S5 E3: Practical and Ethical Implications of Generative AI with Nick Leighton of NettResults

Adrian McIntyre: 0:00:01

From the Public Relations Global Network, this is PRGN Presents. I’m Adrian McIntyre.

Abbie Fink: 0:00:07

And I’m Abbie Fink, president of HMA Public Relations in Phoenix, Arizona and a founding member of PRGN. With public relations leaders embedded into the fabric of the communities we serve, clients hire our agencies for the local knowledge, expertise and connections in markets spanning six continents across the world.

Adrian McIntyre: 0:00:28

Our guests on this biweekly podcast series are all members of the Public Relations Global Network. They will discuss such topics as workplace culture, creative compensation and succession planning, the importance of sustainability and environmental, social and governance programs, crisis communications, and outside of the box thinking for growing your business.

Abbie Fink: 0:00:48

For more information about PRGN and our members, please visit prgn.com. And now let’s meet our guest for this episode.

Nick Leighton: 0:00:59

Hi, my name is Nick Leighton. I’m the founder and CEO of NettResults in the Middle East. We are a PR agency specializing in crisis communication, storytelling for clients who are outside the region who want to get publicity within the Middle East region.

Abbie Fink: 0:01:13

Nick, you and I have chatted on this podcast in the past, and we’ve done it in conversations where we’ve been together personally about artificial intelligence, AI, as we’re calling it now that it’s become a little bit more commonplace and really the impact, both positive and keeping ourselves aware of some of the challenges that might impact our businesses with. But it’s evolved since we’ve started these conversations.

Abbie Fink: 0:01:39

So to kind of set the tone for our conversation on Generative AI, tell me a little bit about your viewpoint on where we are with this concept and how it’s moving and maybe how some of us are maybe feeling a bit more comfortable with using it and talking about how it’s benefiting the work that we’re doing and sort of set the tone for where the rest of our conversation will go today.

Nick Leighton: 0:02:05

Absolutely. It’s a fast moving, exciting space. So previously we had conversations about AI. Today we’re going a little bit deeper. We’re going to talk about Generative AI or Gen AI. And that referred to a class of artificial intelligence where algorithms are designed and generating new content that could be images text music video and it’s all based on patterns learned from data so this is how we produce things using ai these algorithms they work by training on a vast amount of data and it learns the patterns and the relationships and it uses this knowledge to generate new original content. So that’s what Gen AI is all about. And it’s grown so quickly, so rapidly for us that now this is something that is very usable in the world of marketing, in the world of public relations.

Abbie Fink: 0:02:55

So when you’re saying that it’s original content, so we’re still telling it to create something, it is just not pulling from things that have already been previously available. So what is our responsibility then for acknowledging that we’re using this technology, this important tool in helping craft and create this content that we are, you know, working with our clients who’ve asked us to do this, but what’s our responsibility in sort of managing the use of that tool and the expectations of the content that it’s creating?

Nick Leighton: 0:03:31

That’s a great question. So without getting too deep into the technology we are creating original new content so this isn’t if you plugged it back into a system and you know ask where’s the plagiarism where are we copying this from it’s not going to find that because it’s developed patterns and it’s creating something new so today people are pretty scared of that kind of thing so there are certainly publications where you could submit work and you will be ticking a box saying this is not Gen AI related content and the ice is original content that I created and artificial intelligence did not create. That’s where we stand today.

Nick Leighton: 0:04:07

But the question is, how’s that going to develop in the future? So if I use Microsoft Word to create some documents and I have a spell check on there, is that good or bad? If I use a system such as Grammarly to correct my grammar, is that good or bad? So I think what I believe we will see in the future is that the fact that something may be generated using AI is not going to be such an evil thing. No. The thing is, you can’t just throw something into an AI system, ask Gen AI to create something brand new and then use that.

Nick Leighton: 0:04:40

We’re certainly not at that stage today. But we are at a stage where this is a fantastic co-pilot for the professionals in the industry. It allows us to generate new ideas, new content, localized ideas. There’s lots of ways we can use it.

Nick Leighton: 0:04:55

And then the question, which I think you really alluded to, is ethically, who do we say that we’re using AI with? So a general statement about an agency, yes, we use Gen AI, I think would be the same. Be open and be honest and be transparent. But, you know, let’s face it, do we say that we use grammar correction or spell checking today?

Abbie Fink: 0:05:18

I love the idea of referring to it as a co-pilot and I sort of envision this room of humans sitting around brainstorming and then we have our laptop there participating with us as our co-pilot, right, and sharing and adding and enhancing, you know the conversations that we’re having around you know the content that we’re trying to create or the ideas that we’re trying to generate and I think that for me has really been whichever version of that we’re using I think has been the best way for me to explain it to even those of us that are just sort of getting comfortable with the idea of using it is that it is a tool and it helps us and gives us thought starters it gives us a place to build from that it is, at least now, not a complete product, because there are things that need to be added.

Abbie Fink: 0:06:12

There is the nuances of the human connection to it that need to be incorporated. But it is certainly a great place to help get started. And as you said, sort of as your co-pilot, it’s a partner with you in your content creation and your creativity. And are we As a consumer looking at things now that are being written and shared, you know, whether that’s in traditional public relations ways, media relations and such, or advertising, are there ways to understand that it may have been Gen AI created? Or is there an obligation to the consumer to share that, that this is coming from technology and not the human component completely?

Nick Leighton: 0:07:02

Yeah, I think there’s a couple of things to think about there. And certainly it’s fast moving. So by the time that this issue is aired, I’m not sure what the rules and regulations are going to be. I can tell you today as we record it, certainly some of the social media platforms are suggesting that advertisers in certain areas need to define if they’d be using AI. So political campaigns, I think that’s going to be interesting. Just after this is aired is going to be a highly political time in the US for sure. And indeed in many countries around the world. So we’re going to see how that will develop. But my question really would be, do we care?

Nick Leighton: 0:07:37

I mean, if we’re talking about an image and that is portraying someone that we know who’s in the public side, yes, we’d want to know that’s been artificially created. But if we’re talking about text, does it really make a difference? If I bring it back from consumer to business to business, if I’m working with a client, And does it really matter if I personally have written the press release or one of my colleagues have written the press release? Is there a difference there?

Adrian McIntyre: 0:08:00

You know, I think that’s such an interesting question, Nick. Do we care? And I want to take a slight detour here for a second.

Adrian McIntyre: 0:08:07

Because as a cultural anthropologist and linguist, I have studied things about language and how it works and how humans use language. And I have a thought about AI that I don’t hear a lot of people talking about, but I want to throw it out here in this context. It relates directly to what you said about do we care now and will we care in the future? A classic work in linguistics is the book Course on General Linguistics, this guy Ferdinand de Saussure.

Adrian McIntyre: 0:08:34

And he was trying to, first of all, invent the science of linguistics, but he was also trying to understand how language works. And he said something really interesting. This is in the 19th century. He said language essentially has two axes. Like if you think of a grid, horizontal and vertical, the horizontal axis is how words follow each other in a meaningful order. It’s linear, right? Sentences are linear. You start at the beginning and you read through to the end. And that’s part of how they make sense.

Adrian McIntyre: 0:09:00

But there’s also a vertical axis, which is the axis of association. It’s words that are interchangeable with each other, words that are a cluster around which there is meaning. Now, this is all very nerdy, but if you think about it, all that the large language models are doing is what humans are already doing with language. In other words, we know that there are patterns that are correct, and part of becoming a native speaker of a language is to become fluent, means you know how that language generates meaningful utterances, sentences.

Adrian McIntyre: 0:09:35

And the only thing that I think is upsetting people is for the longest time, we thought we were the humans were the only ones that could do that. And now that computer algorithms are essentially doing the same thing. They can, with a statistical model that’s very elaborate, it’s an algorithm, they can predict what words should go in what order and create these meaningful statements. Now, this is just about the written stuff, images, maybe a whole other thing, but it really strikes me as something we have to wrap our heads around.

Adrian McIntyre: 0:10:07

Maybe Gen AI right now is like a really sophisticated, fluent 12-year-old. And it’s able to put together like a precocious language genius of a kid, right? Sort of Mozart of words, if you will. It’s able to put together these sentences that kind of blow us away. We’re like, wow, how did it do that? Well, it did it the same way you do it. It knows all the words in a particular vocabulary. It knows which ones should go in which order, right?

Nick Leighton: 0:10:34

Right.

Adrian McIntyre: 0:10:35

So at some point, we might just have to acknowledge that this is no different than a hammer or a calculator or some other tool. It enhances our ability to do creative work, doesn’t threaten it, doesn’t replace it. And at some point, we may not care whether or not the words came from an algorithm or came from a human. I don’t know.

Nick Leighton: 0:10:57

Right. No, let’s get even more geeky because now you’ve got my linguistics thing going because I studied linguistics as well. So back in the 90s, there was a program that was developed called ELIZA. You can still find it online. You can use it for free. And you can type in a question. And ELIZA worked as a type of therapist. So you would say, I’m scared of spiders. And it would say, reply back to you in text, why are you scared of spiders?

Nick Leighton: 0:11:20

You go, well, I think those spiders are black and icky. Go, hmm, what about black really offends you? And then it would ask, it would reply to your questions. Now, ELIZA had no intelligence. It wasn’t AI. It was purely linguistics and it was mathematical. Because you could go and you could say, I have a phobia against green. And they go, why do you have phobia against green? But you could just as easily say, I have a phobia against XYZ.

Nick Leighton: 0:11:46

And go, why do you have a phobia against XYZ? It was just playing with the language structure. So Gen AI has got to the next level, where it actually is creating something new, rather than just developing a response based on a structure. It’d be harder to go into ChatGPT, and say, I have a phobia against XYZ. They’ll go, what does that mean? Whereas, you know, you’d have to give a real sentence for it to come back with something knowledgeable back to you.

Abbie Fink: 0:12:16

Well, you know, and Adrian, I think something that you said, and this has been sort of the question that we all are getting asked, is this going to replace us, right? Those of us that are in the business of using words and creating content and visual and things. And I don’t believe that it will because we still have to have the human component, we need to understand it and we need to get comfortable with it in the same way that we have with anything else that we’ve maybe pushed back against and said, this is never going to be something that I’m going to do.

Abbie Fink: 0:12:49

And now it’s a daily part of our lives because we can’t function without. And this is likely going to become one of those things. But I do think that there is We’re smarter now, we the humans are smarter now than we were 10 years ago, 12 years ago, 15 years ago when certain technologies came into the workplace and we’re smarter about how we use them and we’re smarter about maybe some guardrails that we’re putting around them just to manage that within our own businesses and the outcomes that we’re using them for.

Abbie Fink: 0:13:25

And we have an opportunity, I think, then to embrace this and then embrace it in a way that makes the most sense for the work that we’re doing. We’re talking about it here in the context of communications, but in all likelihood, this has been part of other industries as well. There have certainly been these types of technologies where we’ve been able to, we just might not be as aware of them as they’re not part of our daily lives.

Abbie Fink: 0:13:50

But as businesses that are using this and using it well and understanding it, but we’re also counseling our clients and other businesses then about the role that Gen AI can have with them. And it’s, I think on us as to the extent that we are considered experts on this to help them understand how to use this responsibly, how to use it in an ethical, honest, transparent manner. So what are we doing as advisors to businesses to ensure that this tool that we have is being used appropriately and ethically and responsibly.

Nick Leighton: 0:14:30

Perfect. So before we get to the ethical understandings, let’s look at maybe three typical challenges or risks that we have. So let’s see how we can overcome them. So I think one of the challenges is to ensure that the quality and your authenticity of this content, and that’s kind of what we hinted at a little bit before, it may not always meet the same standards as human-created content. So we need to have that interaction. And we could talk about real examples later if you like.

Nick Leighton: 0:14:55

There’s also a concern that there’s potential biases that, based on the training of the Gen AI. Now, we as users can’t control that. There can be high standards on people who produce the Gen AI systems that we’re working off, but that’s not something in our control. That’s just something that needs to be aware. If you give bad learning parameters, then we’re going to get poor results back. And then the other risk is that potential misuse or abuse of Gen AI, such as creating fake news or spreading misinformation.

Nick Leighton: 0:15:24

So there’s different levels there that we need to be concerned about. So to ensure that we know what the risks are, we can now think about what are the ethical considerations, responsibilities that we should have as businesses. And first of all, is just being transparent about its use. I think that’s going to be important. And we tell our clients we’re using it. Not every single time we use it, we just tell them at the beginning of the relationship, we use it, it helps us, it’s our co-pilot, to go back to that time again.

Nick Leighton: 0:15:49

And then of course we have to vet all the information that we’re using very carefully. So I’m going to make the assumption that the systems that we use, the Gen AI systems we use, are reputable and there’s no bias within them. But I’m going to check them. So let’s talk about some real examples. So today, all examples from today from my team. So one person on my team was asked to do a SWOT analysis of whether a client should be doing a launch event. Great. So they had some ideas. They put that into a Gen AI system. They go, here’s the question. Should we be having this event? here are my ideas. Can you expand upon this for us?

Nick Leighton: 0:16:26

Results came back, other ideas. The team member that I work with then went through and go, well, no, that’s clearly not relevant to what we’re talking about here. That’s something I hadn’t thought about. Let me put some thought behind that. So the finished report, that sort of analysis, which could have taken over an hour or so, was obviously done much quicker, but also allowed other input that we may not have been familiar with. So it allowed us to do things.

Nick Leighton: 0:16:50

Second use from today was a client had one piece of material and they wanted to localize for a different country or a different target audience. That could be done very quickly. And it allows our team to get input and go, yeah, yeah, this part’s relevant. This part isn’t. This I need to be careful about. One thing that they suggested when we changed something was to spell something out phonetically because in the next language, it didn’t really translate. We may not have seen that straight away.

Nick Leighton: 0:17:16

Also how to change currencies. We thought everything’s going to go out in US dollars, but for this other market, another currency was going to work better. All these ideas that we wouldn’t necessarily have caught the first time around, we had straight away. Great stuff, but it had to be in interaction with someone else.

Abbie Fink: 0:17:34

Well, and it relies then on the knowledge and expertise of your team to fully grasp what the end result needs to be and then evaluates this information against those directives and such that we’ve been asked to do. So, you know, and can’t take out the human component of all of this.

Nick Leighton: 0:17:55

Exactly. So that means putting in guidelines within your practice. Absolutely right.

Adrian McIntyre: 0:18:01

Nick, you’ve spoken about your team’s adoption and several use cases. What would you recommend for young professionals who are thinking of joining agencies or their in-house communicators in terms of embracing and learning new skill sets, new mindsets? As your team has thought through some of these things, what have been some of the speed bumps on the way, some of the obstacles you had to navigate around? What have they had to learn along the way? And what would you recommend others to do?

Nick Leighton: 0:18:31

Okay, well, first of all, let’s start by familiarizing yourself with the basics of Gen AI, how it can be applied in marketing and PR. That’s what we’ve talked about today. Then start experimenting.

Nick Leighton: 0:18:41

Understand what the tools are, what the platforms are, understanding their capabilities and where their limitations are. That’s really important. And then stay informed about developments because it’s such a fast moving area. We need to stay up to date with all that. Finally, I think prioritize that ethics and the responsibilities to use Gen AI in a way that aligns with the values of our agencies and our teams, but also with the audiences. So there’s a number of elements there. Being aware, getting trained, staying up to date.

Adrian McIntyre: 0:19:13

Thanks for listening to this episode of PRGN Presents, brought to you by the Public Relations Global Network.

Abbie Fink: 0:19:18

Episodes are also available on our website, along with more information about PRGN and our members, at prgn.com.