How Marcomms Teams Are Putting AI to Work

The hype around artificial intelligence is loud, but the real story is unfolding quietly in the daily workflows of marketing and digital strategists. It’s not just about ChatGPT or flashy tools. It’s about what happens when AI meets persistent pain points, hidden frictions and fresh opportunities in real work.

To kick off part two of their series on how AI is reshaping integrated marketing communications, hosts Anne Green and Steve Halsey discuss what it takes to move from experimentation to enablement—covering everything from legal and procurement hurdles to the importance of responsible governance and smart integration.

Then, Anne sits down with Loren King, Senior Digital Marketing Specialist at MorganMyers, a G&S Agency, to explore what that shift looks like on the ground. For Loren, that means building smarter systems for agriculture and food clients while staying attuned to how new technologies are shifting everything from brand perception to content discovery. With a background rooted in farm life and a front-row seat to rapid AI adoption, Loren brings a practitioner’s lens to a grounded, real-world conversation.

In this episode, Anne and Loren explore what digital fluency looks like today and how communicators can experiment with AI in ways that are strategic, ethical, and grounded in real needs. From automating repetitive tasks to preparing for agentic tools and AI-optimized discovery, Loren’s insights offer a clear path for professionals who want to stay relevant without getting overwhelmed.

We also discuss:

• Why starting with problems—not tools—leads to smarter AI use
• How image-based search is reshaping brand visibility
• What happens when custom GPTs handle the backend grind
• The new kind of fluency communicators need to stay relevant

If you missed part one of this series—featuring Greg Galant of MuckRack on AI’s rapid adoption in PR—go back and listen wherever you get your podcasts.

Episode Transcript

Speaker 1 (00:00):
I've been thinking a lot about a quote. Essentially, it
says every system is perfectly designed to output what it outputs. However,
a system is made whenever you're getting on the system,
that's what it's designed to do. So if you want
to change what's made, you need to change the design.
I'm not sure that AI fully fits that mold, because
you could be as accurate as you want and there's
still technically the potential for something to come out that
was unexpected.

Speaker 2 (00:30):
Hello, and welcome to Building Brand Gravity. I'm Anne Green,
I'm CEO and a principal at Shens Integrated Marketing Communications.

Speaker 3 (00:37):
Group, and I'm Steve Halsey, one of the principles and
chief growth officers of that same group, and welcome you
to our latest episode of Building Brand Gravity. And you
want to let everybody know what's in store.

Speaker 2 (00:49):
Yeah, we're doing a little two part series here, a miniseries,
if you will. The first one was focused on AI
research and sort of the evolution of where it's at
in our industry right now. That industry is not just communications,
it's really across marketing. And then for today, we're going
to be speaking with one of our own in the
group who I'll talk about in just a moment, discussing

(01:12):
about how we go from AI exploration really to enablement,
especially in the context of some of our largest sectors.
So I'm excited to build on the conversation we had
last time too.

Speaker 3 (01:21):
It should be great and Greg is always always a
great interview, great great friend of the pod, as I
like to joke, and my conversation with Greg Galan, who's
the co founder and CEO of MUCRACH, was really for
those of you. I hope you all listen to it,
but if you have it, you can after you're done
with this, you can go right over and click it.

(01:43):
But it was really on the results of their latest
research on the state of AI. Basically, AI is a
new normal in PR and there were some really interesting
facts that we covered there. Seventy five percent of PR
pros are now using generative AI. That's up from twenty
eight percent just eighteen months ago. Of those who are

(02:03):
using it, ninety three percent say it speeds up their
work and seventy eight percent say it improves the quality
of their work. Does that sound about right?

Speaker 2 (02:12):
I think it's adoption is growing fast. That feels very
right to me, you know, as I talk to leaders
across the industry, as I talk to practitioners, and again
there's comms, there's you know, paid media, the average, there's
creative digital. All of our teams, the teams that I
know at other agencies. Adoption is growing fast. Some of
it is because AI is being baked into so many

(02:33):
tools we use. Mucrack is one of those, Adobe Firefly,
et cetera. But others are people being very intentional and
agencies and clients side being more aggressive about saying we
need to weave this into our workflows, we need to
disrupt our own workflows, and we need to bring this
technology and in all kinds of ways to augment what

(02:53):
we're doing. So I think those stats are powerful and right,
and I think it's only going to grow.

Speaker 1 (02:59):
Right.

Speaker 3 (03:00):
There were two more stats in there that I thought
was really interesting. At a recent awards ceremony, I ran
into Kim Sample of the PR Council, and for those
of you who don't know it, Ann's on the Ethics
Committee and she's really worked to really help put together
a lot of AI guidelines and guidance for agencies. But
there was there were two stats that I found fascinating

(03:23):
in Muckrax's research, only thirty eight percent of PR pros
report having company guidelines and only forty three percent of
AI pros report access to AI training, and even a
smaller percentage always disclosed. At the number I'm just looking at,

(03:45):
only twenty percent of agency pros say they always disclose
when they use AI. And you know, we're seeing right
now some interesting dilemma of we all see the power.
You know, seventy five percent are using it On the
client side. I'm hearing that a lot of brands want
to use it, but there are two things that are

(04:06):
causing some challenges our favorite legal and procurement. So I'm
interested to get your take on that before we get
into Louren's interview. I guess in my mind I jotted
down a list of there's three essential things for us
to think about when we're talking about how we use
AI for media relations, content creation, or analytics, and I'd

(04:27):
love to get your reaction to them. Number one is
defining the acceptable use and boundaries up front, where we
specify how generative AI may be used for strafts, research,
media lists, enhancement, sentiment analysis, things like that. Number two,
making sure you align on the data privacy IP and
confidentiality protection so you're very very clear that you're only

(04:52):
going to use AI tools and not train them on
any data other than what you agree to. And then
the third one is to really codify that governance, that disclosure,
that risk mitigation, share, our AI usage policy, all of
those due diligence. So those are my top three. Define
the boundaries, align on privacy, and really codify your governance.

(05:14):
I mean, how does that tie into what you're here?

Speaker 2 (05:18):
I like those a lot. It's an interesting moment for sure,
and I appreciate, you know, the work with a PR Council,
and it's so funny. You know, obviously the name PR Council,
all the agencies that are our peers that were highly integrated,
so there's marketing, advertising, creative, project management, you know, all
the pieces that reflect who we are today. So it's
pretty multi layered. But looking at it from a communications

(05:39):
standpoint in the toolkit, I have enjoyed being, you know,
one of the heads of that ethics task force relative
to how do we use AI in the best way
with the smart guardrails, because as I say many times,
you know, I'm out there talking about this a lot
there's two sides to every coin. There's the amazing power
of it, and then there's the with power comes great responsibility,
and there the challenge of it too, the deep fakes,

(06:01):
the you know, the AI slop at scale and things
like that. Right, So, what's interesting about our colleagues and
procurement and legal so being on the agency side obviously
and working so closely with clients, those are their peers
and house these are our partners and peers, you know,
across that partnership relationship. There it's coming at them fast.

(06:26):
And the problem is there is a lot of challenge
with these tools. There's protection of intellectual property, there's hallucination.
Hallucinations have been reduced, but not entirely. There's the question
of the transparency of usage. There's a question of I
think there's going to be in a special question of
like what are we paying for, you know, the machine
and the human What is the nature of this relationship.

(06:46):
So one of the watch outs we're seeing is that
as these you know, respected groups, they play a really
important function in our client organizations. Obviously, they are trying
to catch up to something that's just a whole versus
running out of the barn. It's like a whole fleet
of horses. It's just a stampede flying out of the barn,
and they're trying to catch up to put it in

(07:07):
language and legal documentation and questionnaires to be like, how
are you using this? So in some cases we're seeing
organizations that are very large come back with AI addendums
that are huge, long, and overly restrictive because the fact
is AI is baked into everything. I mean, I got
my little phone right here. It's everywhere, and it's baked
into the tools, and it's being baked in in ways

(07:30):
Greg would say this where it's changing every day. So
those overly detailed and restrictive and kind of saying, well,
you almost can't use AI at all. It's just not feasible.
And I think those groups know that others are trying
to do. I think what you SAIDs to you, which
is how do we make sure we're talking about use
cases to begin with? How do we make sure there's
a transparent discussion, Like I would like my own people,

(07:52):
you know, our people at GNS to be transparent about Hey,
I got this. I'm using generative to get me going
on a draft, but I'm also using my brains and
my expertise to edit it and make it better. So
I think that we're seeing the normal sort of churn
in this fast moving landscape. But I don't envy our
legal and procurement colleagues. This is challenging.

Speaker 3 (08:15):
It's definitely challenging for them, and definitely people on the
brand side too with their teams to work through. So
I guess what I would say is, you know, we're
all in this together. So you know, if you're an
agency listener, let's work with our friends at the PRC
and really help kind of educate. If you're on the
client side, you know, just really think about even leaning

(08:36):
in and reaching out to the PRC yourself even though
you're not an agency for best use guidelines, so that
we can help you kind of work within your system
with procurement, with legal to really put together something that
makes sense for everybody.

Speaker 2 (08:50):
I agree, because ultimately it's our partnership together that's going
to find out where the value in our work will
be long term, where we automate roat tasks and get
faster to the value that matters. And we're only going
to figure this out agencies and clients together, you know,
really in concert, and that's actually a good transition to
our guests for today. Our guest is Lauren King. He

(09:11):
is a senior digital marketing specialist at one of our
two agencies, Morgan Myers, a GNS agency. He is actually
just about he's celebrating his third anniversary at Morgan Myers,
but he has a lot of digital experience prior to that,
and he's part of our AI exploration and enablement team,
so he's very, very knowledgeable. We have some super users

(09:32):
across our two agencies, across the group, every specialist group,
every you know, sector and discipline that we offer, and
that group's been such an amazing brain trust as we move,
you know, from more explanation exploration to rubber on the
road and putting all the tools in people's hands, doing
all the training and starting you know, a deeper transformation process.

(09:54):
So what I wanted to talk with Lauren about. Greg
gave us that ten thousand or one hundred thousand foot
view how adoption is changing, and I think that conversation
you guys had had so many cool aspects to it.
He's got a really great bird's eye view. Lauren is
in the ground, especially literally on the ground with our
clients and AG and our clients and food. He is
a real specialist, excuse me, in many aspects of AG

(10:18):
and technology is adoption. So what he and I are
going to talk about, and I'm excited for folks to hear.
You know, what are we seeing on the ground in
terms of tools and approaches and use cases and also
driving more engagement? How is he an advocate for that
self education and upskilling that we all need to do,
both as companies but as individuals. And the third view

(10:40):
he has is it's really hard to just say, well,
what can AI do because the answer is everything. It's
more about what problems can we solve? So let's listen
to the conversation with Lauren and Steve. I'm sure we'll
have stuff to talk about afterwards.

Speaker 3 (10:55):
I can't wait.

Speaker 2 (10:58):
So here on building brand gravity. We are interested, like
everyone else in the world, in the evolution of AI,
especially in the marketing and communications space in which we
work across many, many sectors. And I'm excited to have
a conversation with a colleague of mine, a fairly new
colleague of mine who came to the GNS family through
our acquisition of Morgan Myers, an amazing agency that focuses

(11:19):
on agriculture and food, and that person is Lauren King. Hey, Lauren, how.

Speaker 1 (11:23):
Are you doing pretty good? Myself?

Speaker 2 (11:25):
Excellent? Excellent. So just a little background, Lauren is a
senior Digital marketing specialist at Morgan Myers working on many
different agriculture and food clients. But I love, Lauren that
you have such a background that's really rooted in that space. Literally,
you know, a farming family. You were growing up in

(11:46):
southwest side of Michigan and now you live in the
southeast side, and you and I were kind of spiritual
neighbors because I was born in Toledo and I've got
a lot of family there right down. I sebenty five
from you guys and educated at Michigan State University. And
I think was really interesting for this conversation is not
only is Lauren a specialist in digital and all those
pieces you know, very much an integrated set of capabilities

(12:10):
as many of us work in today, but also doing
things like sharing the American Farm Bureau Technology Advisory Committee
and working with Agriculture Future of America Alumni Advisory Committee
speaking at you know, different types of venues that are
digging into AI. So just to get us started. Laurens,

(12:30):
So tell us where you're coming from today and what
you know, what home is like for you in that
part of Michigan.

Speaker 1 (12:36):
Yeah, like you hinted that, I'm in southeast Michigan, which
is between basically Toledo on one side and ann Arbor
on the other. That's the easiest reference for many people.
Right now, I look out my window, I can see
at least one tractor and plantter accommodation. I think they're
planting soybeans, but the fields by me are a mix
of soybeans, pumpkins, corn. We did have peppers last year.

(12:58):
We'll see if that happens again. So we're very, very
very diverse, very agriculturally diverse. I'm ten minutes from the
nearest town, which is a small main street, and that's it.
I'm twenty minutes from the nearest Walmart, and that's largely
my experience growing up. But thankfully we have great internet, yeah.

Speaker 2 (13:13):
Which is very important, especially for today when you and
our meeting. You are in Michigan. I am in New
York City and we are totally connected and live.

Speaker 1 (13:21):
Which is great. I love that.

Speaker 2 (13:23):
So you're working heavily in digital but when I first
met you, we bonded over many things, many different points
of interest, but especially the evolution of technology today in AI.
So first I'd like to hear a little bit about
how you describe your digital role at Morgan Myers as

(13:43):
part of the g ANDS group, and then after that
I have a follow up question, but why don't you
talk about the kinds of work you're doing now in
the digital space.

Speaker 1 (13:51):
Yeah. So I'm part of our digital team at Morgan Myers,
which means that we touch a lot of different clients.
I don't have one specific client that I work with
day to day, although Mert, Gantal, health, Cattle would probably
be my primary and my role specifically ties a lot
into systems management, whether that's the back end of digital ads,
our social listening capabilities, so tracking trends in real time,

(14:13):
deploying AI when possible. That's been a big part of
what I've done the last six months to a year
or so. And websites really anything that a system within
digital touches is where I tend to specialize. So that
can take the form of pushing out an AD campaign
one day, the next day, it's understanding how a client's
product is viewed online and then the day after that

(14:34):
having a conversation like this focusing on how we can
use AI tools in different ways within Morgan Myers, I.

Speaker 2 (14:40):
Love the agency contexts and especially the kind of role
you play because it is so diverse, and it's also
naturally keeps you as a learner. You have to continuously
learn and grow. So talking about learning and this conversation
you and I are having is part of a small
series we're doing on AI as part of building brand Gravity.
Been having ongoing conversations with something I'm obviously very interested in,

(15:03):
as you know, but I want to go way back
in time because it was so interesting talking to you
about this and preparing for this conversation. When did you
start your AI journey? What was the thing that alerted
you to this coming wave?

Speaker 1 (15:17):
So you know, as a kid, we always had stories
with robots and even Star Wars and examples like that.
But when it really hit me that AI could be
something that could be used was around twenty thirteen, And
a lot of people don't know that Google basically pioneered
a lot of the innovations that make tools like CHATGPT,
and copilot accessible and available today. So they were doing

(15:38):
some work. There was blogs coming out kind of explaining
the underpinnings of how this works, and this is far
beyond what I could understand at the time, but it
got pres enough press and buzz that I was able
to take a look start hearing words like LM or
large language model for the first time. And when I
looked at it, of course, even though it was in

(15:58):
a language I didn't really speak, which is out rhythms
and a data science, I was my interest was piqued
because I didn't realize that you could actually do anything
with artificial intelligence. To that point, I thought it was
something held in asimov novels in the Future of Star Trek.

Speaker 2 (16:13):
I love that. I mean, that was twelve years ago,
and clearly there's many folks out there, maybe some of
our listeners that understand that AI has been out there
developing for quite a long time, not just theoretically but practically.
But I love the idea that your journey has been
so long and that you started to try to get
hands on right away. Because you and I have a
little bit of an age difference, I'm going to assume

(16:35):
you're quite young, you know, twelve years ago, but I.

Speaker 1 (16:37):
Twenty seven, I would have been fifteen.

Speaker 2 (16:39):
Okay, yeah, exactly, So that's what I was kind of guessing.
I had started hearing about it as well, And I
think it's interesting to think back to that moment, as
you said, like, what did you think it was? So
when you first heard terms like large language models, well,
you first heard terms about the idea of AI. I
don't know if generative AI was used in that context.

(16:59):
Sender AI over really in common. Yeah, I feel like
that has been especially chat shep too. One of the
biggest innovations that common nomenclature of generative AI is something
different from maybe predictive or But when you first heard
of large language models and AI over, you know overall back,
you know, when you're a teenager, still, what did you
think it would be and what you could do with

(17:20):
it versus how you're seeing it and using it today?
What's that delta between the two.

Speaker 1 (17:25):
Yeah, So when I was hearing about it, it was
very much within the real of academia that it was
being considered unemployed and where I was thinking it would
be used. So I'm picturing, you know, giant Excel files
with millions and millions of lines, letting some type of
machine learning tool loose to analyze it and find some trends,
or maybe being able to go back and review what

(17:48):
has already existed, or apply government statistical analysis to data
coming out of the USDA. Those types of intense situations,
you know, not something that the everyday person would probably
run into very much, almost validating existing data more than creating.
I do think that I read a little bit about
creating images to the point of they thought it would

(18:10):
ever work because there was no way to make the
hallucinations coming out of an AI tool cohesive in a
way that somebody would actually get what they want. You
could make it make something, but it didn't actually make
any sense. Kind of like the early days of DOLI.
I'm sure if you experimented, whether you've got some really
interesting stuff. I've got some stage down on my phone.
But for me, it just seemed like it would be

(18:31):
stuck in the realm of academia and statistical research, and
it wouldn't ever be something that I could truly see
myself using my dad, using on the farm, or deploying
in a business context. I was very cool.

Speaker 2 (18:43):
I think my version of that is, you know, in
the sectors I've worked in, which are many including agriculture now,
which I love engaging in, and you have so much
stuff there. But I would see it in terms of
health and medical innovation, so things like high throughput screening,
which is critical to the drug discovery process, you know,

(19:04):
things like other types of like the way that alpha
fold is working on protein folding, and so I was
thinking about it more in those scientific context too, and
it's very interesting now to see it leap into a
space that is really transforming everything we touch. I think
the other piece for me that made a lot of senses.

(19:25):
There was such a buzz even almost that far back,
about big data. You know, if you might remember, that
was such the buzzword. There's always a buzzword, but big
data was a thing, and I think what the issue
was was big data was there, but what do you
do with it and how do you deal with it?
Especially when the big data is big, like you said,
the USDA stats or data points, So you know, I
really resonate with that. So there's been so much going

(19:47):
on and the evolution is so fast, like you said,
if you you know created pictures in Dolly a few
years ago, they are quite amusing compared to what it
can do today, which is.

Speaker 1 (19:55):
There were some interesting ones, yes, and all.

Speaker 2 (19:59):
The weirdness with human form and human hands on all
the stuff we know. Right, it feels like so long ago,
but it was only like a year and a half ago.
But when you see everything that's evolving, and you and
I talk about this a lot, and our AI enablement
team across both GS Business Communications and Morgan Myers talks
about this. What are too some of the most notable
and important evolutions that you've been tracking that you think

(20:21):
are really material to the work you do and that
we do well.

Speaker 1 (20:24):
We actually have one that's new as of today or
within the last twenty four hours, which is open AI
focusing chat GPT as in everything app going beyond search
and creation into I'm not sure if you've seen their
shopping edition that they have, Yes, So they're integrating with
Shopify and they're going to be recommending products based on

(20:45):
very specific search terms, no ads, only going off of
the metadata of the product and your personal preferences from
your prior time using the tool. So that'll primarily apply
to paid accounts. But if I have been doing a
lot of researching for a trip to Italy, it'll incorporate
my preferences when I then ask it to recommend a
bath towel that fits for a beach and can go

(21:05):
into luggage. It's a big change, and it's a reason
to not leave the app. It's a reason to have
better SEO if you're working from a marketing perspective, because
that's going to be really, really important and it changes
the way that consumers are going to interact. But it's
also changing how I'm going to interact with a tool
like chat, GPT or copilot will have it before too long,

(21:27):
everybody else will as well. Coming out of that, hope,
please go ahead, I was gonna say. Coming out of that,
another one that comes to mind is integrating with other apps.
We're still on the edge of what that looks like
at you know, I can't log into my Facebook account
through chat GPT yet. I can't log into my United
Airlines account through chat, GPT or copilot yet. But once

(21:50):
you can, that's when eventually you get to agentic tools,
which I'm sure we'll touch on a little bit more,
but you know, making my digital life more cohesive and
having an assistant, so taking the shopping sign and becoming
everything up is one connecting to all your other apps
as the other. Those are the two that I've been
paying the most attention to.

Speaker 2 (22:07):
I think those are so material to our work and
to our lives as humans. I mean, that's the thing.
This is going to be woven in absolutely everywhere. And
the way we talk about AI today and by the way,
one of the things, you know, Lauren, that really bugs
me is when we're not precise when we talk about AI.
We just use this big old acronym when you know,
are we talking about something that's more predictive, cognitive generative?

(22:30):
But the fact is is that it will be that
inside everything and sort of the web that connects us
in a way. And it is interesting you talked about SEO.
There's been so much discussion recently about is this the
end the death of search engine optimization. I tend to
be very skeptical when anyone says something is dead, because
it's usually alive somehow. You know, my agency life would

(22:52):
have been gone a long time ago if I listened
to all of those things, because we're still here. But
this idea of AI optimization, you know, how are you
optimizing for the large language models? And how they're crawling
the web in a very different way and how they're
ingesting information. And so I love that you're thinking about
that from a digital perspective because our clients absolutely have

(23:12):
to think about that.

Speaker 1 (23:13):
Right absolutely, and not just SEO, but how SEO and
your imagery work together. It's a little bit of a
unique experience. But we've been developing a chatbot for a
client that connects to food, and one thing that emerges
when you're making the chatbot is that it was scanning
not just the recipes that we had, but it was
scanning the images associated with the recipes as well, and
it would recommend something different. So, for example, if you

(23:35):
were to search for like red, white and blue or
fourth of July, all of a sudden you're getting, you know,
foods that aren't connected to fourth of July, even though
they have some in this website, but show cherries or blueberries,
and then like a whitish looking dough in the meal
it's or in the baked good itself, and it understood
like the tool understood that. And so even your imagery

(23:57):
becomes an extension of your SEO in a sense. If
you're trying and predict what people are looking for.

Speaker 2 (24:01):
Yeah, it's kind of like a supercharging of what Google
has tried to do with Google Image Search. And obviously
Google is working hard the way it's incorporated Gemini at
the top of the search page not to lose you know,
its brand and the dominance has had. But I think
it's going to be really, really hard at this point
for Google to maintain a choke hold on that. We're
already seeing so much use of chat GEAPT just as

(24:24):
one example, through the app, and from a business development perspective,
you know, more and more we're hearing, oh, we found
your agency through chat GPT, which is not to surprise
me at all when we see the kind of search
volume that's growing around it.

Speaker 1 (24:35):
A billion searches last week.

Speaker 2 (24:38):
Yes, so so much changed. So thinking about the ground
level of this, and that's why I was so excited
to speak to you. You know, like many firms, and
especially on the agency side, because I talked to folks
on the corporate side agency side, many of our clients.
We don't have a lot of restrictions on the agency
side about how we experiment. We have very specif pacific guidelines.

(25:01):
As you know, Lauren, I'm very focused on ethics. We've
been part of the PR Council's work and ethical use
of AI for communications, respect intellectual property. We're very careful
about IP protection and also the ownership of creatives. You
know how things are being used right, So we take
that really seriously. But in terms of experimenting on the upside,

(25:21):
there's very little to hold us back. So what I
wanted to talk to about today was really that ground
level how do digital and marketing and comms professionals, how
are we benefiting from use of ata AI in our
day to day work? How are we exploring and what
are we hoping our colleagues do, say at GNS and
Morgan Myers and other places, And you know, I see
it as there's the working smarter, not harder. They're solving

(25:45):
persistent problems and pain points. There's knocking out those rote
tasks where you're like this has taken away too much time.
You know, I need an assistant to do this, or
my favorite, which is enhancing our strategic value. So when
you think about your workflows and those are your peers,
what are some of the either tools or platforms or
workflows enabled by AI that have really changed your workday?

Speaker 1 (26:06):
The easiest is something that a lot of people probably
wouldn't consider but really has been a game changer for me,
which is automating a very repetitive task of receipt analysis.
We have a lot of receipts come through social media
that need to be split out. For those not familiar
with an agency structure, you're often going to split them
ount on a receipt between different jobs or outlines, each

(26:28):
with their own budgets, if you will. And because social
media platforms don't have the ability to easily update how
a receipt comes to you, you're kind of getting everything
at once. So if I'm running two different campaigns, they're
both going to show up on the same receipt well.
As a result, something needs to go through that and
figure out how much money goes where and then note
it for processing. From our financial side. It sounds very boring,

(26:51):
and it can be very boring, especially, but it can
also be a lot because we have campaigns where we'll launch,
you know, all of a sudden, twenty thousand dollars or
one hundred thousand dollars in a month, and when we're
limited in the size of the receipt, that's one hundred
receipts we have to process. So we have a custom
GPT that's trained to analyze receipts and provide a naming

(27:13):
structure and the math that we need only for us.
So it's trained in the style that we're trying to
go for, it uses the structure we want, it's trained
on our examples. It does not connect to the Internet
in the traditional sense, only through a GBT interface. And
although it's only ten receipts at a time, it's a
game changer for me. My personal estimation is a thirty

(27:34):
to forty percent increase in processing speed more if there's
only a couple different jobs at any given time. But
we have receipts that can go up to seven jobs,
and so I've shifted from having to do all of
the math myself to double checking what's coming out. And
so far there's been one inaccuracy in the AIS part

(27:54):
and five inaccuracies on my part that I mess up
on the re seat structure beforehand, and a ninety eight
percent success rate, I would say, without really any changes
needed at all. It's hard to overestimate that.

Speaker 2 (28:05):
Yeah, I love that example because that's an example of
a persistent problem it's kind of shocking to me. I mean,
no shade to the social media companies, but the way
the billing structure and the receipts come, well, I am
throwing a little bit of shade.

Speaker 1 (28:18):
It's just it's okay, throw the shade.

Speaker 2 (28:20):
Crazy from a business perspective that this is still how
the billing is done. And for any paid media teams.
We have the Morgan Myers team doing paid media, we
have the GS and G and S Business Communications team
doing paid media. It really is a burden. So I
think you've hit those buckets of knocking out road tasks
and solving for persistent pain points. The other thing that
brings up for me is the idea of AI as

(28:42):
an agent that assists to you. And I think you
know we've thought in terms so we first we had
to learn to think about the Internet as an entity.
Then we had to think about kind of that social Internet.
Then we thought about apps. We got trained to think
about things from an app perspective, not just through mobile.
It came through mobile, but now it's like even apps,
you think about it even in a desktop perspective in

(29:03):
a way. So now we're being trained to think about agents, right,
So how would you define for our listeners. You know,
hopefully people are following this, but how do you think
about that word either app or agent in the context
of some of the things that you're starting to build
and use.

Speaker 1 (29:17):
So, for me, an agent i'd consider as something that
can persistently make its own decisions, whether that's allowing it
to operate off of its own schedule, control a specific
piece of software. There's a little bit of nebulousness when
it comes to what is an agent, and I think
that's okay because this is all still so new that
we're not going to have like exact definitions. But it

(29:37):
helps even in the name itself, anybody largely understand what
it is. Think of an intern who you trust enough
to do a specific task for me. I'm exploring a
lot of ways that we can set up agents, either
for creating projects, building out things ahead of time, or
on our social listening side, which I help with a lot.
That could be competitive analysis, so repeat competitive analysis, just

(29:58):
updating me on key and ahead of my chance to
actually review the results that are coming in. So it's
almost priming from an almost real time perspective my work
as I go in to understand a competitor and what
our clients need to know on them. So still looking
at the structures for those.

Speaker 2 (30:17):
Yeah, definition a way of thinking about it, and you're right,
I think we're reaching toward that and we're hearing what
the promise, so that will be It kind of harkens
back to what your trends were regarding connection and connectivity
of the major lms with other parts of our lives
like the shopping piece or social media, airline booking. You know.

(30:40):
One thing that's funny about those connection points is I
actually ran into this this week is with the note
taking apps like Otter and read dot ai and how
you like open the door a tiny bit and they're
in there connecting to everything. And suddenly you have a
read dot Ai reader showing up to your team's meeting
before you get there. And I'm like, I did not
intend this, my friend, but thank you so much for

(31:02):
being so aggressive.

Speaker 1 (31:03):
Yep, getting emails saying oh, here's your recap of a meeting,
not even realizing that the assistants plugged in there, which
double edged, but so far has been beneficial.

Speaker 2 (31:12):
No, it's very funny and obviously in our environment using
Microsoft three sixty five and enabling copilot for all staff
where we are starting to build out apps, you know,
I think they're almost pre agents, but building out tools
in every part of our business, from the client side
to operations to everything it support, et cetera, et cetera,

(31:32):
using it to work smarter, not harder. It kind of
points us in that direction. I read an interesting article
and I'm CEO at our group about how you know
this generation of CEOs will be the last to only
have human employees will also be agent employees. And that
to me was one of those moments like you describing, Hey,
this is how I understand agents that they have some agency.

(31:55):
I had to stop from it and be like, that's
really interesting because I feel like I've been managing technology
strategic asset for a very long time, but understanding without answerpomorphizing,
understanding like agents will be part of that workforce I find.
I mean, does that what do you think about that?
How does that sit for you? Do you feel like
that's the future we're heading towards.

Speaker 1 (32:13):
I do think that's the future. And I've been thinking
a lot about a quote. I'm not going to get
it quite right, but essentially it says, you know, the
output every system is perfectly designed to output what it outputs.
It's like however, a system is made. Whenever you're getting
out of the system, that's what it's designed to do.
So if you want to change what's made, you need
to change the design. I'm not sure that AI fully
fits that mold, because you could be as accurate as

(32:34):
you want, and there is still technically the potential for
something to come out that was unexpected because we don't
truly have a grasp on how the back end technology works.
That's like a human. You know, I've been there myself.
I've been an intern, and I can be trusted with
the best possible instructions and go horribly wrong or go
the wrong direction with it. So there is a bit

(32:55):
of a management approach. It's it's reflecting, it's responding, encouraging,
ask maybe for redo on things. So is there skill
sets that people probably didn't expect and certainly not ones
you'd apply to technology in many cases? You know, when
IT departments started to exist, people might have managed computers,
but they probably expected the computer to be accurate all
the time. And while AI is getting more and more

(33:17):
accurate over time, there's still going to be that element,
whether it's coming from the fact that we're using natural
language conversations that could be part of it because language
is tricky, things mean different things of different people, or
if it's just going to be an inbuilt piece of
the tech as a whole. My guess is, and like
what you were reading and like others have reported on,

(33:39):
management skills are going to be very, very common all
the way down to really anybody in an organization that
touches a computer.

Speaker 2 (33:46):
That's a really great point. I like that a lot.
You've made a point in our AI team meetings, and
we've actually been trying to advocate for this across organization
and also with our clients, which is to get over
the overwhelming nature of AI, because not everybody is feeling
like immediately super technical or hands on like some people

(34:07):
have been exploring it for a very long time as
you have. Others are coming to it in the course
of their job and if there's a big self education process.
But one of the ways we've framed it is how
do you start with a need or a problem to
solve and think of that first versus what can AI
do or what should I do with AI? Why do
you think that's an important way of thinking about this,

(34:27):
especially as practitioners in our space.

Speaker 1 (34:29):
I think it's the easiest way to start understanding the
capabilities of a tool like AI. Maybe we can have
a thousand use cases, we can have a prompt library
that does anything you can imagine, and yet tomorrow we're
going to find out all of a sudden you can
shop in chat GPT. Now, like, there is no predicting
what this tool can do, and there doesn't at the
moment appear to be an upper limit, because if there's
anything digital related, it probably can. But as a person,

(34:53):
you know what your challenges are. You understand better anybody else,
not only what is needed in your role, but so
what is sometimes wrong. I think a good way to
position this is you know you're still an expert in
what you do, So if you're using a tool like
artificial intelligence, you can spot when the output doesn't align
with what's right. And so if you start thinking about

(35:14):
ways that you can solve something that's really annoying for me,
that was receipts, it can open up a whole world.
I mean, eventually I was getting recommendations for how to
code like hard code in analysis tools and connect back
to cloud, APIs and all that stuff that's not in
my wheelhouse. Although I feel a lot more confident now
because I got to go through that exploratory journey. Thankfully,

(35:37):
I found that chatchpt was able to do it on
its own using existing systems with the customer GPT framework.
But that opened my eyes in a way that trying
to read every study never could. Now, when I go
and I come across the new problem, the way I
even phrase it or coach it to an AI tool

(35:57):
is going to be different because I have a better
understanding of how the technology works, which I think is
really important. A lot of people don't give the chance
to really know, like how something is generated back to you.
So that's number one, But then also how it solves
home my existing pain points in ways that I didn't expect.
I didn't the outcome wasn't necessarily what I expected when

(36:17):
I started using the tool to solve a pain point.

Speaker 2 (36:19):
I think it's going to be transformative for us as
we continue bringing all of our colleagues along in this journey,
and everybody has so much to bring to the table.
I'm seeing such creative smart but also you know, appropriate
uses across our agencies and it's really gaining momentum obviously,
but I don't take that for granted because it has

(36:39):
to be super intentional. And what I love is the
idea that we have folks around Morgan Myers and she
and US Business Communications and on our client side too
as partners, different people who are more into this or
have skill sets or they have more experience where they
can say, hey, you considered this, and it's really more
of a consultative mode. It's like the way which we

(37:02):
work with clients, which is come to me and ask
me your questions, ask me what you're trying to solve for,
and I bet you we can find a use case
because you're right. Just listing use cases will make you
insane because there's so many of them.

Speaker 1 (37:17):
And a big piece of this that people often don't
consider is that you can fold a lot of project
and then use aid to figure out what you missed.
That's another way to consider it because you can almost
do a negative parameter, like say I'm planning in events
and you know, I've kind of got a location down,
I've got a lot of themes. Now I have an
idea of the weather, I can upload a version of

(37:40):
that event to chat gypt and say this is what
I already know. What have I missed? It sounds simple,
but you might be surprised at what comes out to
fill in the gaps you might not be aware of.
So that's another spot that we've been exploring. I know
that our SEO experts have been looking at websites and
they're starting to say, hey, you know what is missing
from this FAQ that should be here would support our

(38:00):
SEO efforts.

Speaker 2 (38:02):
Just for example, that's a great example. One of the
concepts I came across. I think it was on Freakonomics.
It was on one of the Freakonomics podcasts with Stephen
Dubner and they had an expert on. It was a
series he did on failure, and he had an expert
on talking about the idea of kind of projecting forward
because you're starting a project to sort of stand at

(38:23):
the end and say, imagine we failed, what went wrong?
And I think that that, to me is a trick
of the human brain to get you out of the box,
like the metaphorical I'm in the box, like how do
you get out of the box? Which is a real
pat thing, but there's a real reality to that which
is I need to get like physically mentally in another space.
And I love that idea of using chat GPT to

(38:43):
do that very experience. You know, either what am I
missing or what will this is the plan? What might
have failed? If we fail? What will have gone wrong?
And it's another way, and that's really again augmented intelligence.
It's pushing us to think about things differently with a partner,
which is.

Speaker 1 (38:58):
The way to think about it. You're unlikely to still
get a result from AI that can really compete with
an expert in a field. It can occur, it's possible,
but if you want repeated excellence, you're better off training
your expert to use an AI tool to augment what
they do, and then your results are going to be
much stronger. I mean, you've read cointelligence. You know that

(39:18):
the data backs that up as well. Human plus AI
is much better than human or AI on their own.

Speaker 2 (39:23):
And you know, speaking of that as we start to
wrap up agriculture and food, but both have a lot
of specialization, especially agriculture. You know, you look at image generation,
is the corn that's going to be generated looking right?
Is that the correct heat is that? You know, it's
just those things that those of you've worked in that
field for years or grew up in that space would know. Intuitively,

(39:47):
it's hard for systems to know. So when you think
about your peers and colleagues out there, especially in the
AG space or the food space or a digital side,
what are some tips for those really committed to engagement
and self education and getting stars. How would you counsel
folks to think about this or dive in.

Speaker 1 (40:03):
I think there's a good three step process that I'd
first recommend, and number one is just simply start prompting,
start asking questions. When chat GPT was released, my initial approach,
which I'm not gonna say is correct for now, but
what worked then was to do a different prompt every
day for a month, whatever I could think of the
most unusual things. Act as though you're this famous person.

(40:24):
What do you think about this historical event? Outline the
recent news. Can you help me write some code? Really
anything you can imagine, usually even connected back to your
personal life. Start up deploying that in a way that
you get a sense of how the tool thinks and
how it works, and really any artificial intelligence option will
work for this copilot. Chat GPT, Google Gemini. They all

(40:47):
should be open for that. But as you're doing that,
you're going to naturally learn how to prompt better because
you're not going to like the responses. You know an
AI might learn from you, which you're going to learn
from it at the same time. So as you're doing
that process, explore prompt recommendations. We've got some really good
frameworks internally, the frameworks out there that you can use.
One of my favorites is to make sure that the

(41:09):
tool you tell it how you want to think. So
act as though you're the CEO of a large marketing firm.
How would you respond to the strategic plan? What did
I miss? Those are the kinds of things you would do.
Don't be afraid of using natural language. My first reaction
was to maybe go with the Google search approach of
this is really an index that I need to specify

(41:29):
and I kind of understand how it works, so I'm
gonna avoid sounding like a human. I'm just going to
search Query doesn't work as well. Natural language is totally
fine to use because the tool adapts to it. So
once you've done those two things where you've started prompting,
and then you start understanding how to change your prompts.
Then you're largely set to go into the broader area

(41:50):
of a whatever that is, whether that's back end technical
analysis for you, image generation, or strategy. I think that
those are the initial things that I would and they
sound basic, but really you need to have ten hours
or more into these tools before you start to do
it right and start to really like your results. My
first AI generated images were horrible. Something between a peacock

(42:12):
and a John Dear attractor came out. I don't really
know what happened there, but not as cool as I thought,
or not as cool as they described as. But nowadays
I'm a lot more comfortable with it. I understand the
parameters and how the technology works, and so not that
I always get everything I won't wish for, but I'm
able to get closer to my ideal. So using those

(42:36):
as difframe as the approaches is what I would do.
And once you've got that. I know it sounds odd
to leave the technical learning and the framework till after
you've experimented, but I really do think experimenting is more valuable,
and then you can go back and say, okay, here's
this system I learned online called craft, which is one
of our internal frameworks that we deploy. How can I

(42:57):
think back on what I've done before in a way
that would make it or had I used this? So
you need to grow at the same time that the
tools learning about you.

Speaker 2 (43:04):
Yeah, that makes I love that. And part of it
is really just getting hands on with it because one
of the things I talk about a lot. You've heard
me talk about this is that when there's something that's
truly very new and you've grown up in a different
time a different set of tools, interrupting your own workflow,
like intentionally is hard because your brain doesn't naturally like,

(43:29):
oh yeah, I have all these uses for AI because
it didn't exist before in that way for you, for
most people. So I love that and I think the
idea and you mentioned co Intelligence, a book by Ethan
Mollick of Stanford. For anyone who hasn't read it, it's
a good I mean it's it feels like again it
was published a long time ago because everything's moving so fast.
But he is really looking at it from a very

(43:50):
human but also a business leader, not a technologist, like
his academic study is business. And so I think he's
thinking about it in a great way and to put
in ten hours to get for me a lot easier
than ten thousand, which is the old mastery you know level.
So getting hands on is great. So there's always one
final question here on building brand gravity, which is what

(44:11):
has you in its gravity these days? What are you into?
What are you reading? What are you thinking about? Lauren?

Speaker 1 (44:16):
Those two that come to mind for me. First off
is the book Doune. Obviously we've had some good Dune
movies come out, but there's a very interesting approach to
artificial intelligence in that series that is informative. It's often
not really considered, but if you like to approach big
ideas through the vein of fiction, that's a good way
to do it. Another one, and I'm really showing my
nerd card here, but and Or, the Star Wars TV

(44:39):
show has just released its second season, and unlike many
there isn't a lot of this technological diployment. It's made
like a traditional movie. There's a lot more real sets.
There's a lot less you know, CGI backgrounds, which I'm
drawn to as I get more into using tools like AI.
I also, on the opposite side, like to go more
towards fully human created things for my entertainment. So I

(45:02):
don't know why that is. It's just kind of the
way my subconscious has been going recently, but I've been
drawn into both.

Speaker 2 (45:07):
I love it. Those are great recommendations, and I think
the continuing relevance of fiction like done, you know, as
we come into these more and more modern societies, you know,
sometimes it ends up like very this topic. But I
think also it's really interesting, like you mentioned as im
Off early on, like how do we as humans think
about living side by Sideman's machines, the good, the scary,

(45:28):
the bad, and let's try to create the world we
want to live in. But Lauren King, again, thank you
so much for being a guest today on building brand gravity.
I really always appreciate our conversations.

Speaker 1 (45:39):
Absolutely wow.

Speaker 3 (45:42):
I mean that's really what makes Lauren's perspective just so valuable.
And what I really liked about that discussion and is,
you know, he wasn't just reflecting on the present state
of technology and brand. He was really pushing us to
reimagine the role humanity should play in the continued evolution

(46:03):
of AI and technology. And I guess my takeaway when
I heard that was that if you're a leading brand
or a team right now. My big takeaway from that is,
don't just react to technology design with intention right, use narrative,
use imagination. And I like how we encourage us to
even consider fiction, you know, as a tool to help

(46:25):
you shape a strategy. And maybe most importantly, and I
know you say this all the time, you can put
technology at the center, but you cannot lose that human thread.
So I really like that idea of let's try and
create the world we want to live in powerful stuff.

Speaker 2 (46:40):
I think it's great. And the thing that's wonderful about
Lauren too is his enthusiasm and his deep, deep driving
curiosity when we get on the call. I mean, we
could have chatted about seven billion things aside from our
work as practitioners and our sectors that were specialized in
the technology. But I think it's that But that is

(47:01):
an example of what we all need to bring to
this theme through the enthusiasm, the curiosity, the excitement and
the ability to connect the dots. And also what I love,
I'm seeing this a lot more, not just an our firm,
but you know, as I circulate around a lot of
sense of shared responsibility to be counselors to one another.

(47:23):
So how do we help our peers and colleagues come
up to speed solve for problems that there are going
to always be earlier adopters. There's going to be technology
coming like it always has, and there's people who it
just is like breathing air. It totally makes sense. I mean,
if you remember that story Lauren told he first was
engaging with AI years ago when he was so much

(47:44):
younger as a teenager, and so this has been a
long journey for him despite him being a bit younger
than you and I So I think that idea of
helping your peers, those of us who are natural adopters
on certain aspects of it, looking to say, hey, come
to me with a problem to solve, because this is
a whole new world. And that is what makes me

(48:05):
excited about the pivot to come I think.

Speaker 3 (48:09):
Well, and it also struck me as I was just
hearing you talk right now. You know, you had made
a comment we do every year an employee survey and
a road show and and and made a comment that
she said, you know, we are going to be the
first generation who is not just going to manage humans.
We are going to be we are going to be
managing AI agents, you know, and just kind of that bifurcation,

(48:33):
so as we're trying to create that world we want
to live, and you know, that's that's something that's out there,
and it it kind of struck me, you know, we
always like to end this thinking about what's in our
gravity and this conversation and that just really reminded me
of a of an article I saw on the Wall
Street Journal recently by Isabella Basquette, and the title was

(48:55):
Walmart is preparing to welcome its next customer, the AI
shop agent, And it was just fascinating to think about
that build from what we were talking about, and part
of me, he's going to read a little bit of this,
but said Walmart is preparing for sweeping changes in the
way customers shop, investigating how to make products appealing not

(49:17):
just to human consumers, but also to the AI agents
that will one day shop on their behalf. At some
point in the future, shoppers will deploy an agent and
tell it that they want to restock on groceries or
buy a new flat screen TV. The operator will then
scan the internet and service relevance products based on what

(49:39):
it knows about the user's preference, and ultimately the agents
may even be able to complete the purchase, including payment.
So wow, talk about the marrying of the human and
the AI. Talk about reimagining a world or thinking ahead.
I think that's like a perfect blend to your conversation
with Lauren and just the change we're seeing as we

(50:01):
try and create this new world.

Speaker 2 (50:03):
It's like we're all be celebrities of personal assistance, except
there'll be AI agents. There's a great little clip that's
been going around in social about a month or two ago,
and I think it's real, but you know, you have
to ask yourself questions about it, but it sort of
showed the interaction between like one AI agent trying to
book a hotel and the hotel had its own AI agent,
and when they both realized they answer promorphizing them, They're like, oh,

(50:28):
let's communicate in more computer code, like you know, like bloops,
bloops ones and zeros than human language, and so they
start to I think it might be legit. Others may
have seen it. If it's fake, I totally own that,
but I think it's a real, real metaphor, which is
there's going to be so much of that AI agent
to AI agent interaction and the question of what really

(50:50):
and Lauren said it to and called what really is
agentic AI? Where are agents like? They're going to be coming?
And I think it's going to be both fast and
organic at the same time. It's funny because one of
the things that I was in my gravity when we
were reflecting, you know, building brand gravity, was this point
about two stats I've heard, or two comments i've heard recently.

(51:11):
One was years about the fact that even entry level
I think it was the CEO of jasper dot AI
said this, and a Fortune CEO newsletter that Diane Brady does.
He said, you know, for entry level folks, we're going
to have to be thinking about their management skills because
they'll be managing AI agents as if they're like an
intern or they're a colleague. The other quote i'd heard

(51:32):
is that this generation of CEOs will be the last
to only manage human employees, that there will be agents
as well. And I've always thought of technology as a
strategic asset, but now I have to think about it
almost as workers, and I think that's a big psychological
shift that we need to make. The other more fun
thing I have in my gravity is many people are

(51:54):
listening to this because it's already shooting to the top
of the podcast list. But Amy Poehler's new podc cast
Good Hang is a joy bomb. It's a bomb of joy,
It's fun. Check Black Tina Fey. She just having a
schell Obama on. It's very interesting to see more of
these types of figures enter the podcast game, and as
we have a podcast too in our own little world,

(52:16):
I'm always looking for you know, what are they doing
and how are they doing it? So I definitely recommend
that to others.

Speaker 3 (52:22):
Yeah, I'll have to check that out. That sounds exciting.

Speaker 2 (52:24):
Well, we're coming to the end of another episode. Thank
you for those who listen to both parts of the
AI mini series. As Steve said, if you didn't listen
to the one with Greg, go back and check it out.
And as always, find us wherever you find your podcasts
and on YouTube, and we really thank you for listening.

Speaker 3 (52:40):
Have a great day, tune in soon we'll be back
with more exciting conversations on how we can build brand
graphy