Sept. 1, 2025

Transforming Media Strategy: How AI Tools Empower Communications Professionals

Transforming Media Strategy: How AI Tools Empower Communications Professionals
The player is loading ...
Transforming Media Strategy: How AI Tools Empower Communications Professionals

Feeling overwhelmed by the AI revolution sweeping through public relations? You're not alone. In this illuminating conversation, Pete Pachal, founder of Media Copilot AI, reveals to host Peter Woolfolk how PR professionals can transform their relationship with artificial intelligence from frustration to fruitful collaboration.

The journey begins with a crucial mindset shift. Rather than viewing AI as a magical "Santa Claus" that instantly delivers perfect content, successful PR teams approach AI as a collaborative partner in an iterative process. This fundamental change in perspective opens up possibilities far beyond basic content generation, extending into campaign automation, targeted media monitoring, and even interactive interview training.
Pachal walks us through specific applications that are revolutionizing PR workflows. Content creation remains the most requested capability, but the real value emerges when professionals provide sufficient context and choose the right AI tool for each specific task. When implemented correctly, AI advances projects from halfway complete to 80-90% finished—though human oversight remains essential due to AI's potential for errors.

For teams struggling with expensive, feature-bloated software platforms, Pachal offers a compelling alternative: creating narrowly-focused AI tools tailored to specific clients or campaigns. This approach delivers precisely what PR professionals need without paying for unnecessary features. Perhaps most fascinating is the evolution of media training, where interactive AI now simulates interviews, challenges spokespersons with difficult questions, and provides constructive feedback—all customizable to mimic specific journalists or media personalities.

Media Copilot offers to download their free "100 Top AI Prompts For PR Pros" at Mediacopilop.ai"

I would very like to get a review from you. Please send a note to me. Thanks, Peter! like to much appreciate a review from you!! Thank you!

Feeling overwhelmed by the AI revolution sweeping through public relations? You're not alone. In this illuminating conversation, Pete Pachal, founder of Media Copilot AI, reveals to host Peter Woolfolk how PR professionals can transform their relationship with artificial intelligence from frustration to fruitful collaboration.

The journey begins with a crucial mindset shift. Rather than viewing AI as a magical "Santa Claus" that instantly delivers perfect content, successful PR teams approach AI as a collaborative partner in an iterative process. This fundamental change in perspective opens up possibilities far beyond basic content generation, extending into campaign automation, targeted media monitoring, and even interactive interview training.

Pachal walks us through specific applications that are revolutionizing PR workflows. Content creation remains the most requested capability, but the real value emerges when professionals provide sufficient context and choose the right AI tool for each specific task. When implemented correctly, AI advances projects from halfway complete to 80-90% finished—though human oversight remains essential due to AI's potential for errors.

For teams struggling with expensive, feature-bloated software platforms, Pachal offers a compelling alternative: creating narrowly-focused AI tools tailored to specific clients or campaigns. This approach delivers precisely what PR professionals need without paying for unnecessary features. Perhaps most fascinating is the evolution of media training, where interactive AI now simulates interviews, challenges spokespersons with difficult questions, and provides constructive feedback—all customizable to mimic specific journalists or media personalities.

Media Copilot offers various entry points for PR professionals looking to develop their AI skills, from accessible monthly sessions to comprehensive six-week courses with personalized coaching. For immediate value, download their free "100 Top AI Prompts for PR Pros" guide, which provides specific language patterns and prompting techniques that dramatically improve outputs.

Don't let fear of falling behind paralyze you—start with small steps toward AI integration and discover how these powerful tools can enhance your creativity and efficiency while maintaining the human touch that defines great public relations work.

Information on NEW podcast website.

Support the show

Newsletter link:

https://www.publicrelationsreviewpodcast.com

Transcript

WEBVTT

00:00:04.847 --> 00:00:05.548
Welcome.

00:00:05.548 --> 00:00:19.635
This is the Public Relations Review podcast, a program to discuss the many facets of public relations with seasoned professionals, educators, authors and others.

00:00:19.635 --> 00:00:23.288
Now here is your host, peter Woolfolk.

00:00:25.060 --> 00:00:31.153
Welcome to the Public Relations Review podcast and to our listeners all across America and around the world.

00:00:31.153 --> 00:00:44.875
Now, apple has ranked this podcast among the top 1% of podcasts worldwide and recently Feedspot listed this podcast as number 13 on its top 70 best public relations podcasts in the United States.

00:00:44.875 --> 00:00:52.500
So thank you to all of our guests and listeners for your continued support and if you enjoy the podcast, please leave a review.

00:00:52.500 --> 00:01:00.564
Question to my audience is your organization empowering your staff with public relations and media AI strategy?

00:01:00.564 --> 00:01:05.194
How are you coping with the fast pace of change and new platforms?

00:01:05.274 --> 00:01:05.334
in.

00:01:05.474 --> 00:01:05.855
AI.

00:01:05.855 --> 00:01:14.763
As AI reshapes the landscape of media and public relations, professionals need practical skills to effectively leverage this technology.

00:01:14.763 --> 00:01:29.147
Media Copilot's program gives participants hands-on experience with cutting-edge multiple AI tools to elevate their communication strategies, enhance real-time monitoring and how to automate critical tasks.

00:01:29.147 --> 00:01:35.069
Their program is designed for PR professionals, media strategists and content creators.

00:01:35.069 --> 00:01:43.433
It provides actionable insights and practical applications of AI to transform how you can successfully engage with your audience.

00:01:43.433 --> 00:01:47.084
Transform how you can successfully engage with your audience.

00:01:47.084 --> 00:01:48.888
So joining me today, from the New York City area, is Pete Pachal.

00:01:48.888 --> 00:01:51.055
He is the founder of MediaCopilot.

00:01:51.055 --> 00:02:00.503
He has been covering technology for more than two decades and has been following artificial intelligence since before Gmail was trying to complete your sentences.

00:02:00.503 --> 00:02:09.352
He was chief of staff at CoinDesk, where he led the publications AI committee and wrote the company's guidelines for the use of generative AI.

00:02:09.352 --> 00:02:15.087
He's also held senior editorial positions at Red Ventures, mashable and NBCUniversal.

00:02:15.087 --> 00:02:17.693
So, Pete, welcome to the podcast.

00:02:19.501 --> 00:02:21.108
Thanks, peter, it's a pleasure to be here.

00:02:22.061 --> 00:02:31.133
Now, needless to say, AI continues to grow and change, so let's begin one by talking about what led you to establish the Media Co-Pilot AI.

00:02:33.480 --> 00:02:47.860
So I'd been fresh off my job at Coindesk, where I was leading strategy and, as you mentioned, I was part of their AI committee, actually led it and wrote their guidelines.

00:02:47.860 --> 00:02:52.471
So that was one of the most rewarding things I actually did for Coindesk.

00:02:52.471 --> 00:03:00.230
And when it was time to figure out my next move, I knew I wanted to do something with AI and I knew media.

00:03:00.230 --> 00:03:03.329
That's the thing I've been working in for my whole career.

00:03:03.329 --> 00:03:19.133
So I decided to combine those two things and essentially give myself a job to figure out how AI was going to affect the media, how it was going to change the media, business and journalism as a result of that.

00:03:19.133 --> 00:03:25.629
And so I didn't quite know exactly what form that was going to take, but I knew doing a newsletter.

00:03:25.629 --> 00:03:35.054
The barrier to entry was low, so I launched that, I launched the Substack and I was, you know, determined to teach myself this stuff and pass on what I learned.

00:03:36.039 --> 00:03:45.694
And really shortly after that, some friends and colleagues started reaching out to me and say, hey, can you come in to my team and maybe explain?

00:03:45.694 --> 00:03:50.561
You know this AI stuff or you know sort of words to that effect, and I thought, yeah, sure I could do that.

00:03:50.561 --> 00:03:58.495
So I went to a couple of the places, usually in New York, and they went great.

00:03:58.495 --> 00:03:59.925
There were great sessions.

00:03:59.925 --> 00:04:03.240
I would talk for hours because people would want me to talk for hours.

00:04:03.240 --> 00:04:10.390
They asked me a bunch of questions about AI and apps and I was very happy to be able to fill in some gaps for them.

00:04:11.199 --> 00:04:21.490
And then the more that I did those, the more I thought this is a really good way to communicate this knowledge in the media.

00:04:21.610 --> 00:04:38.874
And, having dealt with many public relations folks over the years, I knew I had very specific advice for them because I knew I had a really good sense of the workflows, the ethics, all the things that really are of concern to those professions.

00:04:39.379 --> 00:04:43.408
So I launched classes formally in early 2024.

00:04:43.408 --> 00:04:53.213
Those got pretty popular really quickly and a lot of PR firms actually ended up reaching out to me for something a little more custom.

00:04:53.213 --> 00:05:07.824
You know they liked the classes and they liked sort of the use cases I was going over content creation and various other things, but they really wanted something more tailored to their work, other things, but they really wanted something more tailored to their work.

00:05:07.824 --> 00:05:49.810
So you know, that began kind of my sort of custom client business and that's grown and I've partnered with a few other people over the following months and now we have that as a very strong sort of pillar of the business, which is to say like it's essentially kind of a hybrid between training and consulting with PR and communications firms to basically bridge them into this new era where AI is powering everything we do and making sure they can, you know, compete and basically get more out of the resources that they have.

00:05:51.360 --> 00:05:54.591
Let's talk a little bit, maybe bring a bit more clarity.

00:05:54.591 --> 00:06:07.182
So you know, when you said PR firms needed you to help them with what they needed from AI, what did they identify as, let's say, the top five things that they needed, or thought that they needed, from ai?

00:06:09.629 --> 00:06:10.471
oh, good question.

00:06:10.471 --> 00:06:11.413
Yeah, so there's.

00:06:11.413 --> 00:06:12.617
There's multiple use cases.

00:06:12.617 --> 00:06:13.759
I'll get into those in a second.

00:06:13.759 --> 00:06:27.406
Um, the first thing, though, I would say that, um, pretty much any firm needs what, regardless of the industry is, is a shifted mindset around AI, because I think everyone views AI at this point.

00:06:27.406 --> 00:06:51.144
Everyone kind of knows generally what it can do, at least out of the box, and everyone, in my view, starts off with a view of AI as kind of like almost like Santa Claus, right Like you ask him for something, give me a blog post, give me some social copy, give me an image, and it does it for you and it's kind of magical and it's great.

00:06:51.144 --> 00:06:54.882
And then the magic wears off pretty quickly because you realize, oh, this is sort of riddled with errors.

00:06:54.901 --> 00:06:56.990
It's not quite what I wanted oh, this image looks silly.

00:06:59.021 --> 00:07:29.966
And when you shift your mindset away from it just giving you a thing, and you are collaborating on a thing, you know that you adjust your expectations and your mindset so that you're not expecting perfection on the first prompt and that the span of things that AI can help you with goes beyond just giving you a thing, an output, so to speak then that's a real unlock.

00:07:29.966 --> 00:07:34.608
That shift in mindset is key to really getting the most out of AI.

00:07:34.608 --> 00:07:42.286
When you see AI as a collaborator on the many projects that you do, so that's number one.

00:07:42.286 --> 00:07:49.634
And then when you get into PR-specific use cases, there's a number that are predictable.

00:07:50.019 --> 00:08:16.399
So, right off the bat, most people want to talk about content creation and that spans everything from press releases to company marketing material, to you know social copy, you know social copy, and so one is sort of figuring out the prompting around that to make sure you have the right, you're asking AI for the right things and you're also asking the right AI for the right things, because there's obviously a lot of them.

00:08:16.982 --> 00:08:33.432
And then part of that is also providing the context so that it has the right things, the specific data from you to do what it can do to the best of its ability, so that it gets you like, instead of halfway to being done, it gets you more like 80% to 90% of the way being done.

00:08:33.432 --> 00:08:40.019
And you know, I always emphasize it's never 100%, because these things are not perfect systems.

00:08:40.019 --> 00:08:42.870
They hallucinate or get things wrong.

00:08:42.870 --> 00:08:49.826
Human-in-the-loop is pretty crucial with any AI workflow, so content creation is a big one.

00:08:49.826 --> 00:08:56.452
I will say people are very interested in campaign-related things.

00:08:56.452 --> 00:09:21.215
So whether it is automating some of that content, so that, say, you're sending out, whether it's pitches or social copy or whatever, being able to kind of like, have that automatically go somewhere and then be stored in a database in terms of like, whether it was successful or not, you know that's.

00:09:21.215 --> 00:09:23.721
That's a matter of pretty simple stuff these days.

00:09:23.721 --> 00:09:27.211
You used to have to go to engineering teams to like hook systems together.

00:09:27.360 --> 00:09:30.450
Now there's so many sort of no-code automation platforms.

00:09:30.450 --> 00:09:43.649
That's good on loss in terms of like what you can do once you have like your output, and then there's sort of the reverse of that, which is like the media monitoring.

00:09:43.649 --> 00:09:45.287
So everyone's very interested in media monitoring.

00:09:45.419 --> 00:10:03.230
They actually certainly PR firms everywhere spend a lot of money on big platforms that will do this for them to some extent and I'm not saying out-of-the-box AI can replace that kind of extensive platform, but elements of it can be.

00:10:03.230 --> 00:10:17.349
You know, you can actually essentially create your own micro version of one of these platforms in a limited way and get you something that will maybe be very narrowly focused on a particular client or campaign.

00:10:17.349 --> 00:10:24.311
That doesn't necessarily do 800 different things, it does two or three different things.

00:10:24.331 --> 00:10:27.158
And this is kind of one of these interesting empowering things about AI.

00:10:27.158 --> 00:10:42.195
If you have an aptitude to sort of building your own tools around your own workflows, you can really almost get a thin client, so to speak, of the software that you use every day.

00:10:42.195 --> 00:10:47.231
That might seem bloated and you think you're only using like 10% of the features.

00:10:47.740 --> 00:10:53.129
If you are only using 10% of the features, there's a good chance you might be able to replicate a good chunk of that on your own.

00:10:53.129 --> 00:11:01.844
Then there's some really interesting use cases that get into it's media training and you can.

00:11:01.844 --> 00:11:02.284
This is.

00:11:02.325 --> 00:11:27.159
I really like this one in particular because it gets at just how interactive the use of AI can be, so once you start using tools for voice and video, you can start to do some pretty impressive things with regard to to having, like, an interview done right.

00:11:27.159 --> 00:11:36.980
So you could actually have an AI interview you as an expert and then challenge you or give you a softball or whatever you can.

00:11:36.980 --> 00:11:45.653
Actually, it's all up to you what the style of the interview is, evaluate the interview and give you pointers on how to handle a real one.

00:11:46.200 --> 00:11:55.773
And you can even craft these to be replicas, or at least close to replicas of real interviewers, real journalists.

00:11:55.773 --> 00:12:05.051
You could even create your own simulator that behaves like Wolf Blitzer, for example, based on some content that's out there.

00:12:05.051 --> 00:12:18.650
So that's a good one, because that gets into sort of the interactivity of AI, how it's not just here's a prompt and give me a thing, you know, it's much more involved.

00:12:18.650 --> 00:12:30.268
And then there's like all kinds of assistance on very narrow fields and there's different tools that are better suited for one or the other.

00:12:30.268 --> 00:12:50.864
So whether you are doing sort of a campaign assistant, that might be done better by, say, like a Notebook LM, which that's Google's tool that is very grounded, so it's only getting information from the files you give it, or you might want something that's more open, something that can access the internet.

00:12:50.884 --> 00:12:55.044
So maybe you have something that's more along the lines of a compliance advisor, right.

00:12:55.044 --> 00:13:08.267
So yeah, you can ask general chat GPT for compliance advice, but it's going to be sort of hit and miss and often when you're asking AI for help, you don't need the entire internet.

00:13:08.267 --> 00:13:09.190
In fact, you don't want it.

00:13:09.190 --> 00:13:40.870
You want it to be very, very focused on a particular area and you want it to primarily rely on something very specific, and that could be, you know, guidelines, areas of law, areas of specific reports or industry reports, and there are any number of ways to combine both files that you have with databases on the internet and limit it to that, you know, so that you're not getting all the junk mixed in there, like some rando on Reddit talked about some compliance issue.

00:13:40.870 --> 00:13:43.686
No, you don't want that, you want the official stuff.

00:13:43.686 --> 00:13:53.366
There are ways to sort of do that and it's not that hard, once you sort of sit down and know the right tools.

00:13:54.081 --> 00:14:05.586
You know, let me jump in here real quick on this because one of the things that I know when we talked earlier that I found on your website that I think folks will love is that it's free.

00:14:05.586 --> 00:14:09.726
It's the Euro 100 Top AI Prompts for PR.

00:14:09.726 --> 00:14:20.153
Pros and listeners can get a free copy of it from themediacopilotai and it brings up the things that you really talked about.

00:14:20.153 --> 00:14:21.140
You know.

00:14:21.140 --> 00:14:25.591
Here's some of the things, some of the rules you know to if you begin using AI.

00:14:25.591 --> 00:14:40.613
You know setting word limits and set the tone to generate the best results, those kinds of things, and you know some of the right words, verbs to use or, if you want to refine your prompts and so forth, like that, research prompts.

00:14:40.613 --> 00:14:46.173
This information is included from Media AI's website.

00:14:46.173 --> 00:14:58.351
I've used it and, my heavens, it really did help me move faster, very quickly, because it's sort of fine-tuned and maybe more of a laser focus than a shotgun approach to it.

00:14:58.351 --> 00:15:04.510
So I just wanted to bring that to our listeners' attention of the top 100 AI prompts from MediaCopilotai.

00:15:05.740 --> 00:15:12.601
So, just wanted to make sure we got that in front of people yeah, yeah, thanks for cleaning that out.

00:15:12.601 --> 00:15:14.808
Um, yeah, it's a great list.

00:15:14.808 --> 00:15:19.582
I I think it will unlock a lot of possibilities to you.

00:15:19.582 --> 00:15:32.153
If you haven't prompted ai beyond um, sort of that sort of output focused um out of the box experience, um, it will get you thinking about context.

00:15:32.153 --> 00:15:34.057
It'll get you thinking of that sort of output-focused out-of-the-box experience.

00:15:34.057 --> 00:15:35.240
It will get you thinking about context.

00:15:35.240 --> 00:15:39.919
It will get you thinking about the different areas that it can tap into and ways it can advise you.

00:15:39.919 --> 00:15:43.028
Yeah, that's a great download.

00:15:43.028 --> 00:16:19.457
There's also some things you will learn in the classes I teach that are there's some specific prompts, but I think there are some more like broader rules about using AI, and some of them I've come up with fairly recently just from repeated use, and one of them is it's really interesting how, when you tell AI not to do something, it kind of wants to do it, you know, even more.

00:16:19.538 --> 00:16:30.649
It's one of these things scary about AI that they're almost like think like us, a little bit like if I told you right now, don't think of coffee, don't think of coffee, like what's the one thing you're gonna think, think of.

00:16:31.090 --> 00:16:32.836
And AI sort of behaves that way too.

00:16:32.836 --> 00:16:38.072
So I think most people find this that if you tell it not to do something, it almost tends to do it even more.

00:16:38.072 --> 00:16:50.530
So a tip I have about the use of AI broadly is it's really almost a tip for again, bored from human interaction be more positive than negative.

00:16:50.530 --> 00:17:12.268
Like emphasize the things you wanted to do, and it's not like never tell it what not to do, but put more emphasis on the opposite thing you wanted to do and repeat that more so that it is more focused on that, as opposed to be just focusing on the thing you told it not to do.

00:17:12.268 --> 00:17:16.726
Because if you just say this, don't do this, don't do this, don't do this, it's just going to want to do all that.

00:17:16.726 --> 00:17:27.307
So that's sort of an interesting unlock I've discovered on AI that sort of de-emphasize the negative and amplify the positive.

00:17:27.901 --> 00:17:49.279
You know, and I think as I look through some of your material, you probably describe it in here as hands-on mastery of AI tools, and I think that certainly helps people because you know, mastering AI tools certainly makes the output, if you will, more specific to what your needs are, rather than a broad generalization of what you might be looking for.

00:17:51.344 --> 00:18:18.752
Yeah, so that's really important in terms of the tools and for tooling, like we run the gamut, and so it's like there's very specific stuff and very general stuff and, depending on the audience, like for the public-facing classes that I put on that, you know they're sort of come one, come all and anyone can buy a spot and I, you know, will teach for either a day or sometimes for over six weeks.

00:18:19.180 --> 00:18:25.625
But I try to emphasize the more general tools because I don't necessarily want to lock people into a very specific platform from the get-go right.

00:18:25.625 --> 00:18:28.940
So, that said, there's a lot of broad tools for everything from ChatGPP to Cloud to Perplexity, people into a very specific platform from the get-go right.

00:18:28.940 --> 00:18:33.887
So, that said, there's a lot of broad tools, everything from ChatGPT to Cloud to Perplexity.

00:18:33.887 --> 00:18:35.164
These are all very useful things.

00:18:35.164 --> 00:18:38.087
But when I do classes I try to sort of emphasize those.

00:18:38.519 --> 00:18:47.605
But for the more consulting custom packages that we're doing for corporate costumes, we can get very granular.

00:18:47.605 --> 00:18:54.984
There are tools out there that are very targeted on very specific PR functions.

00:18:54.984 --> 00:19:07.094
There is a tool out there, for example, that will pitch in an automated or semi-automated way podcast.

00:19:07.094 --> 00:19:29.133
It's built specifically for podcasts so that if you want to essentially pitch your client to appear on podcasts in a certain area, it is extremely good at finding those and finding the contact information of the people to reach out to and then we'll essentially pitch you.

00:19:29.133 --> 00:19:34.030
And that's a crazy narrowly focused thing, right, if you think about it.

00:19:34.030 --> 00:19:36.566
What is the call for that?

00:19:38.303 --> 00:19:44.171
Well, you know, let me just jump in there right quick, because right now most of my guests are actually coming from pitches.

00:19:44.171 --> 00:19:48.140
I have to ask them, you know, how did you happen to find my podcast?

00:19:48.140 --> 00:19:56.811
Well, there are organizations that are set up to look for that deal specifically with podcast and public relations podcast, to be specific.

00:19:56.811 --> 00:20:03.584
So right now I'll read what's coming and if I like what it is I'm seeing, then I'll reach out to you and we'll go forward from there.

00:20:03.644 --> 00:20:05.991
But yeah, it's not.

00:20:05.991 --> 00:20:06.815
I don't think it's.

00:20:06.815 --> 00:20:21.092
I think it's a fine tool.

00:20:21.092 --> 00:20:23.253
You know what I mean?

00:20:23.253 --> 00:20:24.535
Like, I mean, it's just using.

00:20:24.615 --> 00:20:26.797
AI in this way that you might not.

00:20:26.797 --> 00:20:29.960
Other tools that will pitch journalists in a very specific area.

00:20:29.960 --> 00:20:32.343
There are other tools that will.

00:20:32.343 --> 00:20:45.017
The media training thing I mentioned will automate a lot of that and do it in a very sophisticated way so that you are interacting with video avatars or voice avatars.

00:20:45.017 --> 00:20:58.332
And there are others that specialize in like list curation, you know, like, and so we can talk about all of that and I do some recommendations through the business I do.

00:20:58.332 --> 00:20:59.623
So it's all.

00:20:59.623 --> 00:21:05.925
It's all tailored to whoever the client is Like what do you need, what are you most interested in?

00:21:05.925 --> 00:21:17.046
And we can sort of recommend a whole stack that usually is pretty affordable when you compare it to the big loaded software that I was kind of alluding to earlier mm-hmm.

00:21:17.626 --> 00:21:40.770
Well, you know, one of the things that you mentioned in here I think you basically touched on is that students will gain is that you know they're data-driven insights, where you can actually leverage AI to interpret real-time media trends and other things like that, so that they'll know how to maybe format the messages or what particular media to go to that kind of information.

00:21:40.770 --> 00:21:48.599
So the specificity is what I'm hearing is that there's a lot of specificity can be developed from this if you use it correctly.

00:21:50.154 --> 00:21:51.541
Okay, for sure.

00:21:51.541 --> 00:21:57.534
Yeah, the specificity is kind of the biggest unlock, right.

00:21:57.534 --> 00:22:26.961
So like I say a lot about at the outset of a lot of my trainings is that, out of the box, ai has more to offer a novice that wants something passable than a professional who wants something good, and that's frankly why people come to me because they're professionals and they want something good and they want to figure out the best way to do that and it's like, well, the way to upgrade your AI experience number one is like the AI should really know you.

00:22:27.463 --> 00:22:29.510
You know, yeah, of course you're not getting great results.

00:22:29.510 --> 00:22:35.873
You're just pinging some generic chat GPT bot to do something that doesn't have the right context.

00:22:35.873 --> 00:22:55.525
So you know, step one is sort of building some assistance, either in chat, GPT or some other platform that has the right context, and then, once you sort of have the right context and the goals, it's like oh actually, in addition to context, I wanted to do something, and probably beyond just giving me, you know, a piece of writing or an an output.

00:22:55.905 --> 00:23:12.385
It's like I wanted to call a tool or a function and it's like, ah, okay, then you can either hook that up next to wires with some tooling or buy a tool eventually that has already done that and figured out a lot of the prompting as well.

00:23:12.385 --> 00:23:20.603
So that's the big advantage of a lot of these software platforms and it's just that they figured out the best prompting to get consistent results.

00:23:21.324 --> 00:23:41.255
They've got some methods to connect wires, essentially between things, so that it can call function and bring them into the middle of the process and, in the case of what you might call an agent, interpret the output that they get from these other things and do something with that.

00:23:41.255 --> 00:23:41.880
So that's all an agent is.

00:23:41.880 --> 00:23:57.829
It's sort of combining a bunch of these tools and then giving it the right instructions on what to do when it gets things from those tools, and then the more you do that the more you can sort of have it just do more of the work on its own.

00:23:58.319 --> 00:24:10.691
And the big trick there is sort of figuring out where to insert your human judgment, which is absolutely still needed, you know, because when you start talking about looking up tools and agents, people always start to get a little worried about.

00:24:10.691 --> 00:24:16.920
Wait a minute, what am I here for way with?

00:24:16.920 --> 00:24:17.501
What am I here for?

00:24:17.583 --> 00:24:52.926
and so you know, there's a lot of existential dread and a lot of injuries, industries, pr and communications included, but the there's, there's absolutely a need at several stages in any workflow or for humans to be in the loop, and there's certainly stages where I call for you know, like, when you want that human authenticity, when the relationship is important, yes, AI can get you close to like, basically do a lot of things prior to that interaction, but you still got to do it as the human, because that's what we're doing here.

00:24:52.926 --> 00:24:54.487
That's who your customers are usually.

00:24:56.621 --> 00:25:07.685
Well, you know again, as I go through that 100 plus program that you have for free, because I'm listening to the things how you help people.

00:25:07.685 --> 00:25:30.182
You suggest certain verbs that they can use and you know how to refine their prompts by using certain words you know, such as edit or combine or synthesize or interpret for me, and even some research prompts you know, such as list the common challenges of, describe the target audience for, explain the article to a five-year-old and, with a link, all of those kinds of things I help.

00:25:30.182 --> 00:25:39.756
I can see how people can get better output, if you will, from AI, if the specificity is there that they're looking for.

00:25:41.662 --> 00:25:59.474
Well, especially for your own customization, like that one about you know, explain it to me like I'm five is a good example of understanding that you as the user are also an audience, right understanding that you as the user are also an audience, right.

00:25:59.994 --> 00:26:14.145
So the AI out of the box kind of doesn't know you from Adam and the more you can tell it about yourself, the bar is going to give you stuff that's going to be useful to you.

00:26:14.145 --> 00:26:31.828
So there's a very underutilized feature in ChatGPT, for example, called custom instruction, and I often leave it like I've been teaching it ever since I've started teaching AI because it's been there for a while and I've been tempted now and then over the years to kind of stop because I figured everybody knows about it.

00:26:31.828 --> 00:26:34.446
But I find that people don't.

00:26:34.446 --> 00:26:40.849
Usually People just kind of leave it untouched and I like, oh wow, no, this is a great way to just out of the box.

00:26:40.849 --> 00:27:00.424
So go from out of the box to something that is much more tailored for you and you can tell it all, anything you want to do in terms of like how you want it to respond, and so to think about you and what, how, like what your comfortable comfort level is with certain language and what sort of how you like explanations.

00:27:00.444 --> 00:27:02.692
You might like them longer or shorter depending on the context.

00:27:02.692 --> 00:27:14.769
But I also find it's a very useful place to kind of end run some of the annoying habits ChatGPT has, because it tends to be over eager.

00:27:14.769 --> 00:27:24.186
I think a lot of people when they ask chat GPT for something and you were expecting maybe three bullets, it gives you 20 or you know, it just keeps going and going on something.

00:27:24.819 --> 00:27:34.651
And so you kind of have to tell it, limit your responses to say you know 200 words, unless I ask for a long, something longer.

00:27:34.651 --> 00:27:36.987
That's a good custom instruction.

00:27:36.987 --> 00:27:47.421
Another one I like to use is that if you ever use, you know, if you've used ChatGPK for a while, you've noticed that you ask it for a thing you know.

00:27:47.421 --> 00:27:53.547
You're like give me a recipe for an arugula salad and it'll go like okay, I'm going to do that recipe.

00:27:53.606 --> 00:27:54.148
Here's a good one.

00:27:54.148 --> 00:27:55.303
And it says gives you the recipe.

00:27:55.303 --> 00:27:57.628
And it says give you the recipe, and it says okay, I just gave you a recipe, Did you like that?

00:28:03.940 --> 00:28:08.435
And those sort of I'm about to do something, I've just done the thing, kind of bookending comments start to get annoying for a while and you can actually tell it to stop doing that.

00:28:08.435 --> 00:28:16.013
You can essentially go you know, stop self-referencing, don't explain what you're doing, just give me the thing I asked for and then it'll stop.

00:28:16.013 --> 00:28:18.585
And it's a bit of a.

00:28:18.585 --> 00:28:30.849
I guess you might consider it an advanced move for ChessGPT, because I think people don't even notice those things until you're sort of using it all the time and you start to realize how annoying it is to have this stuff.

00:28:30.849 --> 00:28:37.855
But it's so responsive, I mean, you can tailor it in so many different ways.

00:28:38.419 --> 00:28:56.991
You can even tell it to be working like references to Narnia or something in every response, which is a nice trick I do when I'm showing off construction because it's very amusing to watch it sort of bring up things like Aslan and the White Witch or whatever.

00:28:56.991 --> 00:28:58.906
Never in many responses.

00:29:00.260 --> 00:29:11.067
You know, one of the things I think we might want to do before we get too far down the road is to let people know how they can get in touch with you, because you know the information that you are providing here right now.

00:29:11.067 --> 00:29:20.582
Obviously, a lot of folks have a need for it, so why don't you let them know how they can reach out to you and get involved in a program of folks have perhaps have a need for it, so why don't you let them know how they can reach out to you and get involved in a program that you have?

00:29:22.667 --> 00:29:23.750
yeah, absolutely, thank you.

00:29:23.750 --> 00:29:25.917
They can always email me.

00:29:25.917 --> 00:29:31.268
My email is Pete PEG at media co-pilot AI.

00:29:31.268 --> 00:29:32.369
It's all one word.

00:29:32.369 --> 00:29:34.374
That's the easiest way.

00:29:34.374 --> 00:29:36.837
I'm really accessible on LinkedIn.

00:29:36.837 --> 00:29:41.167
I'm there all the time and so you can reach me there.

00:29:41.167 --> 00:29:42.631
I'm on a few other social networks.

00:29:42.631 --> 00:29:47.108
I probably check them out a little less often, but you can reach out to me on any of those.

00:29:47.108 --> 00:29:50.785
If you want to just Google my name, it's got a weird spelling.

00:29:50.785 --> 00:29:58.444
My last name it's P-A-C-H-A-L, but you should be able to find me pretty easily and, like I said, the best way is email.

00:29:58.444 --> 00:30:00.309
Second best is probably LinkedIn.

00:30:00.971 --> 00:30:01.172
Mm-hmm.

00:30:01.172 --> 00:30:16.034
Well, look, Peter, let me say this you provided us with an awful lot of information on your program, Anything that you think we might that you may have overlooked in providing us about the program.

00:30:17.721 --> 00:30:33.885
Yeah, I just want to emphasize even though I'm excited about the new thing that I have with my partners about doing things for PR teams and corporate comms teams I teach and have classes for all levels, so I do a monthly class.

00:30:33.885 --> 00:30:38.661
It's live, it's only an hour.

00:30:38.661 --> 00:30:54.644
It's essentially an introduction to AI, but it's meant to give very practical advice and tips, like some of the stuff I've just talked about today, in a very short period of time and very focused on PR and media.

00:30:54.644 --> 00:31:09.349
So we're talking about, you know, content creation, media monitoring, assistance that are going to help you with those things, and how to sort of get more out of the tools you're probably using, which are usually ChatGPT, quad, perplexity.

00:31:09.349 --> 00:31:11.567
Those are the most common ones, I think, in this profession.

00:31:11.567 --> 00:31:12.509
So that's an hour.

00:31:12.509 --> 00:31:14.724
It's a very affordable class.

00:31:15.359 --> 00:31:30.000
You can find it at Mediaacopilotai and I also have longer programs in the fall, one for journalists and one for PR and comms, and those are more extensive.

00:31:30.000 --> 00:31:36.570
Those are going to be real workshopping, six-week courses.

00:31:36.570 --> 00:31:38.434
They're great.

00:31:38.434 --> 00:31:43.531
We see a lot of great feedback from them because they're really focused on changing the way you work.

00:31:43.531 --> 00:31:43.772
Right.

00:31:43.772 --> 00:31:46.749
It's not just demos, like we have demos.

00:31:46.749 --> 00:31:52.532
They're great, but we will handhold you through building them into your own workflows to accelerate what you're doing.

00:31:52.532 --> 00:31:55.478
It's like a real course, and we have coaching to your own workflows to accelerate what you're doing.

00:31:55.478 --> 00:32:06.087
It's like a real course and we have coaching, and we even have all of our students complete a project so that we will work with you to build a new workflow for yourself so that you're getting more out of AI.

00:32:06.087 --> 00:32:08.788
We see a lot of good response from that.

00:32:08.788 --> 00:32:09.884
We think it's really valuable.

00:32:09.884 --> 00:32:17.431
So, yeah, so we do a lot of classes in addition to the custom work we do at PR Team.

00:32:19.401 --> 00:32:21.429
Well, pete, let me say thank you so very much.

00:32:21.429 --> 00:32:38.713
That's the reason I called you, because when I read about Media Copilot, I said wow, this is something I think a lot of our listeners will have an interest in, because technology is moving so fast, just trying to keep up not only with what's coming, but also effectively use what's already here.

00:32:38.713 --> 00:32:50.853
So, the fact that Media Copilot is available, I'm glad that I had a chance to have you on, and I'm sure a lot of our listeners will take advantage of the fact that you've got something that they can use.

00:32:52.661 --> 00:32:53.605
Yeah, I appreciate that, Peter.

00:32:53.605 --> 00:33:03.750
Yeah, I appreciate that, Peter, and I know a lot of people sort of spin a little bit of what you just said in a negative way and they say, hey, you need to get skills or someone else is going to outcompete you.

00:33:03.750 --> 00:33:10.432
I like to sort of flip that a little bit because I know probably a lot of people listening are thinking, oh, I'm really behind the curve, I don't know this stuff.

00:33:10.432 --> 00:33:13.053
Believe me, a lot of other people don't either.

00:33:13.053 --> 00:33:17.954
So you know, be assured in that you're not alone.

00:33:19.336 --> 00:33:25.958
But you know there is a lot available to you and I'm happy to be your guide if you would like me to.

00:33:25.958 --> 00:33:29.262
It would be an honor.

00:33:29.262 --> 00:33:38.892
To be honestly, we could start really slow on one of my classes, one of my monthly classes, and then, if it's right you know, feel free to step up into one of the courses or even work with your bigger team.

00:33:38.892 --> 00:33:44.653
But don't panic, I guess, is my main message there.

00:33:44.653 --> 00:33:48.410
You might feel like you're at the beginning and you're overwhelmed.

00:33:48.410 --> 00:33:52.240
A lot of other people feel that way and it's just baby steps to the future.

00:33:52.604 --> 00:33:52.865
Mm-hmm.

00:33:52.865 --> 00:33:59.268
Well, once again, let me say thank you so very much to Pete Paschal, who is the founder of Media Copilot.

00:33:59.268 --> 00:34:05.848
And again, don't forget to take advantage of that free piece of information the 100 Top AI Prompts for PR Pros.

00:34:05.848 --> 00:34:08.789
That's available, obviously, at mediacopilotai.

00:34:08.789 --> 00:34:12.213
I've gotten a lot of mileage out of it.

00:34:12.213 --> 00:34:14.103
There's a few days I've had access to it.

00:34:14.103 --> 00:34:22.594
And again, thank you to Pete Patchell, our guest today, and also thank you, my listeners, for listening to the Public Relations Review Podcast.

00:34:22.594 --> 00:34:32.411
We certainly would like to get a review from you and also tell your friends about us, and we'll look forward to having you as a guest on the next edition of the Public Relations Review Podcast.

00:34:33.481 --> 00:34:43.768
This podcast is produced by Communication Strategies, an award-winning public relations and public affairs firm headquartered in Nashville, Tennessee.

00:34:43.768 --> 00:34:45.726
Thank you for joining us.