Generative AI Update - News, Tools, Prompting, Crafting A Company Policy

Hear the latest updates on generative AI for marketing, including tools we're kicking around, features we love, and prompting best practices for creating AI-assisted technical content.

Morgan Norris, TREW Brand and Content Strategist, is on top of how AI tools can enhance our technical content writing process (not derail it). She's been researching and piloting various tools and staying abreast of the ever-evolving legal and cultural implications of generative AI.

In this episode we cover:

• Marketing AI news, including a statement and guidance from the U.S. Copywrite Office
• Generative AI tools we are trying out (see full list below)
• Generative AI functionality that benefits us most as technical marketers
• Best practices for prompting AI tools during ideation, research and writing
• How to create a company policy on generative AI

Morgan will host a webinar in May to take a deeper dive into many of these topics. If you have questions or topics you'd like her to cover, drop her a line on LinkedIn. 

Resources

• Past podcast episode: Generative AI for Industrial Content Development - A ChatGPT, Writer and Jasper Showdown
• An Industrial Marketers' Guide to Generative AI
• U.S. Copywrite Office - Generative AI Resource Page
• Writer
• Jasper
• ChatGPT
• Demandwell
• Canva
• GlossAI
• Marketing AI Institute
• Morgan on LinkedIn

Transcript:

Well, the world is changing rapidly when it comes to AI for Marketing, and we promise that we would help keep you on top of the latest and help you understand what to use, when to use it, and how to use it. So this episode will be an April roundup of the news of AI for Marketing recommended use cases, tools that we've been playing with, functionality we're excited about, and best practices for using generative AI to write technical content.

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. I'm joined today by our regular guest, Morgan Norris, Senior Brand and Content Strategist at TREW Marketing and very rapidly becoming our agency expert in all things AI for Marketing. So, welcome to the show, Morgan.

Thank you. Thanks for having me back.

Yes, and thank you for raising your hand to be our agency expert on AI. It is changing ever so rapidly and is requiring a lot of research and just staying on top of things. And so we had promised in a past episode that we would have this be a regular conversation and just have updates whenever we felt it was necessary on the podcast. And here we are with an April roundup for AI for Marketing.

Yeah, everything's changing so fast and I think it's helpful to have some guideposts to look to, to say, okay, here here's what's working, here's what we're focusing on right now, and here's kind of how to move forward for the next month. Right?

Yeah, right. Because who knows what may bring? But we'll start with April. So let's start with news. What's happening in the news? What's changed? I know there have been a lot of headlines lately.

Yeah. So there are a few updates that I'll focus on. The first one is around Copyright Law update. So the Federal Copyright Office posted an official update around AI content, saying if work is generated specifically and exclusively by a machine, then that work lacks any human authorship and therefore it can't be copyrighted. And so the biggest thing there is just knowing that when AI technology is actually generating content for you, developing content, it's not yours, it's not necessarily proprietary. And so just making sure that what you're you're taking kind of what you're generating from AI tools and adding your own kind of spin and detail on it. Right. And so just word to the wise and a caution. And if you have somebody else creating content for you, a freelancer, an agency need to make sure that they are actually creating that content. They're the ones putting it together. So that's something that is definitely paramount as people move forward.

Okay, so you're saying if it's completely created by an AI tool, then it's not yours. So you could put it up on your website, but what, a competitor could also just copy paste? Is that like, they could what it.

Means to not yes, they could. Somebody can just lift it and use it. Now, it is a little bit mushy and kind of a moot point because we're also saying news is showing right now while there are AI detectors that they're not very reliable going to know. Okay, so how are we going to prove that you pulled this from a content generator?

Got it. If you changed it 10% or 15% did, it becomes yours. Okay, so I feel some mush on this.

There's some mush there, but it's just we're leading in the direction of we're not just going to spit out a bunch of content, put our names on it, and publish it. Right? Got it. And that goes the same for, like, you should feel good about publications that are creating content, knowing they've got that in their reputable. Publications are not just generating content straight from AI tools.

Yeah.

Okay. Thank you. All right, what else you got?

Okay, next one. There was an open letter published. Elon Musk was a signer of it, and that got a bunch of eyes on it. An open letter calling for developers to pause these giant AI experiments. This kind of AI summer. Have a summer break. And everybody the question that came up around this is why are we pausing? Why would we pause technology generation? And the thought here is that as a society, we need to come around and put some practices in place of how we use this so it doesn't get out of control. I bring this up just because it's recent news. Will there be a pause on AI developments? No, people are going to keep working on things, but it does just again, same as the copyright law, there's an unsettling feeling in people that we don't have a lot of parameters around this and so know that that guidance is going to start coming and the kind of regulatory stuff will start coming. It's not here yet, but I think all these things just kind of nod to a need for some kind of framework and structure to make sure we're creating systems that are safe and reliable, I think, for people.

Got it. Okay. Anything?

I have more headlines. So data privacy is an issue and this comes up. This is kind of tied into that open letter too, I think, helping people understand what types of information they're giving away and what the privacy laws say about it. So our privacy laws don't necessarily specifically address AI generated content, AI generated designs right now. But what happens is when you input something, when you prompt Chat GBT or you prompt one of these tools for something, in order to do that, you have already agreed that whatever you're prompting with that, you're handing that over. And so there's not a lot of regulation now around, well, what can that company do if I'm sitting within a large company or say I'm sitting within a government organization, and, oh, now I'm going to play with this tool, and now that AI company now owns whatever I inputted. So it's just kind of trying to understand and put some parameters around that. It's an interesting piece.

Well, I was just thinking of the scenario of new product development, right. And we have this new product coming out. It's our IP. So, yeah, you could unwittingly be handing over all kinds of information.

Yeah, you could be inputting how to address a merger and acquisition or trying to gain information on a company. And now where's the line? Right? How freely does that AI company own those queries? Do they want to do something with it? So that's kind of this need for some both regulation and greater awareness of where the lines are. And because of the lack of regulation right now, it was even just a few days ago, italy said they banned use of Chat GBT within the country. And there's other countries that have done that, but they're the first kind of Western country to do that. And so that's coming. And they're trying to put AI tools up against GDPR law and figure out what that looks like as the EU has led GDPR. And so there are more to come on that more to come on data privacy. The last thing I wanted to touch on, which is kind of news, but kind of just a point floating around there, is from what we're seeing at TREW pricing on AI tools is all over the place, but it's coming down. Tools that we evaluated four or five months ago and were priced at tens of thousands of dollars are now in the four figure range for the year.

And so not to say that that's the same for all the tools, but know that these companies are trying to figure out what their revenue streams look like. And so if there's a tool that you like that you've started playing with and you're interested in, I would encourage people to talk to their sales department because it seems like there's a lot of flexibility right now, and a lot of companies even pull pricing from their websites. And so it just says, contact us. Do you want to keep find out.

What the price is this week?

Yeah.

Negotiate.

Yeah.

Okay.

Good.

Well, thank you for the news roundup. If I'm a marketer listening today, what should I be considering bringing in AI for? What are some of the recommended use cases that you think people should try out right now, today?

Yeah, so AI is still such a really great brainstorm tool. So you're sitting there, you've got a blank page or you have a single topic, but you can really use that to kind of beat that writer's block and get you started, get you from blank page to some ideas, different angles on a certain topic, and you can also use it to do some preliminary research. Now, while I think it's a good starting point for research, basically if you're looking for kind of a definition, if it's something you would go to Wikipedia for, I think going to kind of chat GPT or something similar, is that's helpful? If you're looking for kind of a single answer, I still would say if you're doing research on a topic and you want varied opinions about something and you want to kind of get the lay of the land about a certain topic, especially for us when we're working with really technical topics. To research a technical topic often means you've got to look at all different sources. And so if you're doing that kind of intensive type of research, stick to your trade publications and your company websites and things like that and then form your own opinion before you just ask somebody to summarize the information.

Right. But as far as the tools go, they're good outline generators and then the transcription kind of the audio and video transcription is pretty good. Language translation is still pretty good also. And so those are things if you have a company operating in multiple countries and you need some language translation, you can use that. And then the rewriting and repurposing and summarizing tools are getting better. They're better than when we last talked. And so there are some tools available. You think, I wrote a blog post and now I want to feature it in a newsletter. And so I need to kind of summarize it and make it compelling for a couple of sentences that are going to go in the newsletter to promote it. That's potentially a tool you can use. Again, you're still going to need to weak it. But somebody said, I went to the Marketing AI Institute's virtual Summit for Writers last week and Mayhabib from the company writer talked and one thing she said was having these AI tools is like having an army of interns at your disposal. Yeah. And you're not going to take something that an intern wrote and publish it.

Right. It's going to need your professional expertise. That's why they're an intern and not you haven't hired them yet. So it's kind of that army of interns there.

Okay.

Another use case. And I know it's not necessarily what you focus on at TREW, but just a leap forward in reporting for marketers. So the ability in HubSpot of course, those of you using HubSpot probably know that they've added this great AI functionality, where instead of going through and filtering and creating your own report, where you have to do all these, drop downs and think about parameters. You just kind of ideate and say, I want to report on the first content instance that led to a sale last month for North America, and I want to see that in a chart or a table. And so just ideate that report and boom, there it is. And of course, it's not always perfect. You may need to dig in and make sure that it's pulling the right stuff and whatever, but a big efficiency time saver for marketers that I think is another fun use case. Not so much generative AI, but I'm excited about that.

Yeah, I love that.

Yeah.

So, Morgan, I know you've been playing with some tools lately, and it seems like it's funny, but it seems like everybody's talking about Chat Sheept all the time. Right. But really we're focused on tools that are more dialed in for marketers. And so what have you been playing with lately? What's caught your eye?

Yeah, so writer W-R-I-T-E-R continues to stay on our radar, but they've added some cool new functionality. So for a company that has insane repositories of content, often public companies with huge marketing teams, you can create a piece of content and then add your domain and it will crawl your domain and the content that you just wrote. And it will recommend links like what words should you hyperlink and where should you send those? Which is, I think, really cool functionality. Again, if you have been at that company for years and years and were involved in creating the content, you probably know. But if you're not, man, think about being new in a job at a company and writing a piece of content and then having a tool that's going to give you kind of that domain knowledge of everything that's been created before.

I think wonderful.

Yeah.

Here's where to cross link. And it's so easy to miss some of those, especially like when we do our research report every year, we would have to okay, now let's find all the instances we looked at last year's research. We want to update that to this year's. What a time saver. Great.

Yeah. And then there's a few other ones. I think the biggest advancements right now in the last month have been around SEO. So this is a little bit out of my wheelhouse, but what's happening is SEO and content development are converging. They're becoming kind of an a more integrative process than they ever have been before. And so there's a couple of organ companies demand well, is one, they create kind of SEO friendly outlines so you can put in your topics that you're creating content about the keywords you're going after. And it'll give you kind of what you need to cover in an article that you're going to publish on your site. And you can also input something you've written already and have them kind of judge it against the keywords you're going after. So you could use that to optimize content that you've already created and make that better. There's a company called I don't know if it's Gloss AI or it's Glossai, but it's Glossai Co. And this is a video repurposing tool. They have almost all these tools you can go play with for free. So I would encourage people, if they're interested, like, go hop on and do something class AI.

You can upload a video, and in the free version, you can upload maybe like a one or two minute video. It'll clip it for you and in different kind of sets and give you prescribed, like, LinkedIn posts for that video. So it'll write the text, put the hashtag in, it'll highlight in purple, things that you need to specify. Like, it'll say you upload a video that has multiple speakers, and it can do that. But again, you think about where we were four years ago when if we put together a webinar after the webinar, if we want to clip it, we're going into some Adobe tool, we're cutting up that video, we're figuring out the transcription, all this stuff, and now this will get you 80% of the way there. You customize that writing and you're ready to publish. And so, just a great time saver and to allow you to repurpose things that you've already created. Right. I would also say we've got clients who have tons and tons of webinars, and we go back and promote those, but what an easy way to go promote those webinars. So that's one, another one I really liked was called Summary.

So this is interesting, and I don't know if this will catch on. Summary, you can put in a long form piece of content and it'll summarize it, but it also has this API that you can integrate. Again, I'm a little bit out of my wheelhouse here, but this API that you can integrate into your site, and when you hover on a link, it will basically pop up a summary of where it's going to send you. And so you can see if that's like, is that where I want to go before I click? Which I think is interesting as some of the Google algorithm changes happen and we're looking for engagement on a page and things like that, but you don't have to click away. I don't know how it works on mobile. It feels like there's a lot of things that work out. It just looked cool when I saw it, so I liked that. And then lastly, I'll say, canva we talk about canva a lot? It's a really low barrier design tool. Low barrier to entry, like your intern can use it, your mom can use it. And there's a lot of templates in there, but they added a ton of functionality.

Things like create animation Tool, where you can drag your mouse across, drag your cursor across a path, and that will become the animation path. Oh my gosh. How easy is that? Wonderful. You don't have to click a bunch of little lines and all this stuff. So great. They've also got magic, right, which will generate content for you. And that was available only in their Canva Docs tool, and now it's available through all their design tools. And then they've got if you're a Canva Pro user, they've got a brand hub as well. And one thing I thought was cool here is it'll check your designs for Visual Inconsistencies. So again, Canva has always been great about you can upload your brand toolkit in there, so you've got your colors and things like that. But again, now say you've got a new person, you've got an intern in there, they're messing around and it's not the right blue, but nobody notices until it's on a big trade show booth thing. But it'll just kind of automatically check for those visual Inconsistencies, which I think is so cool. That is helpful.

Yes.

I feel like we're getting some of the things that for years we've been like, why can't this just be easier? And AI is making that easier for.

Us and a little bit smarter. And I know we've been saying, I don't want something to write for me, but I want someone to take my long form video, my long form piece of content, and split it up and parcel it and help me be smarter with SEO and do all those things. And it sounds like that's rapidly approaching and becoming better in multiple different tools.

And I think that functionality is even better. Like the clipping and the summarizing and that kind of stuff is a far better win than just this kind of summarized generate content generation.

Yeah.

So I know people ask all the time, okay, yeah, there's all these tools, just tell me what to use. So I'll be that annoying person to say, okay, if I want to do that, Morgan, I want something that's just going to take pieces of my content that I wrote and parcel it out like that.

What's the best tool?

Do you have an opinion?

I think that right now, writer is probably a good tool for that. I think. Let's take a white paper and turn it into four blog posts. What's interesting is Chat GPT could do that for you. And I'll talk in a minute about prompting and how to prompt. Well. However, I wouldn't at this point hand over all of that information to Chat GPT right. If I wrote this whole white paper, I don't yet want to enter that whole thing into an AI tool. And now they also own that content, right? And so I'm leery to at this point kind of recommend fully that we do that. I think we take pieces of what we've written and we can put them into a program like Writer and we can have it kind of iterate in another direction that maybe we wouldn't have gone or it'll provide some kind of SEO friendly general information, but I'm hesitant to really hand over an entire piece of content.

Yeah. Okay. Fair.

And what a great reason why, before we move on to prompting and other best practices, I think of Jasper and Writer as kind of like the same tool or very similar in functionality. What are some key differences?

Writer works more like the writing process. If you are a writer at your core Writer, you are going to like the Writer tool. So Writer, in the same way that you don't go from topic to entire blog post, you go from topic to research to outline to a draft to an iterated draft. Writer works like that. It's built to work like that. Jasper is a little bit more SEO forward if you're coming from I'm thinking of Aaron on our team. If Aaron is trying to put together web content to double down on a couple of topics, jasper is going to do that well for you.

Okay, interesting.

Yeah.

And then who knows where both of these tools and others, of course, will go. So that's why we have to talk about this all the time.

Right. And I think we'll see functionality start to converge. A few of these will start to win out for sure. Definitely. Writer is staying on top of what they're doing. Canvas staying on top of what they're doing. Jasper is staying on top of what they're doing. There's a lot of these other little ones that I think I could only imagine they're startups and they're sitting there going, this could go ten different directions and we're open to any of them, right?

Yeah.

But there's not like a clear business path probably for all of them.

Okay, well, let's round out our discussion with just some best practices for how to use generative AI specifically for those writing technical content.

Yeah. So what's happening is what you generate is only going to be as good as the prompt that you put in. And so, again, we don't normally go from a topic to an entire blog post. A tool is not going to magically do that for you and write something well, but the difference between prompting with just a question, give me a list of ten best practices versus a detailed prompt. And by detailed, I mean if you sent me out to write an article on something, I would have kind of all these parameters. And we start with our basic journalism questions of who, what, where, when, why, how. But if you start to answer those and put those in the prompt, you're going to get a lot better results. So things like chat, GPT write as if you are a technical expert in blah blah, blah topic and you're talking to me, a technician operating this specific machinery. What type of questions might you need to answer or something like that. But the more specific you get in that prompt, the more specific of an output you'll have, the better kind of things you have to work with if you're looking for actually some generation and not just like a brainstorming tool.

And so I think this is something that we'll talk about more as time goes on from a TREW perspective and we can do some kind of in depth trainings on. But the prompting is definitely key. It requires you to think like a content developer without these tools and then just use the tools to aid in your process so you're not having to think a different way. But it's not a silver bullet. You're having to be really specific about what you're asking for and then you're going through drafting and iterations just like you normally would.

Yeah, I know you were showing me two different examples before we hopped on this call. So one was, what are some things that design engineers are worried about in this specific industry or within this specific application within this industry? And that's one scenario, but then the other one is going even one level deeper than that and being more and more specific and then the output just got better and just more tailored to that audience.

Makes sense.

Yeah. And these prompting tools, they like references, so you think that it's supposed to be providing you the information, but when you prompt, you can also say, here's an example of the tone and format that I'm looking for, and reference something that you've already written. And so you're kind of training it to learn, training it to create on things that you've already created. Again, if you have nothing, if you have no content and no ideas, it's just not a silver bullet for you. It's not going to be helpful. But if you've got a process in place and you can use these tools to help you along the way, they can be really valuable.

Cool.

So, Morgan, I'm starting to see little notes pop up on websites just in my consumer life, if you will. Like, I was looking for reviews on a product and it said this blog post was created by a human, not an AI robot. And it just looks like this one little phrase. And then we've seen other I think maybe you sent the example of Wired Magazine having a whole policy around this. So tell me, just what are some best practices around that?

Yes, that's great. So a couple of tools you can use. So the Marketing AI Institute has actually an AI use manifesto on their site that you can start with. It's published under Creative Commons. You can list that and start to edit it from there. But the recommendation right now is for content published on your site. Put something on your site that says, here's how we think about AI, here's how we use it, here's how we don't use it. And you can trust that for all of our content. And then for every little piece that you publish, you don't need to put a note on it. You can put something in your footer to reference your policy on AI tools, but you don't have to kind of note it every time I think we see it noted when people are trying to say, hey, we're still humans over here. This is human thought that's getting put into what we're doing. And so you might see it noted more when people are saying, we actually wrote this. We didn't generate this. And I still think as time goes on, there is going to be a continued increase for personally generated content.

I think things like this unscripted discussion that we're having, that's what you okay, fine, fine. I can go to Chat GPT and have it pull a summary of an AI update for me, but I want to hear about people actually using the tools. I want to hear about what we're talking about. I want to hear questions people are asking. And I think we're going to just start to really crave these personal discussions. And so that said, as far as your website goes, have something that says this is how you use it, this is how you don't that's referenceable and then move on.

All right, good.

Well, we're going to continue to stay on top of this, especially you, Morgan. And I think you raised your hand for a webinar later in this quarter. So to that, if you guys listening are on our TREW Marketing email list, we'll make sure you hear about that once we get a date nailed down. And any parting advice for marketers listening that are ready to get started with generative AI?

I think just pick a couple of tools. Don't get yourself a list of 100 tools and feel overwhelmed by it, but pick a couple of tools that you're interested in and start playing around with them. I would also love we will do a webinar, and I think that the Q and A is so important in the webinar. And so if people are listening to this and over time, if you have questions, feel free to shoot them to me and I can start to incorporate that into the webinar so@TREWmarketing.com you can email me and we'll kind of get the research and answers together that you need.

Great.

Oh, my gosh. Thank you so much, Morgan. This is very interesting. I appreciate 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.

I Would love your reviews on this podcast cast. So please, when you get a chance, subscribe and leave me a review on your favorite podcast subscription platform. Thanks and have a great day.