17 min read
Generative AI for Marketing - Maintaining Trust and Building an AI Policy
Wendy Covey : 6/8/23 10:00 AM
Is your team using AI tools? Here’s everything you need to know to maintain trust and authenticity.
In this week's episode, Morgan Norris, Senior Brand Strategist at Trew Marketing, shares everything she’s learned about Generative AI. Morgan has become our resident expert on all things AI tools and has sifted through dozens of tools to find out which ones are most valuable to marketers. Morgan shares the key features of the best performing tools for tasks like idea generation, editing, and content writing.
We also highlight the ever-present need to stay authentic and trustworthy when developing and publishing content – something we still need human oversight for. Last, discuss what liabilities could come up if you’re using AI tools incorrectly and the direction these tools could take in the future.
“If you’re not staying on top of generative AI technology, your competitors are. Tools are evolving so quickly. There are constantly new tools that are popping up that help you work smarter, work more efficiently, and free you up to let you focus on things that are going to grow your business.”
- Morgan Norris, Senior Brand Specialist
Resources
- Webinar: Improve Your Prompting to Use Generative AI for B2B Content Development
- Generative AI for Content Marketing
- How to Create an AI Policy for Your Engineering Company
- Morgan Norris on LinkedIn
Transcript:
Well, for those of you that are regular listeners to this podcast, you'll know that we made a pledge to help keep you up to date with everything generative AI, and today's episode is no exception. I'll be joined by Morgan Norris, a regular on this show, and we'll be talking about new tools that we're experimenting with, how to use AI while still being authentic to your brand and creating content that's trustworthy, and also the importance of creating an AI policy for your company. Let's do this.
Welcome to Content Marketing, Engineered. Your source for building trust and generating demand with technical content. Here is your host, Wendy Covey.
Hi, and welcome to Content Marketing Engineered. On each episode, I'll break down an industry trend, challenge or best practice in reaching technical audiences. You'll meet colleagues, friends, and clients of mine who will stop by to share their stories. And I hope that you leave each episode feeling inspired and ready to take action. Before we jump in, I'd like to give a brief shout out to my agency, TREW Marketing. TREW is a full service agency located in beautiful Austin, Texas, serving highly technical companies. For more information, visit trewmarketing.com. And now on with our podcast.
Hey, everyone, and welcome to another episode of Content Marketing Engineered. Today I'm welcoming back a regular guest on the show, Morgan Norris, who's the Senior Brand Strategist and Content Strategist at TREW Marketing. Welcome, Morgan.
Hi, thanks for having me.
What's it like today in DC?
It is going to be so hot. We've had great weather, so nice. City is full of tourists, lots of traffic, but today it's going to be Texas weather. So in the yeah, there you go.
Well, here in Austin, the tourists are present as well, but that's okay. I'm out in the Hill Country, so I'm out of the fray of all that, which is nice. Well, today we're doing another episode of all things generative AI. And this is, I think, the third or fourth time now that we've done this early June to give context to when we're recording this. And Morgan, just maybe before we get into all the updates that you have for us today, why is it important for engineering and industrial marketers to keep up with all things AI as they come about right now?
Yeah, great question. So it's important because partially, if you are not staying on top of this, your competitors are. And so tools are evolving so quickly that there are constantly things popping up that can help you just work smarter, work more efficiently, free you up to focus time on things that are going to grow your business, maybe automate some tasks that your staff doesn't love doing, or automate things that normally just take a lot of time. And so for us, I think staying on top of these tools means staying on top of what's new. And we're trying to keep up with these regular updates to give people the resources they need. And it's just helpful to kind of stay on top as you move forward, because it's never the right time to get involved and to get up to speed. That was a month ago, but the second best time is today. Just jump in and start exploring some tools. So we want to give people the resources to do that good.
And feel a little bit more in control rather than fearful of what's out there and where things are going, because there's a lot of good that can come out of this. And then some concerning stuff that needs to be mitigated, which we'll talk about today as well. Yeah. Good. Well, first topic, let's talk about how we can use AI to reach new audiences.
Okay, this is a topic that I love. So as these tools are evolving, what happens is you think about technical audiences and engineers and clients that we work with. We're spanning from new grads to people who are approaching retirement and these types of people that span of people consumes information in really different ways. And so a few things that are coming up are kind of video tools. So a lot of times, we'll be working with maybe an organization that's been around for a longer time. They're trying to attract new talent or their customers or younger engineers. Man, they want some video they know. Our research shows that younger engineers are watching more video for work, especially in comparison to older engineers. But sometimes that video just feels really hard to create. It feels like a huge burden. You've got subject matter experts who are incredibly wise and experienced, but they don't want to be on camera. And so what does that look like? So, there's a few tools that I've looked at recently. You did a demo of a tool called Gloss AI. It's just Gloss AI is the URL, but what they do is they can take written content and create some kind of broll type of video out of that.
It definitely is harder for super technical topics, just like we all know. Finding imagery and broll that acutely shows kind of what it is that you're talking about. It can be difficult, but you can get some kind of abstract video in the background with the voiceover through their tool, and so that's really helpful. There's another one called Synthesia, and I actually got served up a video like this recently. I picked up a medication from the pharmacy, and then I got a text, and it said, hear what the pharmacist has to say about this video. And sure enough, I clicked through, and it's a person there talking, and it's all lined up. But after using a couple of tools, exploring a couple of tools like this, I know that that is an AI created video. However, it is a person talking you through this kind of medication, and you could see the benefit for that of even, like, help videos or things like that. There are people who will search through help forums to troubleshoot your products, but there are other people who they just want to talk to somebody, or they just want to see kind of how something works.
And so with synthesia, you can actually just literally type in words. You pick out a person. You pick out are they in a lab coat? Are they in kind of work attire? And they will then read that script, but it's all auto generated. And so some of those tools are more expensive, but they're not as expensive as going to a production studio or having a professional videographer kind of come and do all this stuff on site. And so you're able to take the content that you already have and change formats to formats that were previously felt really unreachable for you.
I got you.
Boy.
I'm not it sounds so efficient to be able to do that. And I'm assuming that because you had a good experience when you saw it with the medication, that it doesn't have that weird look to it.
Like, some of it it doesn't.
Like Wally pictures where people have distorted eyes and stuff.
It doesn't okay, so these look like.
Real humans, even though they're not. It's an AI generated talking head. What do you think about the fact that you're sort of representing this person as an employee? Or maybe you're not, but you know what I'm saying. It's not really a person that works for your company. It's an AI generated person.
I think you've got to have a disclaimer. You've got to have a disclaimer on there. Like, when you see commercials, and it's like, these are actors. These aren't actual people using the product.
Making claritin and then jumping through the field of flower.
Yeah.
The disclaimers are good analogy.
The disclaimers are helpful. And we'll talk about that. I want to talk about that at the end, about kind of how you handle AI as a company and how you let people know what you're using.
Fair enough. Okay, good. Any other tools on your radar?
Another thing that we've used that I would really recommend, especially for technical audiences is a tool called chat PDF. So it takes a chat PDF.
Here's.
Why I like it. With Chat PDF. You take a PDF, you take something that you have already created, some knowledge base that's a white paper or an ebook or it's got all of your expertise in there. And then you can input that into chat PDF and start asking for specific kind of excerpts or summaries or a bulleted list or take what we've been waiting for. Yes. Take this highly technical content and summarize it in 200 words at a 10th grade reading level.
Yeah, the level of okay, I love it.
Right? So all of a sudden, you can take that really technical paper that you wrote when your mom asks what you do, and you can put it in language that a non technical person could understand. So it's things like that, I think, that are particularly valuable for our audience. Because what's so unique about kind of B to B technical communication is it's all of this IP that's stored within the minds of your subject matter experts and within your company. And that information a lot of times is actually not sitting in Chat GPT. I had a client recently say they're like, well, can't we just have Chat GPT write something about this? Well, what they were trying to write about was basically research that they have discovered recently in our productizing. Well, that's IP, it's patented IP that was even very recently patented. So Chat GPT stopped collecting data a couple of years ago, and so it doesn't hold even the source material to describe what their technology is. However, if we do the due diligence with them and write up the technical paper now, we can use that as our base source material and start pulling pieces from our material.
So it's a much better approach than trying to ask a tool to write something that hasn't been created yet.
Yeah. And then I just keep hearing more and more about people running into false sources. So if you do ask Chat GPT and it's not out there, there's a chance that we'll just make up sources and make up stuff.
Wow, okay. We haven't even talked about this before, but I actually had a tool, right? I was kind of like brainstorming. So I'm using Chat GBT to just give me some just kind of get my mind flowing, right? What was super interesting is it served up a couple of specific products as answers and I stopped. I think I was even asking it about using AI tools or something like that, and it inserted a product that I then Googled, which was a specific AI technology for it's like an older technology for voice messages or something like that. And it just gave me pause for a second because I thought, for them, that's potentially an opportunity for product placement or am I being advertised to through this process? So, anyway, it's not something I've completely explored, but it was enough that I was like, I don't like that. Right.
And if it was product placement, now we've lost the transparency of that that we used to have in normal Google searches.
Right. Anyway, so I don't think it was I haven't heard anything like that, that they're doing that, but it just made me stop and go, oh, that's definitely a possibility to come. Right. You think of where these things are headed and where they're going to go and yeah, it would be like if Google served up sponsored content that didn't say sponsored on it.
Yeah. Well, speaking of Chat cheap et, I've heard more and more people using it and Bard for Brainstorming buddy, right? I hear a lot of people adopting that. That's an exciting use of those technologies. Just curious in your experience and I think Jamie, I know she's been experimenting a lot, which tool do you like better and why and how are they different?
Yeah, Jamie really likes Bard and I trust Jamie a lot, so I'm kind of shifting that way. I think that my familiarity is a little bit more with Chat GBG, though, we're also starting to use some a little bit stronger tools like Writer, which is a content development tool. And while Chat GPT can help kind of answer questions or give definitions or summaries, writer is one that is designed to kind of work how a content developer works. So it'll sort of assist you along the way to topic ideation and then forming an outline and then it'll kind of make suggestions along the way. And one thing I really like about Writer is they just implemented a feature where if you've got a specific, if it's serving you up some information and there are data points in there, it'll highlight those and go, you need to fact check this, like you the subject matter expert needs to this. And so I think barred Chat GBT for an ideation kind of pool for sure. But these more refined tools for specific purposes are getting, they're getting better and smarter and even a little bit more human as they evolve.
Saying things like that, you need to fact check this or it'll produce something, it'll produce a few sentences and then it'll actually call out its own writing and say this sentence should be streamlined, it's too wordy or something. It's a cool resource in that way and it's not presenting itself as the be all, end all for content still requires a human.
Yes. And is it fair to say when it comes to original content development for, let's say, a white paper or case study, you're not necessarily saving time yet using these tools for that original piece.
But.
You have, I don't know, help making it more creative or making it a stronger piece in some way.
Yeah, I think some of that research gets streamlined. I think the time saving comes on the back end when you're then creating a blurb for an e newsletter or a social media post or something like that, that's where a lot of the time savings comes in, the derivative content. Yeah. Because from a research perspective, even the tools are good for Ideation. But at the same time, if you only used Wikipedia for your research, I would say you're probably not going to be very well read on a topic. Definitely still going out there to get different opinions and stuff is helpful.
Yeah. So where are some of these tools when it comes to search optimization?
Yeah, so that's one thing that's helpful. Writer, I'm not getting search optimization help necessarily from Chat GBT or from Bard, I would say. I think we'll see those probably evolve though. I can't totally get my mind around the kind of eating your own dog food cycle of like Bing having Bard enabled. But they are search engine, so I'm.
Not sure you want to surface on the very tool you're asking to surface on.
There's that, but the writing tools like Writer or like Jasper will help you come up with kind of like good keywords or say they'll make recommendations on. You can put in terms that you're kind of trying to focus on and then it'll help you integrate those in a little bit more of like a natural language type of way. Not overly creative though. What I'm finding with technical content is a lot of times you'll put in a really technical term or like more long tail type of term. And then the suggestions it'll come up with are if it was radio frequency signal generation or something, it'll come up with what is radio frequency signal generation? How to use? Yeah. And so you're like, okay, that wasn't awful.
So they're not a replacement for our tried and TREW tools like SEMrush and Maz and some of those specialized SEO tools.
Yeah, not yet. I think we'll continue to see those integrations though. I think it'll get and it's not going to be like I think over time the good writing tools and the good SEO tools will just start to come together more naturally for a better user experience. And that is Jasper is doing that somewhat. They've got an integration with an SEO company called Surfer. I don't think it's as good yet for the kind of deeply technical stuff that we work on, but I think probably from a consumer perspective, that's a pretty helpful tool right now.
Okay, good deal.
Yeah.
Well, Morgan, as companies start to adopt AI, whether it's for content development or programming or all sorts of other functions, what do we do here about being a trusted brand and being authentic to who we are? How does that live alongside AI technology?
Yeah, so a couple of things, and I will keep saying this, but I think that as time goes on, we are going to crave individualized, individualized but authentic content because everybody can go produce an answer with these tools. Right. But I think that we are going to continue to want to see people that we know we can trust. We're going to want to hear people talk. We're going to want to want to be in the physical presence of people who are experts in their field. And so one thing that keeps coming up is this kind of trust and authenticity. So I think this goes a lot of different ways. So I think back to doing.
We.
Would do a website for a manufacturing company that had a lot of their manufacturing at their in house facility in the US. And those clients were so proud of that because they had so much control over what was happening in their manufacturing facility that a lot of times they would want a video inside their manufacturing facility because that spoke volumes to their customers to show. We are doing all of this stuff in house. We've got our eyes on it. And look, here's a video. You can come visit our manufacturing facility. You can walk through it. And I think that that type of authenticity and transparency is going to continue to be important. So for designers, this might look like a time lapse of pad drawings, right. For writers, this looks like your drafts. It looks like seeing how you've input your own voice in there, it looks like quotes from subject matter experts at your company. It might look like for coders, this looks like kind of screenshots of how you've gotten through the process to get to this really authentic end result. And you've showed human oversight at each point in the process. And so I think that that's going to be important.
And people need to kind of every AI tool you use, you need to take a step back and say, what's the kind of human oversight here? And how do we show that we're being authentic in what we do? And I think that'll continue to be really valued as we go forward.
I go back to what you said about how writer says fact check this and fact check this, how you need to be very diligent about knowing where your claims come from, where your information comes from, and doing that research, knowing your sources.
Yeah, well, so then that speaks to kind of creating internally within your company some kind of AI policy. And you can come at it from a lot of different angles. But if you've got a large organization or small, what if you find out a year from now that one of your teams or even just an individual employee has been using some tool without enough human oversight? And then now you've created problems that surface later. So by kind of creating an internal document of this is how we handle AI within the company, is something that I would urge every company to do. We published something recently that a blog post recently that people can even just use as a framework. But the first things you've got to look at are the kind of legal ramifications the government has asked us to consider. Data protection and privacy, IP liability, discrimination, bias, and then security in any endeavor. Right. But we look at how those relate to AI. We've talked about before. Solely AI generated works can't be copyrighted. We've talked about the fact that most AI generative AI platforms require a terms checkbox that states something to the effect of whatever you put in this tool can then be accessed by the tool.
And so I haven't heard horde stories of things like that happening yet, but we are very click happy to just yeah, sure, whatever. Get me through the terms box so that I can get to what I want. But you think about that and you think about, well, what if you're putting in product information or IP that is not out there in the world? I think about public companies kind of putting material news in there that's not yet announced and things like that. We will see some stuff like that that becomes a problem. So just that very first one, data protection and privacy, just taking a step back and going, oh, as a company, let's make sure we're not using a tool. We're not putting anything in a tool that we wouldn't put out in the public. Right. It's things like that. And then kind of writing a corporate statement that answers the questions of how you use AI, how it contributes to what you offer to customers who's accountable for anything created by AI? And mind you, that answer should be the company. If you're giving something out, if you used an AI tool to create it, the liability is still on you.
You're claiming that can't go back and blame a tool. What limits do you have on AI, and then what level of transparency do you have with the tools that you use? So those are kind of the questions that you could sit around a table with marketing, with internal communication, with engineering, with sales, and get everybody on the same page so that when new tools come up or new processes come up, you at least have a framework to work through to decide what to do. Yes.
And I don't know why, but this brings to mind. Recently there was an attorney that got in deep water and if you read about this, where he used AI to come up with his whole whatever it was case consideration, and it had fake precedents listed in it that the AI tool just served up, and he just ran with it, didn't fact check it. And now we'll be kind of the first attorney, and I don't know what's going to happen to this person, but it's okay. So the company, were they aware he was using AI to do his job? Was he using it appropriately? Obviously not. Was it fact checking? It not enough human intervention. And look at the result. And of course, to your point, that law firm can't just say, well, it's AI's fault, right, because they served us up false information. Absolutely. Okay. Well, as a marketer, obviously they're one component of this larger AI policy. But if your company doesn't have one today, the marketers can certainly lead the charge or be the catalyst to say, we need to get something out there and help communicate that.
Yes, for sure. And that we kind of publish something on our site to that point. And it's definitely tweaked towards kind of marketing and how we use AI tools for marketing. We're also really transparent about what tools that we use. So I think this is interesting, and I think as time goes on and in the name of transparency, being able to say, we use a tool called, like Happy Scribe to transcribe this podcast, or we use things like that, there might be tools that you're already using that you don't even know. But we've said kind of what we are and aren't using tools for and what those tools are. I don't think there's any hindrance in giving that transparency, because when clients start to ask questions, when your customers start to ask questions, they're wondering, I think. And in my experience, clients are wondering kind of, are you using a tool to do something that now I could be doing myself? Or are you using a tool to do something way faster or cheaper? And now I'm overpaying for it. And so I think we want to get ahead of those and say, here's what we do.
Here's how we use them. Here's what we don't use them for. As we start to kind of adjust and move forward, you shouldn't be afraid to say that. If you are, you need to evaluate what you're doing and how you're relating.
To your clients trust and authenticity.
Right.
We'll circle back to that.
Yeah, for sure.
Good. Well, Morgan so as I mentioned, it's early June right now, so depending on when someone is listening to this, we have an upcoming webinar that would, I believe, be on demand afterwards. But tell me a little bit about your webinar.
Yeah, so for the webinar.
We'll talk.
About kind of some of this stuff, putting policies in place, and then we're going to talk about prompting and how to use generative AI for B2B content development. The webinar is on June 15 at noon Central Time, and it'll be live. We're going to do some live prompting, which should be fun because a tool is going to fit out something different every time you use it. You can see how I do it on the fly there, but what's happening is the IP that you insert into a tool or the prompt that you insert into a tool, really becomes like your IP. It becomes kind of your source code for what is going to come out of it. And so crafting, really deep, specific prompts will start to help you get better content out of a tool or even a better starting point or better brainstorm, buddy. Right. If I bring in an intern, that's never it's their first day, they're probably not going to have great ideas. But if they've sat in meetings for the whole semester and then they come in and I want to hear what they think, they probably have a lot better ideas.
So we can kind of give some of that information to a tool and get better out of it. So we'll work through that.
Great. Okay, well, I will put a link to that webinar in the show notes and otherwise. I know we have a live document that you're keeping up to date, a generative AI guide for industrial marketers. So I'll put a link to that and anything else?
No. From that guide, you can also click out and see our policy. And I would just encourage people feel free to use that as a guidepost. We used the Marking AI Institute's guide as an initial guidepost, and so that's licensed under Creative Commons. You can use it, but I would go from there. Just start somewhere great.
All right, well, thanks for your time today. Morgan thank you.
Thanks for joining me today on Content Marketing, Engineered. For show notes, including links to resources, visit trewmarketing.com podcast. While there, you can subscribe to our blog and our newsletter and order a copy of my book, Content Marketing, Engineered. Also, I would love your reviews on this podcast, so please, when you get a chance, subscribe and leave me your review on your favorite podcast subscription platform. Thanks and have a great day.
Wendy Covey
Wendy Covey is a CEO, a technical marketing leader, author of Content Marketing, Engineered, one of The Wall Street Journal’s 10 Most Innovative Entrepreneurs in America, and she holds a Texas fishing record. She resides in a small Hill Country town southwest of Austin, Texas, where she enjoys outdoor adventures with her family.
About TREW Marketing
TREW Marketing is a strategy-first content marketing agency serving B2B companies that target highly technical buyers. With deep experience in the design, embedded, measurement and automation, and software industries, TREW Marketing provides branding, marketing strategy, content development, and digital marketing services to help customers efficiently and effectively achieve business goals.