related to our release. So, again, this isn't going to replace the need for traditional media training, but it can potentially enhance the way we deliver media training. I mentioned this briefly as well. So, sentiment analysis and this is a really good example of where we can use something like chat GPT for sheets.

Speaker 2: 28:25

So, working in a busy consultancy, we know that there's an awful lot of time spent on analyzing media coverage, analyzing it for key messages, analyzing it for sentiment. If you play your cards, why order? That can be automated. And the example I gave again was chat GPT for sheets. Plug in for Google Chrome and it can basically analyze all the media links in your spreadsheet. And typically I mean, what do we get for a client in a typical month? 50 pieces, 100 pieces of media coverage. That can take a huge amount of time to analyze, can't it.

Speaker 2: 28:56

GPT for sheets, can basically analyze it for sentiment, for key messaging, for whatever you want, at the touch of a button, and that gives you more time to spend on creativity, on strategy and more value added work. Finally, idea generation, and this might make some of the creatives bristle a little bit, but AI can be used to enhance the creative process. It's never going to replace the creative process but ultimately it's a really effective way to kind of get some of those thought starters.

Speaker 1: 29:24

So, for example, Sometimes it's a bit shit though, isn't it?

Speaker 2: 29:27

But it's another tool. I think it's another tool, just as Google's a tool for when it comes to creativity.

Speaker 1: 29:31

Yeah, but you can't go, like you know, say I'm an in-house marketing director, I'm not going to go, I work for eDAM. Write me four marketing campaigns that are going to be used in 2024, and then it'll do it. And then it produces them. And it's like take a little tour around the UK and produce it. It's just really bang. Average is what I'm saying.

Speaker 2: 29:50

Yeah, it's about understanding the use cases. So I would say it's never going to come up with a killer campaign. It might come up with some insight for you, but it's things like blog titles. So if you're struggling for blog titles, if you're launching a campaign and you need a name for the campaign, in actual fact, without giving too much away, I think, the name of our podcast was inspired by AI, wasn't?

Speaker 1: 30:11

it? Yeah, it was.

Speaker 2: 30:12

Yeah.

Speaker 1: 30:13

And the name of today's episode has been as well.

Speaker 2: 30:15

Yeah, so there we go. So it augments the creative process. It doesn't replace it. And be aware of the limitations. Like Chris said, it's never going to come up with an amazing, award-winning marketing campaign for you. And well, it could do if you have an amazing prompt.

Speaker 1: 30:29

but it probably won't at this point, will it Not yet, but it could do soon. The tools to help you do your job more effectively. It's there to help you do the jobs you don't want to do, for you to be more effective.

Speaker 2: 30:40

Now a fascinating use of AI, which we've actually been experimenting with over the last few months, is persona development. Now we do an awful lot of persona development. At Prohibition. We have a whole suite of tools that help us develop really effective and impactful brand personas. So, for example, we'll use things like social media listening tools, audience insight tools, everything like that.

Speaker 2: 31:05

All of that's still absolutely relevant, but what we can actually do as well, which is a starting point, is put in a prompt.

Speaker 2: 31:11

We'll share this prompt actually in the show notes, but you can actually put a prompt into Bard or ChatGPT and ask it to come up with an audience persona. So the types of thing you would include job title, the business category that you're in your mission, your company size, where you're active geographically, what your business objectives are and the kind of desired output. So just to kind of bring that to life, I'll read a prompt that we put together Build a persona for a marketing director of a £10 million not profit in the Southeast, with a mission to provide better access to healthcare resources in underserved communities. The goal of this person is to create a more effective fundraising campaign through digital marketing. Please include specific details about goals, pain points and decision criteria. Snappy, very snappy, but actually you put that prompt in and it comes up with some really interesting and quite thoughtful audience personas which you can then build on and then refine using these have you just talked us out of a load of work.

Speaker 1: 32:17

Yeah, you can use these things to get you started. I think, Especially if it's as brands out there that don't have big budgets to work with agencies and to help them develop sophisticated personas. But yeah, it's a good starting point. It's better than having nothing. When people say you meet owner managers and you say, okay, who are you marketing to? Are products available for everyone?

Speaker 2: 32:34

And then final points, it's starting to impact the creative industries as well and creative. So everything from storyboarding for a video to generating images, there's a whole bunch of tools out there that can do this for you. And, again, it's never going to replace a skilled illustrator, but using something like Doorly 2 can generate images. There's. Isn't there an AI video generator as well? Will, yeah, we've started experimenting with that. Actually, it's called Runway and essentially you put in the criteria for your video and it will actually come up with AI generated short video clips. So if you wanted, for example, a couple in pub horse running through field, well, that's what an image that is. Horse running through, hello, lloyds. Chris riding bareback on a horse running through a field in the mountains, hello ladies, whatever the images you want, it will generate that in seconds and then you can actually use that as the basis of a video storyboard.

Speaker 2: 33:32

And the same is for images Doorly 2, the tool I mentioned, which creates a slightly creepy images waxwork images but you can actually generate bespoke images. You don't have to pay for stock, you don't have to actually commission something yourself and as a starting point, it's a really effective way to actually start putting together a really accurate and engaging storyboard, and our video team have started using that. So the point is, there's a million and one applications that are hyper relevant for the PR, the marketing, the creative industries. What we would say is you need to be using them, you need to be experimenting. It's so much more than just chat, GPT. There's a thousand and one different tools out there and think about what they can do and how they can actually drive efficiencies. That's what it's all about, isn't it? It's about actually looking at what eats up a lot of your time where can there be efficiencies and how can you use these tools smartly, just as the computer took over from the typewriter whenever that was I actually used to use a typewriter years ago.

Speaker 2: 34:27

In your first agency.

Speaker 1: 34:29

No, I had an electric typewriter.

Speaker 2: 34:31

It was like it in between my mum and my mum Like a word processor.

Speaker 1: 34:34

Yeah, and then I went, moved to a word processor and then moved to a laptop, which is.

Speaker 2: 34:37

Imagine how noisy the office would be with 30 typewriters going. Honestly, do you think we're ridiculous?

Speaker 1: 34:41

I remember my mum typing stuff out and using tip X to get rid of like it just seems crazy now.

Speaker 2: 34:47

Yeah, you can't ever make a mistake, can you?

Speaker 1: 34:49

No, okay, before we go into this bit, I've got a question for you, zach AI in your job video editing, sound editing how are you using AI in your role and what you do, and how often are you testing different bits?

Speaker 3: 35:03

Well, I use AI every day for what I do. It could be from sound design to storyboard and as you just mentioned earlier but there is literally an AI software or something coming out all the time that makes my life easier, and every evening, I'll scroll through Instagram and I'll find out a new technology that's just come out or something that's being updated, and then I go and implement that you know, maybe not the next day, but the next time.

Speaker 1: 35:32

So give us a couple of examples of the good stuff, because you've tested loads, because I make your test as much as possible, don't we? So what?

Speaker 3: 35:39

do you?

Speaker 1: 35:39

Yeah, so a couple of good examples that you use. That's helped you in your video process.

Speaker 3: 35:42

I'll keep it specific just to the podcast. Why not? While we're on the podcast, Every podcast I'll run. We've got these brilliant microphones, great sound, but we run it through an audio enhancing software that'll just get rid of any background noise and make the audio sound that much better. And then, you know, to create our snippets, we'll run it through a piece of AI software which will rank the most engaging moments. That will do the best on social media.

Speaker 1: 36:10

So this will be this moment, right now.

Speaker 3: 36:11

Yeah, because I'm in it. But yeah, it will.

Speaker 1: 36:15

We should just retire, Chris, yeah that's it, Just get an AI Zach.

Speaker 3: 36:20

It will rank basically moments from the podcast.

Speaker 2: 36:23

Yeah, okay, and talking of retiring from the podcast, chris, we're actually looking at what the most extreme example of using AI could be for us. So I'll explain this scenario and then I'll explain the implications of it. So we could auto generate a podcast script. That's pretty straightforward, isn't it? We could put in some prompts and say generate a one hour script for a podcast. We would then use an AI text to speak app. So we would actually generate a voiceover reading the script. So we'd have a voiceover basically reading that hour long script. We could then speak a single word into voiceai and it will clone our voice.

Speaker 1: 37:01

They're getting better by the way. Every week they're getting better.

Speaker 2: 37:03

Yeah, it's crazy, and we could actually clone one of our voices to then read this auto-generated script and what you'd have is a totally AI generated podcast. And would it be funnier and wittier than what we can do?

Speaker 3: 37:16

Probably Just to hop in here again. I don't know if you know this, but I have genuinely done this for both of you before. I've taken about half an hour's worth of your audio from a podcast and there was one part that the words just weren't coming out right or the sentence didn't sound right.

Speaker 1: 37:31

That doesn't sound like us.

Speaker 3: 37:33

So I wrote a whole sentence or a part into this voice cloning thing and I took your voice and I made you speak a sentence which you'd never even said before and neither one of you noticed it at all. Was it slanderous?

Speaker 1: 37:48

It wasn't. I decided to be mature about it. You can be chief editor of this podcast, right? I want you to insert at this point now a line that Will says that's quite funny that he hasn't said yeah, a joke For the podcast, let's not get it one liner, and here is the one liner AI generated.

Speaker 2: 38:07

He is going to hate me saying this, but he once had a client meeting in Cumbria and started driving there from Leeds but ended up at the Angel of the North in Newcastle, more than 75 miles in the wrong direction. I am funny right. Vote, ai will yeah.

Speaker 1: 38:22

Okay, yeah, yeah. So how was that?

Speaker 2: 38:24

Funny, making you laugh.

Speaker 1: 38:25

Probably funny than you. Well, that's been honest, Probably was actually Okay.

Speaker 2: 38:29

So what we are seeing, which is quite reassuring in a way, isn't it? That PR professionals are concerned by the risks presented by AI, and this kind of will lead us on to the ethics which Chris will talk around in a moment. So Muck Rack, the marketing blog, has done some research on this, and it posed the question which of the following do you think are risks generative AI poses for PR professionals, if any? So coming, let's have a look. Coming top was unscrutinized AI output lowers the quality of content. That's something we've talked about, isn't it the fact that suddenly there's going to be a glut of AI generated content, and does that mean an overall dip in the quality of content out there? It probably does, actually, but then the search engines and the social networks are going to get more sophisticated, so probably we won't see it.

Speaker 2: 39:21

Secondly, younger, slash, newer PR professionals don't learn the principles of the profession and rely heavily on tools. I actually think that's a bit of an unfounded fear. That's a bit of a it's a bit of a Luddite attitude, isn't it? Because actually it's just the evolution of PR, isn't it? It's using tools in different ways. I suppose the danger is, if PRs enter in the profession, don't learn basic writing skills and rely, you know, naturally, on generative AI Writing, I think, and creativity will always be core PR skills, won't they? Everybody has to be able to write to be an effective marketer or an effective PR. Professional Clients think they don't need content creators anymore. That's 56% of people are concerned about that, and that's quite an interesting one, and then audiences get overwhelmed with so much content. It's harder to actually get cut through. 33%, and I think that's that's actually the same point as the first one, isn't it? But I actually think the algorithms will naturally filter out this kind of AI generated noise, and the very bullish 3% are not concerned by any aspect of AI, which is probably that's probably leaving the industry, yeah they probably are.

Speaker 2: 40:25

They're retiring, so can we avoid it in PR, chris? Is that? Is that? Is that likely, or are we no?

Speaker 1: 40:32

I don't think we can avoid it. You can't avoid innovation If you, if you're sat in, whatever job you're in marketing, comms, pr, whatever you have to look at using these tools to help you speed up your job, and other people are using them. So we have to all be testing and learning. Testing and learning, seeing what works, seeing what doesn't work. That's the point of today's podcast is try and help people steer them in the right direction on thinking, just thinking about little things that you can do and use, like how can you perfect your AI strategy? We've been advising clients on how to, how to roll. We do like an AI workshop with our clients so we'll go in and analyze what they're doing and which tasks we can help them optimize with various AI tools, and that's more of an in-depth couple of hour session, but it really really helps because you're identifying the things that they're doing in their work processes and you give them a new sort of strategy and it can save them tons of time. That's the point of it, isn't it?

Speaker 2: 41:22

Yeah, and that's the killer benefit for me. You know, as kind of marketers and PR people, we've never got enough hours in the day, have we? You know, literally you could work and work and work 24 seven and you still wouldn't get your your to-do list done. But the point of AI is it can drive real efficiencies and actually give us that breathing time to focus on strategy, on creativity and on real adding value, and that's the major benefit as far as I'm concerned. I think there's a lot of fear. Well, a lot of this is scare mongering, actually but you know the fact that AI is suddenly going to replace the need for skilled workers in the economy and I think, obviously certain industries that that is the case. But when you look at marketing and PR, I don't think it's going to necessarily replace people. But people will have to reskill, won't they? I think there's going to be a need for people to learn these. Yeah.

Speaker 1: 42:14

There was a really important brand new report entitled the Impact of AI on UK jobs and training by the unit of future skills in the UK government Department of Education and that looked at which sectors were most at risk from AI. If you're a roofer and you're listening to our podcast, good news guys, ai is very little risk. Similarly, with farmers and people that work in manufacturing, they've got less of risk and they're less exposed via AI. However, professional occupations such as management consultants, accountants, psychologists, pr professionals, marketing people yeah, it is quite a big risk to our jobs, but not right now. It's coming, though. We've got to be, we've got to prepare for it.

Speaker 2: 42:58

Yeah, I mean I thought things like accountancy in the legal profession. I mean, so much of their time is on putting together very, very complex legal papers. They're obviously the whole business model is billing by the hour. I'd be a little bit concerned if I was a lawyer. Maybe it means our accountancy fees are going to drop next year, chris.

Speaker 1: 43:18

Yeah, thanks, I don't think so.

Speaker 2: 43:20

So we've talked a lot about AI. The use cases is coming, whether you like it or not. I think that's clear. The one thing we can do to get better at AI is to look at how we craft our prompts. So the better your prompts, the better your results, and I think there's a tendency to look at AI and think it's a bit shit, it doesn't really work. But that's down to your prompt, isn't it?

Speaker 1: 43:40

Yeah, it is, but I've got some good news. So prompts are really big and Will's about to just explain a few of the prompts, bits and pieces. But in November, in 2023, so very recently OpenAI rolled out custom versions of chat GPT and I'll say what it says here. Anyone can easily build their own GPT. No coding is required. You can make them for yourself, just for your company's internal use or for everybody. Creating one is as easy as starting a conversation, giving it a few instructions and extra knowledge and picking what it can do like searching the web, making images, analyzing data and you can go. If you go to chatopenaicom, you can create your own, but there's loads in there that have already been created, like customized chat GPTs for yourself. Really, really perhaps seen some good examples of them, so check that out Okay, so quick guide on how to create effective prompts.

Speaker 2: 44:34

First of all, we need to treat prompt creation as an art and we really need to spend time on refining and testing and learning when it comes to creating great prompts. There's actually some guidelines. You can look at OpenAI. We'll put this in the show notes. Openai has produced a really useful guide to creating effective prompts. Change your mindset as well. So, typically, ai tools won't be rubbish. They tend to do what you ask them to do, and if they're not producing the results you want, then probably you're not asking it to do it in the right way. I actually read somewhere that you're supposed to be polite with.

Speaker 1: 45:09

AI. Yeah, there's data that proves. There's data that's proven. If you say please and thank you to these AI bots, you get better results. So that's imposing and this is something. So please listen.

Speaker 2: 45:20

This is something Stephen Waddington was talking about, wasn't it around how we?

Speaker 1: 45:23

Previous episode plug.

Speaker 2: 45:25

Yeah, absolutely, how we're imposing human characteristics on computers. In effect, it's like the way we say thank you to Alexa when asking her to her it. Yeah. I say her, I think it's a her isn't it Well turns out, what voice you've got, I've got. You've got Chris's voice on your voice. Yeah, I find it very soothing.

Speaker 1: 45:46

I did that. That'd be awful. That would be awful. You just give me abuse?

Speaker 2: 45:50

Yeah, I would. And finally, practice and share team knowledge to get better at creating prompts. So don't just work in a silo. When you find something that works, share it with your team. Have a central resource where you're sharing knowledge about what prompts work really well. So we've got this at prohibition, for example. So the persona prompts I did a little workshop on it. We've now got a whole range of different persona prompts that can be used as a way to kind of enhance our broader persona offering. So communication is absolutely key. Okay, should we talk about a few tools? I mean, this is fraught with risk, isn't it? Because we mentioned the tool and the chances are, in a week's time it's been updated, it's changed, it's been discontinued.

Speaker 1: 46:33

But but useful tools for AI yeah.

Speaker 2: 46:37

Yeah, so let's go through a few tools that we love using at the moment.

Speaker 1: 46:42

So, zapier, so you can create zaps to do things. Well, now you can use, you can plug AI in and basically it can automate those tasks that you were zapping. So I mean, zapier is always if this, then that type technology. So you say, if somebody sends me a tweet, add them to RCRM. Well, ai can enable it to do much faster. So check out, zapier. If you've never used it Show notes, well, this is for people that have got podcasts. It creates show notes, converts your lengthy podcasts into brief summaries, finds specific information, highlights key points and more. We've got our own sort of one that we use for our podcast. We've already covered GPT for Sheets and Docs, which is redactive and analyzes documents and sheets Google Sheets and Excel sheets.

Speaker 2: 47:27

Yeah, I mentioned Runway earlier. That's that AI-powered video editing and video generating software. We actually use it to put together video storyboards, but there's a lot of other uses of it and there's with most of these, there's a kind of a premium offering and a freemium offering isn't there, and then we use.

Speaker 1: 47:44

There's a tool called RightSonic that we'll mention earlier. Rightsonic is built on ChatGPT, but it's a really sophisticated, like blogging content platform. There's others out there, but this uses the latest version. Depending on how many keywords you wanna use, you can make your content really, really specific, and it's now got an AI built into it that'll analyze your content and then make recommendations over what will work better on Google. So I did warn earlier about AI content with Google. I say that the best content is written by humans, so add your own references and examples of real life. But it is a great starter and it's paid for. But it's about I think it's about 300, 400 quid a year. But it's worth a play around with because you'll be able to turn out some interesting blog content.

Speaker 2: 48:29

And we mentioned the fact that it's not just ChatGPT. There's thousands and thousands of other tools which can be, to be quite honest, can be pretty hard to keep up with at the best of times. There's a tool called there's an AI for that, and that's essentially an AI database. It's got about 7,000 AI tools on it at the moment and essentially you can search via keywords and look at the most appropriate tool for you. So all of this talk about AI, chris, it sounds pretty risk-free to me. Is that a fair assumption?

Speaker 1: 48:58

No, it's not. There's quite a lot of risks and the ethics and the risks is one of the reasons why there was a. There was an AI summit up in I believe it was in Scotland with. I think it was in Scotland was in Scotland.

Speaker 2: 49:12

Was that when Elon Musk was there?

Speaker 1: 49:13

Yeah, who's Elon Musk was there, wasn't it that was?

Speaker 2: 49:15

quite a surprise and quite a coup for the prime minister, wasn't it Suddenly appearing out?

Speaker 1: 49:18

Yeah, but who's listening to whom there Is? You know, there's a whole debate around that.

Speaker 2: 49:24

Yeah, where does the power lie? It lies with Elon, doesn't it so?

Speaker 1: 49:28

your question about AI misinformation. I'm gonna go, I'm gonna, I'm gonna contraintrate on how that it can be used in a negative way. So there was a story on the BBC in December where the journalist created a one of these custom GPTs and they called the entire. They created it and they called it crafty emails and told it to write text using techniques to make people click on links and download things to them, and basically the bot was able to craft highly convincing text for some of the most common hacks and scams. So some of the scams that we've talked about that we were talking about before the show was high mum. That tech scam where people had sent texts Well, crafty emails was able to write the text pretending to be a girl in distress, using a stranger's phone number to ask a mum for money for a taxi, which is a really common scam.

Speaker 1: 50:19

The Nigerian Prince email it did that. It did the smishing test, one which is? It created a text pretending to give away free iPhones. So it had used social engineering techniques like need and greed principle apparently the AI has so. And then there was a crypto giveaway scam, which was pretty bad, and it's also doing phishing emails, but bespoke to. So misinformation and the way that AI is being used is scary. We do need to be aware of it and they need to get a hold of it. It's why they had the AI summit here in the UK with Elon Musk, as Will just said. Yeah.

Speaker 2: 50:58

But there's no legal policing of this really is there and that's unlikely to change.

Speaker 1: 51:04

How do you police something that's global as well, in multiple languages? Now you've got AI that's writing in English, but it doesn't need to be in English. You could translate into 75 different languages within three seconds. It's insane. So how do you police that? How do you stop the Chinese or the Russian governments from hacking things using special phishing emails that are pretending to? There's all kinds of different things that need to be considered for AI and the weaponization of it, but it can happen and it is happening already.

Speaker 2: 51:33

Yeah, and another flash point will be the UK to a degree, in the US elections next year. We already heard about the Cambridge Analytica scandal a few years ago, potentially influencing the Trump you know, the Trump presidency him getting into power. It's quite alarming actually to think about the role AI could play when it comes to the next US elections, isn't it? I thought I'd done, you know, and how will that be regulated? I've got no idea.

Speaker 1: 52:01

Yeah, that is worrying how it's going to be used and similarly used on social media platforms to scrape data and create fake profiles with people in the right tone and sending messages to people. All that stuff can be done. Similarly, the fake images thing I love the fake images and then I want to see the Pope in the big white puffer jacket which just went viral. That was amazing, that was, and it was really good. You check it out. If you haven't seen it, just Google fake Pope and white puffer jacket.

Speaker 2: 52:31

That's quite a realistic generated image, isn't it? It's not typical of the images, it's really it's quite clear as well.

Speaker 1: 52:39

It had millions of people downloading it. There's other issues as well. There's like the legal side of things. So Getty Images sued an AI art generator called Stable Diffusion in the US for copyright infringement and then, basically, when we did the webinar, we showed an image of we had Jordan Henderson tackling a Tottenham player, and then the AI generated an image that was very, very similar, but the logos look different and the faces weren't quite there. And yeah, so Getty Images is suing for using, because what it's doing is scraping the images and then recreating them even if you haven't got them. Who owns that copyright? And that's why they're suing them. So the argument is that these systems have been trained using images from Getty Images, which it hasn't given permission for.

Speaker 2: 53:25

So yeah, and that could open a whole can of worms, couldn't it? I mean, you think of every stock image or stock video provider. They're going to follow suit, aren't they?

Speaker 1: 53:32

Yeah, the UK, us and EU have all got different copyright laws, so there needs to be, we need to come together and figure out a route forward on this.

Speaker 2: 53:41

The other issue for me is the pace of change is so rapid, isn't it so? You know, you think about what's happened in just a month in terms of AI development, and this is quite alarming. I mean, it's got massive implications. It's got implications on the advertising and marketing industries. So the question is will we be able to keep up? And this taps into that idea of digital Darwinism, the fact that the pace of technology may increase at such a rate that organisations and companies can't actually keep up.

Speaker 1: 54:10

I think we need to keep up. It's whether the police and the legal systems can keep up. They can't even keep up with streaming, never mind AI. Ai is crazy. I think it's going to be the legal systems and the frameworks, that's around, the ethics of using AI, but all countries are using it. That's the worrying side of it. It's like you know, the UK can do whatever it wants to protect, and the US can do it create its own laws but what about the other? What about states that aren't under the Western control? And I'm not saying they're better or worse, I'm just saying that are they going to get involved?

Speaker 1: 54:41

It doesn't stop some other rogue state doing some sort of weaponisation and using it in a different way. But I would say, if you're worried about keeping up on AI, just keep listening and reading to various bits and pieces. Try and stay up on top of it as much as you can, because there is new stuff coming out all the time. But all it's going to help you do is speed up things a little bit quicker and that's it. Don't worry too much, don't get the fear of missing out thing. I think it's just keep reading and be open to innovation.

Speaker 2: 55:09

Yeah, for me it's just using the tools. I mean there's so many out there, just experiment with tools and look at what's right for you. That idea of governments keeping up I mean when we did the Andrew Bruce Smith podcast he was telling us that graduates in AI in the UK and now being snapped up by Silicon Valley and they're looking at kind of starting salaries of $500,000 a year. So the top AI talent is being snapped up. So where is the expertise in government? I mean in central government? There just isn't anybody with that level of expertise. So how on earth is government going to?

Speaker 1: 55:43

police. This all it's quite alarming, isn't it? That's a creative AI government.

Speaker 2: 55:48

Yeah, quick discussion before we move on. Should we actually divulge the use of AI, and I've seen different schools of thought on this. Some people will put a footnote on a blog post that they might have used to research a piece of content. Would we divulge the use of Google, though?

Speaker 1: 56:11

I just think that's over the top. The world's most annoying legal requirement is do you want to accept these cookies? It's pointless. I get that 100% content that's just created by AI is not ethical. But to get everybody, everybody's then got to say I've used AI for 25% of this article. I think we've just got to accept that everybody is going to be using it. Currently, 75% 80% of people are using it regularly. So I think we're going to have to just assume that people are using it. But I've seen the reverse of this world. So I've seen people. There's a guy called Christopher Penn who I read and listen to sometimes and his newsletter starts with this newsletter was completely written by 100% the human being. That is Christopher Penn, which I always think is a bit weird, but it's quite interesting that he's saying and he talks a little bit about AI and he's basically saying that it's all done by humans for his side of it. So it's quite interesting.

Speaker 2: 57:13

Yeah, the ethical side of it. There's a lot of literature on this and we'd encourage you to read it and make your own mind out. Really, we'll link these up. The CIPR has got a really good guide. Ethics to Artificial Intelligence in PR. Is that a combination of Stephen Wellington and Andrew Smith?

Speaker 1: 57:32

There's a panel.

Speaker 2: 57:32

I think yeah, but the CIPR is a really good resource and if you go on their website there's a whole load of kind of downloadable PDFs which give you the kind of the lay of the land. A lot of it's quite common sense, I think, when it comes to the ethical side of things, but they're definitely worth a read if you're serious about communications.

Speaker 1: 57:51

And then we've plugged it already, but if you've liked what you've heard about today, I definitely recommend you come in to listen to our podcast with Andrew Bruce Smith, which is called Unleashing the Potential of AI and Marketing, which is where we go into other areas a little bit different, but it's similarly very, very interesting. It is a fascinating area and it's moving quickly, as Will said so yeah, check that out.

Speaker 2: 58:11

And just to conclude things, Chris, do you want to? Let's do an AI-generated thank you from Chris. So you're going to hear it now.

Speaker 1: 58:19

Thanks for listening to the socially unacceptable podcast everybody. Please like and subscribe. If you enjoyed today's show. It means a lot to us, and I'd like to say a special thanks to Will for joining me today and being an incredible co-host, but most of all, thanks to our brilliant production team. From the filming to the editing, you do an amazing job.

Speaker 2: 58:39

How was that?

Speaker 1: 58:40

I don't know how I've heard it. So thanks for listening to Socially Unacceptable this bonus episode on AI. Don't forget to click the subscribe button. We need every single subscriber we can get. We're also looking for our furthest listener around the world, so if you're not in the UK, drop us a message and let us know where you are listening, because we find that fascinating. If you did enjoy the show, please leave us a review. All reviews are read. I'll reply to all of them if I can. So please let me know. And hi, mum. So thanks for listening. All about AI. Unfortunately, we haven't got an AI-generated ending, so we'll just say thanks very much for listening and we'll see you in the next episode.

Speaker 4: 59:19

Thank you for listening to Socially Unacceptable. Please remember to subscribe to the podcast and leave us a five-star review. Don't forget to follow us on social media on Instagram, TikTok and LinkedIn at ProhibitionPR, and Twitter at Socially UA. We would love to hear some of your career fuckups so we can share them on the show. For more information on the show, search ProhibitionPR in your search engine and click on podcasts. Until next time, please keep pushing the boundaries and embracing the socially unacceptable.