This is an unedited transcript created using Otter.Ai. As such, it may contain minor errors.

 

Chapter One: Introduction to AI

 

Paul Sutton 0:00

Hi. I’m Paul Sutton and this is digital download the show where I talk to topic experts on digital marketing social media and public relations, about the things that matter in today’s communications industry.

 

Chat GPT has revolutionised the communications industry by introducing advanced conversational AI technology is natural language processing capabilities have transformed how businesses and individuals interact. Chat GBT enables more efficient customer support through chat bots, reducing response times and improving user experiences. It’s also streamlined content creation, helping marketers generate engaging copy and assisting writers with inspiration and grammar checks.

 

Furthermore, chat GBT has enhanced language translation services making global communication more accessible. In essence, it’s ushered in a new era of effective intelligent and convenient communication across various sectors.

 

Now, if all that sounds a bit, well robotic, that’s because what I just said is what chat GBT itself told me, but run that through a voice simulator that can effectively clone my voice and your friendly neighbourhood podcast hosts is probably out of a job.

 

Both hype and concerns over the impact of aI have exploded over the last few months. From a marketing communications perspective, there’s a huge range of articles and advice from listings of the ways AI will change the industry to portents of doom. Log on to LinkedIn and you can barely go a day without being exposed to 10 Chat GBT prompts type content. And all the while many, if not most communicators are going about their day jobs and pretty much ignoring it totally. Now that, to me, at least is insane, given it’s so much bigger than just communications.

 

In June, Mo Gawdat, who is the former Chief Business Officer for Google X, which is the Google innovation arm, joined a swell of high profile tech specialists in speaking up about the dangers of AI. He talked about humans limited intelligence compared to AI, and explained that if you could apply an IQ to chat GPT for it would score about 155. Now, Einstein was 160.

 

He went on to say that if Einstein had tried to explain the theory of relativity to someone with an IQ of say, 80, that person would have no concept of what he was trying to explain. And taking this further chat GPT for is something in the region of 10 times more intelligent than chat GPT three was. And so it’s not unreasonable to believe that when chat GPT five is released next summer, you’ll have an IQ of somewhere in the region of 1600. We humans will have no idea about some of the things you can come up with.

 

Whereas to some there is a big existential threat of Skynet and robots taking over. And although that is a small possibility, that’s not really what we should be worried about. The fact that AI will be way smarter than us within a year is a much more significant threat. Gawdat  projects that by 2045, ai will be a billion times smarter than us. I put this to today’s guest, Stephen Waddington, who in June released a management paper assessing the impact of AI on the communications industry.

 

Chapter Two: Defining Intelligence

 

Stephen Waddington 4:04

I’d argue with that point. Because it depends how you define intelligence. If you define intelligence that is access to large amounts of data and to be able to spot trout ponds, and since the thighs concepts from that data, then yes, if you find intent if you define intelligence as human creativity and the ability to build and create new knowledge, then no, because AI hasn’t been demonstrated to be able to do that at the moment. It can act on a large data set, draw trends patterns, respond to queries and build texts based on its dataset but he can’t yet build new knowledge.

 

And we as human beings we seem to want to map human characteristics on AI actually is it’s artificial. It’s not intelligence is artificial. And it appears to have human like characteristics? Because it can work so quickly? Yeah, thanks to machines, and it can work on such large datasets. So duty for that you said, the hours 45 gigabytes of information from across the web, Google Scholar, Wikipedia, Google Books. And you know, we don’t know how did the all the different vendors stop publishing the size of their datasets? Because it gave a bit of a willy waving exercise. Yeah. massive amounts of data. And that’s the point, you can draw out seemingly intelligent contracts from that data.

 

PS 5:35

Yeah, intelligence says it is a difficult word to use. I totally agree with you. Because, yeah, it is drawing something human out of it. Do you? I mean, you’ve looked at this area quite a bit, do you have any sense of, I don’t know the right word fear about what could be coming five or 10 years down the line or, or not?

 

SW 5:58

Hopefully, I’ll be out there, you won’t care. But you know, the reason you make a good point, the reason I have doubled down on it is to continue to be relevant, and understand trying to understand what’s going on. And the you know, so I’ve been looking at this for the last five years through the panel, the CIPR or, you know, and it did a piece of work. One of the early pieces of work he did was to map the tools, around public relations onto you know, the skills of public relations, it figured out, we’ve talked about this, we’ll figure it out around 38 40% of the competency of public relations, and marketing could be done by tools or will be able to be done by tools.

 

But, you know, 2023, here we are, and they got it pretty, right actually, looking at it, because to your initial point, anything where you’re trying to research information, you want to write you want the first idea, the first of anything, you know, you’re acting on knowledge, it’s there already, open AI would chat GVT completely democratised AI systems when in November it launched to work, you know, via a web browser and gave us all access. It’s interesting.

 

Chapter Three: Impact on the Communications Industry

 

PS 7:13

You mentioned the the AI and PR group, actually, because I was looking back at that. Because I remember very well that that piece of work being done five years ago. And I was thinking if we’d have had this conversation, nine months ago, I would have looked at that and thought, but that overshot the mark a bit. And then suddenly, it appears. So it’s quite an odd one that the way things have moved very so fast very quickly. Yeah. Was Was that your aim? I mean, it’s a bit difficult to say now. But prior to the launch in November of chat, GBT, were you under the same impression that maybe we’d have overshot those?

 

SW 7:55

Yes. In fact, Andrew Smith and I, on behalf of the CIP are we’re doing an audit through the summer, last year 2022, looking at the tools market, and we got access to Scott Brinklow, the HubSpot guy at HubSpot, he has a database called the chief martech stack, he gave access us access to that and allowed us to, you know, start drilling data from it. And we were starting to pull some really good analysis. He’s got 100,000 10,000 Sorry, 10,000 tools within that. Okay, they’re all very much focused on, you know, automation, though. And improving workflow.

 

Yeah, we were close to publishing this, when along came chat GPT and open AI and we completed the cable shuttle project for nothing? Well, no, because, you know, because we put it on ice, let’s say and said, Hang on, we need to pause there and reflect on this. And so the piece we ended up publishing in February, March was, you know, much more progressive, contemporary and relevant.

 

But yeah, I mean, the technology has been there, the technology and machine intelligence has been there since it’s existed since the 1950s. It’s just computing power has got so much better. And the ability to build large datasets has got so much better. And it’s the inflection of the two that makes this now possible. So, you know, there have been AI tools that we looked at in 2018. But they’re very crude, you know, it’s based on not generative AI, but actually reductive AI. So taking large bodies of text and creating tweaks, social media posts, and so forth.

 

PS 9:45

Yeah. Based on on what we’ve seen in the last six, nine months. It’s kind of not unreasonable to me to believe that the impact of AI is going to be similar on the marketing comms industry as the internet Yeah, as social media, you know, those big things that have happened over the last 1520 25 years, whatever it is, it strikes me that I mean, potentially is bigger than any of them actually, what were the major ways that you think it could impact the industry?

 

SW 10:18

I think anywhere, so any of the professions that are reliant, so if you’re relying on a body of knowledge, you know, you can now train a machine with that body of knowledge. So, you know, they are going to be undoubtedly the most the most impacted. So you think about any other professions, accounting, public relations, marketing, undoubtedly, legal, undoubtedly, there’s going to be a huge impact.

And that impact will be anywhere where it, you know, you touch large amounts of data. So legals a very good example, you know, previously, you’ve had, you know, people coming into the profession, whose role has been to, you know, study case law and get to grips with large briefs. Now, you know, machine can do that, and provide a summary of of, you know, large transcript, interview, conversation, you know, to a lawyer and do it really, really effectively and quickly.

 

PS 11:17

And then there are two areas to this aren’t there because you’ve got the external things, when it comes to PR. There, it seems perfectly reasonable, that there’s going to be some sort of reputational or could be some sort of reputational impact on an organisation. And then you’ve also got the internal things, which is the things that maybe we can do something about, which is just sort of workflow stuff.

 

Chapter Four: AI and Ethics

 

SW 11:38

Yeah. So from, from the perspective of the popular Well, I’ve perspective of any profession, there’s, you know, what is the impact on my organisation and the externalities around that organisation. Part of the issue around AI at the moment is even the platform’s that generate creating the technology, don’t fully understand how you know, its potential and how it can be used. And, you know, us spied very, very quickly, how easy it is to, to use a tool like, charge up to create, you know, massive volumes of text, data, whatever you like, images.

 

And in doing so, you know, you just plug thinking, just imagine you plug that into the web, plug right into, plug that into, you know, a network like Twitter, and the ability to completely pollute, pollute, a network is just, you know, it’s just so apparent, I have a huge fear about what’s coming up in terms of the election in the US, which we know is going to happen in the next next 18 months. And also in the UK, we’ve got an election coming up next year. Yep. You know, and massive, I’m personally huge concerns about that.

 

PS 12:52

In terms of false narratives being made by AI, or just generate massive amounts of content

 

SW 12:59

so yeah, just the ability to this technology being weaponized, to produce massive amounts of text to produce false information, you know, to do it at scale and to be to be weaponized. And then that relates to, you know, the ethical applications of, of AI, you know, it’s application for, for good and bad. You know, and we’ve just got no regulation in any of this area, either, you know, the issues related to copyright. And with no regulation in this, you know, if if you pass some text, or if you ask chat, GVT or bar to write, you know, a profile of a Paul Sutton and it comes back with a profile who owns the copyright to that test. You take that stage further, if you are using it to create ideas, and you patent those ideas who around him because no human has been involved.

 

PS 13:56

I think that the whole area of sort of weaponization, if we like to use that word and regulation was something that that interview I mentioned with Moe Gordo, that’s what he was getting at really, is that. So it’s like all of these things when when we’ve had them. It’s not the technology itself, that’s not the problem. It’s the way Pete knows. You know, bad actors getting hold of this stuff could be really, really problematic,

 

SW 14:22

it could be really problematic. And, you know, against this, we’ve got a context where, you know, just just rewind, a little bit non you know, open AI stole a march on the market by releasing GPT 3.5. The first data set in November, probably ahead of when it should be had no, it has no got very limited guardrails. And even now on Google, I’ve been working on this project called BOD for a similar length of time, and had, you know, had hadn’t put this out into the market. Because it was concerned its ethics teams were concerned about it. exactly some of the issues that we’re talking about.

 

Yeah. And, you know, as a result of open AI launching in November, you know, it’s fairly followed fairly quickly by Google and barred because, you know, they didn’t want to miss out. And, you know, there’s the famous example of one of the first demonstrations are barred completely hallucinated got the wrong answers, this got this term hallucination, where they literally just make stuff up. Yeah. And you know, it not something like 10% off the share price or Google for a day. So, you know, there’s massive, massive concerns. So the tech firms themselves, Google, open AI, and Facebook, and I’ve been pulling back Twitter have been pulling back their ethics teams rather than investing in Yes. Because there’s been so critical.

 

PS 15:49

Yep. Absolutely. Do you think there’s a possibility even of new discipline springing up around? PR, basically, related to AI? I mean, AI, for example, it could it be, I don’t know, reputation management related to AI? Do you see what I mean by that?

 

SW 16:07

Yeah. So think about you use the analogy of the internet, which gave rise to the search industry, and then Social Media gave rise to the content industry from a marketing perspective. Yeah, I think if you’re optimistic, the same things got to happen here, in that there will be an aspect of relationship management around the handling of data by these models, how they are used and implemented by organisations within the public sphere, and use cases

 

And then, you know, you can already see a whole range of sort of functional activities that are required, in addition to existing workflow. Or maybe they replace or augment existing workflow, but related to things like fact checking, related to, you know, writing instructions, and getting the best performance out of these, these machines. Now, arguably, these are all public relations functions, you know, we’re related to a machine rather than, rather than a human being.

 

The hallucination things interesting. So the way these technologies work, they are predictive, they’re based on you know, huge amounts of data, you query them, and they’ll come back with responses based on themes around the thematics around the issue or query that you made. And so this has given rise to this term called hallucination, because just as often as they’ll get it right, or get it wrong, if something isn’t in the large language model, it will literally make up the responses, which is what happened in the Google instance. So you know, for example, go and ask Bard or chat GPT, to write a bio of yourself, and almost certainly will get the first paragraph or I might even get the second paragraph, right, because you know, this, you operate in the public relations industry, there’s spoken at conferences, and stuff in well known. But once it gets behind the second or third paragraph, if you can’t find any, any more information, it will draw information from biographies of similar white men in their 50s. Just make stuff up and just make stuff. And that’s got to be that’s a matter of concern.

 

Chapter Five: AI Tools

 

PS 18:22

Have you done an analysis of the number of tools out there that? I mean, I know it’s changing all the time in this sprint that mean, God knows how many there are now. But have you done that sort of analysis?

 

SW 18:35

It’s moving so quick. So the the open AI API has been used and incorporated in and it’s driven this third party market for a whole new set of tools. Yep. So tools that sit on top of the database of a large language model, I use those terms into interchangeably. And if you like, provide a level of application on top of those. So things like darly for creating images, dollar two for creating images, things like, you know, I discovered one recently, cog Gnosis literally only in the last couple of weeks that starts to provide, you know, you’ll you’ll ask it to do a task, it will break that task down into discrete functions, and then it will go away and work through each function, literally mimicking the activities that you will do as a professional or knowledge worker. And yeah, there’s fuge tools, which is a AI a database of AI tools. I mean, there’s around 2000 tools in that now. There’s Product Hunt, you know, if you subscribe to their newsletter, every day, you know, you’re getting a dump of 1020 30 new tools. There’s a lot of shiny new shit that you could literally jump on every single day and it does become absolutely exhausting.

 

PS 19:50

Are you trying a lot of these personally?

 

SW 19:53

Yeah, I’m trying a lot of them personally. I mentioned somebody called Andrew Smith. He devotes a huge amount of energy and now do choose to follow him on Twitter or LinkedIn, because, you know, he’s, he’s going well, we haven’t got the time to doing the work for us. And I really appreciate that. But you know, I’ve stopped looking at the shiny new toys, because, you know, it’s it’s software developers serving, pushing functions into a market that, you know, might or might not be particularly useful. And it’s much more useful to break down the tasks that you have within your domain, and then look for, look for tools to support with that.

 

So I’ve kind of flipped how I think about this, and work and work on this. But you know, for all pretty focused on public relations for almost every aspects of public relations, activity, there were there is a tool now to help them support what you’re doing, you know, this podcast, you’re, we’re recording it, you know, we could have a little bot sitting alongside from otter or fireflies, that within 10 minutes of the meeting finish will create a transcript for us, and will pull the key bullet points out of that transcript. And that sort of thing is terrifying. There you go, you’ve got, you know, for a half an hour of text, you’ve you’ve which will probably take, you know, three or four hours to transcribe create notes, find the key themes from it, and you’ve destroyed three or four hours of work.

 

PS 21:24

But it’s a great example, because you’re right, I mean, I going back to I mean, I’ve, when did I start this podcast? 2018, I think. And at that time, so we took him five years ago, the transcripts, technology around was pretty poor. You know. And because of that, I’ve never really transcribed them. And I know that’s a bad thing to do. Anyway, full stop. But nowadays, yes, you use a tool like auto you transcribe your podcasts from that you plug that transcript into another system, which can then write you five blog posts, for example, about that one podcast, you could then carve that into God knows how many social media updates, you could create a white paper, I mean, there’s, it’s endless, which would take minutes, I know it needs polishing. But that would previously have been hours and hours of work. And it’s just done.

 

SW 22:18

So I’ve started doing a bit of a party trick at conferences and events where I’m invited to speak in, I demonstrate how AI could provide the function of an account Exec. And you know, we talk a lot about generative Aikido about reductive AI is equally as important, I think, as n is disruptive. And that’s an example of, you know, the transcriptions an example of reductive AI, because he’s taking a large amount of data and making making so help you helping you make sense of it. But you know, I do this demo where over the course of 10 minutes, I’ll take a white paper, upload that into chat GPT you know, the three or 4000 word paper, you ask it to write a press release, comes back with first jet first draft of a press release, it’s okay. Typically, it will be enough for you or AI to take and and work on it as a first draft.

 

That’s only going to get better and then say then say right, I don’t like the headline, can you generate me 10 alternative headlines? And all that happens within 30 seconds live? Can you create me a quote relevant to the PR industry then the you know, management, to marketing and they’ll create your three versions of a quote? Can you write me an email pitch as if I’m going to pitch this to a business media? All right. And it all does this in real time and live and that’s terrifying. It is because suddenly you just see so much work disappear.

 

PS 23:50

But you do you do hear people say, Well, yeah, it can do that. But he doesn’t do it very well. And this tends to come from copywriters actually, you say, Well, I can do it way better than that. 

 

SW 23:59

Richard Bagnall. In fact I’ve heard him speak. So he was chair of American news. Karma leads, SEO karma. You he did this really depressing piece about PR singularity where we’re now taking what journalists are writing, were framing a story or a pitch in the context that a journalist based on what a journalist has written to make it relevant, where pitch and get the journalist is then you using AI to help them write a story of publishing the story. And then, you know, the measurement industry is and then using automated tools to figure out you know, the sentiment and positive as that’s definitely where we terrify me.

 

PS 24:52

when you look at all this stuff, whether you’re thinking about the ethical issues, the tools, what they can do the impact act on stuff. It’s It’s overwhelming. I mean, I find it overwhelming. I’m not afraid to admit that. I have tried to keep up to date with this, but I haven’t I haven’t put anywhere near the, perhaps the time that you have into really understanding this stuff. What should we be focused on because a lot a lot of it, especially when it comes to tools is white noise, I think you’d agree with that.

 

SW 25:24

A lot of this stuff is fenders pushing solution, exactly Chinese solutions onto a market that doesn’t really know what he’s buying, doesn’t understand it. And you can very quickly end up with you know, it doesn’t $20 subscriptions a month that you don’t really ever use,

 

Chapter Six: Getting Started with AI

 

PS 25:38

just going on that which are the tools that you think if there is a toolset a skill set that we should be adapting that are the key ones if people re listening to this. And they don’t know where to start, what what are the sort of half dozen or so the

 

SW 25:55

go for chat GPT because you know that that’s a fundamental platform technology based on a large language model. Googlebot is also, you know, an autos a similar tool. But I jumped very early on to chat GPT. And I’ve stuck with that, that will give you, you know, a fundamental grounding in a lot of both the reductive and generative application. So you can start to use that at a command line prompt level, to ask, you know, basic queries, you can ask it, you know, to become support you in doing research support you, as an analyst support you in creating content, and it will start to give you a very good understanding of technology, and what’s capable.

 

And then there’s a series of applications, alongside chatty Petitioner chat DPT focus is very much focused on tax, but then can look at things like dolly to that, or how you create images. You know, so we’re working in a tax environment with with charging parties. So chatbot, essentially, but dolly will help you, you know, create images, and none of that’s not work area of explosive ever explored as much. But, you know, it’s, it’s really interesting, you know, functional level design. So those are two, those have covered you for text and images. And then there’s an application layer that’s starting to sit on top of these tools now that are helping you with different aspects of of your job. So I’ve used an application for for a couple of years now called Beautiful AI instead of Paris power replacement for Power BI.

 

And now it’s like, interesting is like having a an assistant sit by your side or designer in this case, that helps you to lay out slides and presentations, and just gives you nudges along the way, and really helpful suggestions. And it will also it’ll also find images for you or, you know, create images for you. So I’m using that a lot. I mentioned, you’ve talked about otter fire files is another really good tool, just it takes they large amounts of data that you might create transcripts in particular is good. And will you know, help you make sense of that. So write your transcript, find the bullet points in that. And the latest tension, slightly terrifying, the lighter latest version of ifl.

 

So I used recently, it will mark out the percentages of who’s spoken during a conversation, and then give you a sentiment scoring. Free, plentiful and miserable your words, existential. Honestly, I found the reductive applications, the most useful so there’s an application called Word tune that I use it almost every day, where it will take a lot, you know, a large 10 page PDF, and break it down, page by page. Basically write crib notes for you, you know, so all these handful of tools that I use, and so tools is one thing the other and you know, immerse yourself in those, start with a large language model, and then look at some of the applications that sat on top of it, focus it always around your workflow, I’d suggest set the next thing you have to do, though, is think within a business context, you know, what guidelines are you going to put in place?

 

You know, in lots of organisations, particularly in their PR context, and now looking, looking at this, you know, because what level of transparency Do you provide publicly when you’re using these tools? You know, how do you counter bias? How do you, you know, how do you counter issues like copyright? And often there are answers. Yeah. No, and then equally, there’s not that much advice come out. Advisory coming out of the PR industry. Yeah. On, you know what best practice should be because we literally at the start of this,

 

Chapter Seven: Disclosing AI Use

 

PS 29:51

yeah, it’s an interesting point, because using that example, I mentioned let’s say we take the transcript of this podcast, and I produce half of it 1000 Blog posts from the transcript and I post them on my website. And I have a light touch edit on them, let’s say, you know, I use them as a first draft, but it’s a light touch. Do I then on those blog posts? Should I be saying this? This article was written with the help of chat GBT, or whatever taught by us?

 

SW 30:21

I don’t know. I don’t know the answer that Yeah. You know, it’s everyone from the office to copyright those who are looking at this issues. And there just isn’t any regulation to to cover it. I mean, you know, you don’t disclose when you’re using the spellcheck or grammar checker or family. Yeah, you know, we’ve accepted those as tools that are part of our workflow. I’ll be quite interesting, if you did, but, you know, I think they’re going to become I think these things are just going to become part of our day to day workflow. There’s, there’s, there’s a point related to that Microsoft, which is an investor in open AI, has said, it’s going to launch a tool within the office 365 environment called copilot, again, sitting on top of the chatter YouTube large language model, as an assistant, they’re calling it co pilot with our Hangouts in a really interesting piece of positioning. Because, you know, it’s clearly there as an analyst or researcher to sit alongside you, and to support you in what you’re doing day to day. But that’s how in between now, and that’ll be launched in the wild by the middle of next year. Okay. That’s the point where no one can avoid this.

 

PS 31:39

I think it’s a fair point. Anyway, this is happening, this, this can’t be turned back, it won’t be. 

 

One of the things that’s been on my mind, actually, for quite a while now, I, this gonna sound a bit odd. But I’ve been through this process of thinking. So I’m in my 50s. Now, and I do a lot of training, right on social media strategy, or digital marketing strategy, whatever it is. And I go into a room. And this 50 year old guy comes in, and he’s training 25 year olds, and I’ve been thinking, Do you know what, there’s only a certain lifespan that I can go on before they start rejecting me as an old granddad?

 

And I was thinking, well, if I had to tweak my career, there’s got to pay a significant role there for the prompt side of things. So it’s all very well saying, right, I’ve got this AI tool in front of me. The skill is, is learning to prompt it in the right way to get the result that you want. And I think when when we were saying, these are the these is this is a set of tools that you should be getting to know. It’s not a case of Oh, yeah, I’ve used GPT. Three times. It’s spending time actually learning, learning the way to prompt it, that that strikes me as the big skill that’s probably just around the corner, if not there. Now

 

SW 33:03

So first of all, I correctly you’ve got, you know, your your career that’s come before now is always your pension in terms that you’ve got the wisdom of age. And, you know, to be fair, we’ve been through economic cycles, multiple, we’ve been through, you know, innovation cycles, in a PC, mobile phone, internet, she’s talked about social media, we’ve lived through all of those, you know, I too, I’m in my 50s. And so you’ve got the experience that you can bring to bear. And you know, the store, we talked about this before, so many times, PR really did miss SEO, and it was slow on it was slow on social media and other industries grew up around it.

 

Well, you know, potentially the The same could be happening here. And, you know, the PR industry is not good. PR and marketing industries aren’t good at innovation and adopting new forms of technology. And the data that’s coming out of, you know, the industry associations is very negative in terms of, you know, from a risk perspective, we we need to put all this back in the box. Well, you know, you you just can’t sense this the same arguments that wouldn’t you know, when Facebook and Twitter first, first emerge, so, you know, actually your future proofing your career, and you bring a huge value in, in your experience to bear.

 

Second point, though, prompting, yeah, I mean, so numerous organisations from McKinsey to the World Economic Forum have said, you know, prompt engineer is the role of 2023. Because, you know, that’s the way you train and get the best performance out of machines. And yeah, it’s, it’s certainly my experience, these machines will regenerate speed. So if you don’t like the first draft or something, you ask it, give it a fresh instruction, and he’ll come back with a better response and you know, absolutely worked.

 

Chapter Eight: Teaching Yourself About AI

 

PS 34:58

Obviously, you’ve spent more time than I have doing this stuff as well, is that something you? You will have learned that it’s something that you consciously have learned how to do? Or is it just? I don’t know. It just did in your view on that, really.

 

SW 35:13

And so it’s something I’ve I’ve taught myself, absolutely taught myself just through experimentation, but then also reading. Now there’s tonnes and tonnes of resources, you know, from Twitter, through to Facebook groups through to LinkedIn conversations and Reddit, where, you know, people are experimenting with this stuff out in the wild. And so yeah, I’m just jumping on that bandwagon.

 

PS 35:38

Yeah. Okay. You mentioned there, that the PR industry is generally slow to adapt to stuff, new technology, specifically, which you can’t deny I totally agree with you. What do you think we need to start doing now, as an industry? If someone’s listening to this, what do they really need to start doing now? Because it won’t be long before it falls behind. You know, in this case, it’s coming on so fast. And you likely say there will be other industries springing up around it. What do people need to start doing?

 

SW 36:12

So, anecdotally, I’ve now asked this question at the start every time I do a speaker conference or an event and, and typically it’s around 50% of people have, you know, experimented with it. Everybody’s you know, everybody’s aware of it, you can’t not be it’s such a huge issue within the public sphere. You know, there’s a piece of work came out of Annenberg, USC Annenberg, in this stage, which says university that tracks PR issues, said around a fifth of practitioners were actively experimenting with it. And that’s seems about right, to me.

 

And I think at the moment, you’ve got to, you know, if you want to future proof your career, you’ve got to, as we’ve discussed, experiment with these different technologies and understand them, you’ve got two assertive opportunities, I’d suggest, within your day job, the first one is to be the person who informs management about where these technologies are going. I mean, AI was stuck within stock, it was typically within research and development or, you know, innovation within an organisation before November 2022. After that, you know, it’s very firmly become a management issue. Yeah.

 

And, you know, management’s trying to get to grips with the impact of this on, you know, business models on reputation on ethical issues. So, you know, that’s naturally the domain of the public relations practitioner in understanding in the horizon, scanning the public sphere to, to help organisations figure this out. So, you know, this is an opportunity to engage with penetrant, just as we did during the pandemic, you know, so So it’s an opportunity to get ahead in your career. Second one is, you know, the opportunity to understand the impact on workflow and be the person within your organisation that people go to help and assistance to understand that a lot of agencies, not so much comms teams, but agencies now have built innovation teams to look at these tools start looking at workflow, and they have been hacked days and so forth, to try and get ahead of this.

 

PS 38:18

Just to finish off, and do you feel broadly optimistic about AI as a whole? And then about its impact on the comms industry and where that’s going.

 

SW 38:30

So I’ve got another 10 years of career. To your point. I’m so yeah, I’m naturally going to say really, I think it’s, it’s a, it’s a disruptive innovation, we’ve seen those many times that you know, the PC, the PC to the mobile phone, the internet, social media, it’s light that we don’t yet know how big it is, and what the potential impact of it might be. Undoubtedly, you know, it will have a huge impact and it’s showing already the potential to be disruptive. It provides us with an assertive opportunity to go and engage with management but also to change your workflow. So yeah, I’m hugely optimistic.