The intersection of artificial intelligence and modern fandom is transforming how we discover, experience, and connect with live events. In this enlightening conversation, Janette Roush, Chief AI Officer and SVP of Innovation at Brand USA, shares a treasure trove of insights on the evolving relationship between AI and the live experience economy.
Most marketers approach AI with a mixture of curiosity and anxiety – an understandable response to technology evolving at breathtaking speed. Roush dispels common misconceptions, explaining that while today's AI tools aren't perfect, we're merely in the "training wheels" phase of what will eventually revolutionize the entire ecosystem of live entertainment.
The discovery process for experiences is already being dramatically reshaped. Unlike traditional search engines prioritizing paid results and keyword-stuffed content, AI-powered discovery offers truly personalized recommendations aligned with individual preferences. This represents a fundamental shift away from volume-based recommendations toward connecting fans with experiences that genuinely resonate with their unique tastes and interests.
The conversation explores essential tools for marketers ready to embrace AI, with recommendations including ChatGPT Team, Claude, Google Notebook LM, Midjourney, and Descript. Looking ahead, Roush envisions a future where personalized AI assistants become our primary technology interface, proactively recommending experiences based on our preferences, schedules, and interests.
Whether you're a venue operator, event producer, or marketing professional, this episode provides critical insights into how AI will reshape fandom while offering practical guidance on navigating this exciting transformation. Ready to explore how artificial intelligence will forever change the relationship between fans and live experiences? This conversation is your essential starting point.
Recorded Tuesday, April 22nd, 2025
Host: Damian Bazadona, CEO & Founder, Situation
Guest: Janette Roush, Chief AI Officer and SVP of Innovation at Brand USA
Producer: Peter Yagecic, Innovation Advisor, Situation
Slides: https://situationlive.com/aipdf
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Peter Yagecic (00:05):
You're listening to Fandom Unpacked the podcast,
an audio version of our regularlive stream series where we
unpack modern fandom with someof the brightest minds in sports
and entertainment.
I'm producer Peter Yagisic andI'm joined by our host for
today's Q&A, Situation's CEO andfounder, Damian Bazadana.
Damian Bazadana Our guest isJeanette Rausch, Chief AI
(00:27):
Officer and SVP of Innovation atBrand USA, the destination
marketing organization for theUS, dedicated to driving
international inbound travel.
Today, we'll be talking allabout the ways generative AI is
impacting live events and fandom.
Here's Damian, with our firstprompt.
Let's roll.
Damian Bazadona (00:40):
You have to curb me and Peter.
Talking about AI for 30 minutesis going to be a challenge.
It's going to be tough, butwe'll do it.
We'll do it for hours.
So AI is at the center of fandom.
As you know, on our Livestreamseries, we spend a lot of time
talking about fandom.
That's the whole thing, and AIis literally at the center of it
, of how it's going to evolve itand change it, and I think it's
really one of the most excitingunlocks I've seen in my career
(01:02):
in the digital space.
That is exciting.
So I know on the call today Isaw the registration list.
We've got sports folks here,theater folks, attractions,
everyone from the liveexperience economy at large.
So it's exciting.
So I want to just dive right in.
Jeanette, you spent a lot oftime with marketers across
hospitality, travel, liveentertainment all these live
experiences.
How would you describe thegeneral mood around the broader
(01:27):
topic of AI?
Because you know, I'm assuming,if you, for the people on this
live stream right now, if youhad to guesstimate how they're
feeling right now when they'rethinking about AI, how would you
describe that?
Janette Roush (01:37):
I'm going to guess that it's a combination of
curiosity overwhelm, somenervousness.
I'll say probably 20% of theaudience is currently using paid
tools of some kind so that yourdata stays a little more secure
and you have access, kind of,to better capabilities.
And then I think the other 80%most of you are probably using
(02:01):
free tools and playing aroundwith it every once in a while
and maybe kind of asking well,what's the big deal, what's the
point?
Or this is the thing that'sgoing to steal my job and you're
not quite sure what to do withit or what the next step is.
Damian Bazadona (02:16):
From where you're at, it's funny, I think,
of curiosity, anxiety.
Curiosity and anxiety are likemy eating habits at night too,
I'm like shouldn't do this, butit looks so good, and I've seen
you speak at a lot of differentevents and what would you say is
the biggest misconception?
You're hearing, you're like inlarge rooms, all these types of
marketers, everyone's asking youa question, given that's a
(02:37):
significant part of your role,what do you feel like is the
biggest misconception thatpeople are getting on it right
now?
And, by the way, let me justqualify really quickly.
I'm using the broadest term ofAI.
It's like air at this point.
It means everything.
It's been around everywhere.
We're talking a lot, I think, alittle bit more in the genitor
space, the idea we're going tobe leaning a little bit into
chat, gbt, that ecosystem, justto qualify a little bit, because
(02:57):
it's a massive topic.
Janette Roush (02:59):
But talk about the misconceptions you're
hearing that it can either solveall of your problems or that it
cannot measurably help yourproblems in any way.
So it's kind of living too muchat the extremes.
There is a certain fallacy ofworking with AI that everything
that it gives you is going to beperfect and amazing and almost
(03:22):
a mandate.
You have to use this becauselook at how much language you
know.
This was trained on.
Like, how much smarter is thisthan me?
What do I know?
Whatever it said, let's do, orthe total opposite, which, and
I'd say, neither version ishelpful.
Damian Bazadona (03:39):
Yeah, I feel like particularly in
misconception.
It gets a lot wrong.
And I'm going to a lot of thecontext.
What I'm speaking is I'm sortof in the chat gpt lane, uh, for
a whole range of reasons, butit's probably the most
frictionless to sort of jumpinto and I now use it as an
unlock for pretty mucheverything.
But it's fascinating when Ispeak with other people about it
.
I see those two camps often and, um, it gets a lot wrong.
(04:00):
It's we're like on the trainingwheels right now, like Like
we're just getting.
You can see the opportunity toit.
How do you envision AIreshaping the way fans discover
live experiences and thinkinglike travel planning or finding
that next concert?
And when I'm poking at, thekeyword in that sentence is
discover, because we're going toget into other stuff later.
But the discovery is rapidlychanging with the relationship
(04:23):
of search and how that's looking.
What's your take?
Janette Roush (04:27):
You know somebody I was running a workshop with a
friend of mine at the Oregon'sGovernor's Conference on Tourism
last week gone to Medium andwrote this ode to a toothbrush
that they had found throughChatGPT.
That is like this small Germanbrand and I guess they had put
(04:52):
in a chat query something thatwas you know, not just what's
the best toothbrush in the world, but also with some of their
own personal specifications forwhat they were looking for.
And it came back and said thisthis is the toothbrush that you
want.
And they're like well, that'swho am I to argue that I'll give
that a try?
And then they were like thiswas the best toothbrush I have
(05:13):
ever seen in my entire life.
This is changing toothbrushingfor me.
And then my friend discoveredthat article through ChatGPT.
So it kind of became thisvirtuous discovery cycle and I
think what it shows us is it's alittle unpredictable what these
large language models?
What?
Because it's not arecommendation engine, it's a
(05:36):
next word predictor, and so Ihave heard from people they're
really nervous that it's goingto Latin.
All of the recommendations thatcome out of AI that we're all
going to be sent to the same top10 restaurants near me that
would show up in a Google searchresult, and I think what that
(05:57):
misses is this personalizationaspect that our own AI is going
to add in.
So where I see all of this goingis and I you know there's going
to be a number of differentpaths for how people use AI but
I think the primary way is thatwe're going to have on our
phones the one chatbot.
(06:17):
That's kind of our main source.
So if it's chat GPT, then it'schat GPT and that it will.
Chatgpt just turned onpersistent memory a week ago
across all of your chats, so nowit knows everything about you
that you've ever had aconversation with it about.
(06:39):
More targeted to me, jeanetteRauch, than just here is, you
know, the top ranking toothbrushon Amazon, which you know.
Great, that's volume.
But do I want to make everydecision based on volume of
other people purchasing?
Not necessarily, not as a fan,and so figuring out and kind of,
(07:02):
and so figuring out and kind ofunlocking.
Damian Bazadona (07:20):
How do we get our productions, our sports
teams, our live events tosurface in these large language
models?
I think that's like the bighill for all of us to climb
right now, so let's dive into it, because that's, I think, the
huge question In the marketingspace.
I would imagine a lot of peopleon this call.
A significant amount of theirmoney probably goes to well,
definitely goes to some Googleproduct at large and or a meta
product, but let's just usesearch.
That's like the gateway in, andwe all know how to play that
game.
It's a paid model.
Yes, there's some organic playand there's different placements
(07:40):
.
You can play in math, but,generally speaking, and so you
can now look.
Well, at least you used to beable to fully understand that.
Google now is clearly, I thinkit's in a clear identity crisis,
trying to identify what is itgoing to be with this Google AI
overview.
How do you think, though,brands should be preparing for
it?
Like, what are the steps Againin the past?
Well, I'll pause there, becauseI'll just keep asking questions
(08:02):
.
What do you think they shouldbe?
How do they should be preparingfor discovery?
What would you be?
What do you think?
What are you recommendingpeople should?
Janette Roush (08:08):
do.
The best advice that I haveheard is just do what you ought
to be doing anyway and creatingthe good quality content that
through your research andthrough all of the work that
everyone does to determinewhat's going to be the most
moving.
To do it, because right now welive in this weirdly gamified
(08:33):
world of content where, you know, for Google search, everything
is written with keywords in mindand everything you know we have
, all we've done this toourselves, because the internet
is full of recipe websites thatyou have to scroll through a
life story to get to the recipe.
And that has happened becauseGoogle rewards those websites
(08:54):
because they've chucked in morekeywords.
But it's not the actual bestanswer to the question that
somebody might be trying to askyou know, how do I cook this
particular kind of food?
So what large language modelsif they continue to center, like
the person and the experienceof the person using it?
(09:17):
You know, and that's probablyan open question that might not
always be true, but let's assumethat's pretty true right now
then it's not going to force meto read all of the junk before
the recipe.
It's just going to give me therecipe as succinctly as possible
and so, like that becomes theopportunity is don't try to
(09:40):
trick or confuse the largelanguage models by, I don't know
, having content on your websitein every single possible
language, because large languagemodels like to go to
in-language examples before theygo to other language examples,
depending on what you speak whenyou're typing into the engine
and like you can waste a lot oftime gamifying, or you can just
(10:05):
do the right thing create thecontent that users need, answer
the questions that they need toknow before they purchase the
ticket.
Damian Bazadona (10:14):
Yeah, no, I think we're about to see the
biggest shift I think I've seenin a long time, as it relates to
consumers connecting with thethings that they want.
And there's Google, and I loveGoogle as a company and I think
they're going to figure it out,it's no question.
But the current search ontravel sucks.
It just sucks.
I'm sorry what it does.
It gives you every choice, andevery choice that's paid for,
(10:37):
and you're kind of going am Igetting anything of value here?
Until you start digging in?
And there's too much work.
Until you start digging in, andthere's too much work.
And again, a long hill to go,even on the ChatGPT side.
But I don't know from personalexperience.
I just find it a lot better,way better experience.
Let's talk about how AI willactually change the actual
(10:57):
experience for fans.
Let's think pre, during, post.
I'm a Knicks fan.
That was not a great game lastnight.
Peter Yagecic (11:04):
I'm so sorry, but that was not a great game last
night.
Damian Bazadona (11:05):
But like, how might AI help me, whether that's
wayfinding, gamifying, likewhat are you seeing, what are
you thinking, even in sports ortravel or any other just
vertical, just curious how youthink it's going to impact the
actual experience itself.
Janette Roush (11:18):
Wayfinding.
I think there are hugeopportunities there Wayfinding
and translation.
So again, that personal devicethat lives on your phone.
Imagine you're a wheelchairuser.
My device, my Google account,would already know all of the
dimensions of my uniquewheelchair, because this is an
(11:40):
issue across travel right nowthat there's a number of
companies working to solve insome way, because a hotel might
say the room is accessible andthat is a absolutely meaningless
adjective.
I don't know if my you know ifit's a roll-in shower or, you
know, do I need to have twopeople with me to help transfer
(12:00):
to the bed or is that equipmentalready there?
And so for a future version ofGoogle Maps to understand where
is my hotel and what is the pathto the theater or to the arena,
that includes curb cuts atevery curb so that my wheelchair
can make that trip.
I think that's like the level ofaccessibility help that we can
(12:24):
expect in the future.
You know, we already seeamazing things for people who
are blind or have low vision andthe ability of you know.
Imagine a pair of meta glassesthat can do audio descriptions
for the user as they're outnavigating the real world.
So I think some of the greatestopportunities are in that arena
(12:47):
.
And then I think it comes down,like for the fan experience
I've kind of been thinking about, like what would it be like to
be able to experience the gamefrom the perspectives of
actually the players on thecourt?
Damian Bazadona (13:02):
Yeah.
Janette Roush (13:03):
And so, like that's, I'm cheating now and
like moving ahead of like what'sthe ultimate dream 10 years
from now?
But imagine being able to see amusical, but now you're seeing
it through the eyes of Elphabainstead of through the eyes of
the audience member.
Like experiences that wouldnever be available to us.
I think that's the kind ofthing AI can make real us.
Damian Bazadona (13:28):
I think that's the kind of thing AI can make
real.
Yeah, you know, just buildingon.
This is really not necessarilyabout the experience itself, but
almost the transaction processof being able to map out where I
want to sit.
I will tell you, if I I am anaisle snob, I will pay a premium
for an aisle.
I will sit on the roof ifthere's an aisle seat, and so
any.
It does not take a genius to beable to watch me for like 10
minutes and go oh, this guywants an ILC, and so pretty much
no matter what and I knowthere's a lot more automation
(13:49):
happening on that front of botchbeing able to shop within as
agents, excuse me, which wouldbe incredible, which I know,
peter, you've been testing with.
I just think there's obviouslya massive opportunity on that
front.
Speaking of Peter, peter, arethere any questions coming?
I just want to make sure I'mbeing mindful of any questions
that might have come in.
Peter Yagecic (14:09):
Yeah, we do have some audience questions coming
in.
Going back to your toothbrushexample, jeanette, deep research
gen AI seems like a much bettercomparison shopper than earlier
LLM models.
Just ask deep research to helpfind the cheapest tickets for
any given Broadway show thisweekend and see what you get.
I think the question is a yearfrom now, will any humans be
doing this research or will itall be our personal AI agents?
(14:32):
But maybe could you take aminute to kind of let the
audience who may not be familiarwith deep research we were
talking about it right before wewent on camera the difference
of kind of what deep research isdoing, and then you know will
it take over all of the researchthat we're doing?
Janette Roush (14:54):
So deep research, and this is annoying because
the people at these companiesare so terrible at naming their
products.
And so Google Gemini came outwith a deep research back in
December, which then OpenAI cameout with their own version of
deep research a month later.
I think Perplexity now offersdeep research.
I think Claude has rolled itout, but they've called it
something slightly different.
(15:15):
But what the technology isdoing is essentially I think of
it as going down an internetrabbit hole.
So all of these services thatall offer some version of this
type of work, it's you lay out aproblem for it and then it may
ask a couple of clarifyingquestions or let you see its
(15:37):
research plan, and then thelanguage model goes off and
spends up to 30 minutesresearching on the Internet to
see what are the best ways toanswer this challenge.
You have given it, and it isgreat for comparison shopping on
products.
It is really good if you'redoing a competitive analysis.
(15:57):
So you know, brand USA, we'relooking to move more into the
business event market, so that'sgetting more conventions to be
held in the United States andfor you know and this will be a
new piece of work for us Like,we're not going to have sales
(16:29):
teams that are literally bookingthe convention centers, because
that happens at the local level.
So what is our role?
And deep research does a greatjob of taking kind of a big,
needy, diffuse project like thatand coming up with a
McKinsey-style 50-page reportthat is based on 30 minutes of
(16:52):
research.
It is insanely valuable.
Peter Yagecic (16:55):
More are coming in if we want to do another
audience question.
Damian Bazadona (16:57):
Do it, do it, do it, do it.
I've got 400, but you go.
Peter Yagecic (17:00):
Starts out?
Security, security, security.
Should we operate under theassumption that anything we put
into chat GPT is not private andcould show up anywhere?
Is there a way to protectourselves?
Which I think the answer tothat is yes, but, Jeanette, I'll
let you take a whack at that.
Janette Roush (17:13):
Yes.
So I have done a lot ofstudying on AI governance and
kind of what are like how do weresponsibly work with AI in our
companies?
And so, like I'll say for mepersonally, I am a privacy
fatalist.
I have no assumption thatanything I ever do or say is
private anymore.
So I'll put my blood work inthe chat GPT so that I know what
(17:37):
my doctor is going to saybefore she says it.
But that is not how I wouldapproach privacy and security.
You know, on behalf of ourbusinesses or on behalf of Brand
USA, and so for both of those,the first step is use a paid
model, because that gives youthe opportunity to turn off a
toggle that says that thelanguage model can use anything
(18:02):
that you put into it as trainingfor future language models.
So that's the first step.
It's going to in most cases, ifyou're using a chat GPT team
account or an enterprise account, that's going to give you SOC 2
security.
So again, it keeps your datasafe from hackers when it is
moving to the cloud where it'sgoing to be analyzed, and then
(18:24):
while it is at rest in the cloud.
But then there's separatequestions about what is okay to
put in a large language model.
That isn't necessarily relatedto security.
So I'd say yes, I would feelpretty comfortable knowing that
I have a SOC 2 compliant closedsystem putting most items into a
(18:47):
language model.
I'm not going to put the schemafor a new iPhone into it if I
work at Apple and I'm usingChatGPT, which is heavily
invested in by Microsoft, butthey're not doing that anyway
and most of the stuff we'redoing is not going to be at the
level of state security but interms of what else is not okay
(19:10):
to put into a language model PII.
So you know personallyidentifiable information.
It is a violation of GDPR toput any you know names and phone
numbers, email addresses of anEU citizen into a tool like
ChatGPT, because the EU says youhave to give like affirmative
(19:32):
consent in order for somebody touse your information in a
language model.
So that's just a best practiceto not do that, because it's not
your information to give away.
And then you want to look atthe third-party permissions of
things like research documents.
So one of my favorite use casesfor AI personally is to create
(19:54):
a project inside of ChatGPT orinside of Clod.
It looks like little foldersand you can put background
information into those littlefolders but some background
information, like a researchreport from a third party.
If it's syndicated research,they may say in your contract
that you're not allowed to sharethis information with a third
(20:16):
party.
A language model is legallyconsidered a third party.
So it's time for all of us.
You know, as small businessowners I bet on this phone or
people who work for really,really big brands you want to
make sure that you are talkingto your vendors.
Like, rather than ooh I don'tknow if it's okay to do that
(20:37):
Like, let's get it out of.
You know the secret realm.
It's okay to say I'd like toput this stuff into a large
language model that has trainingturned off.
Can we add those terms to ourcontract?
Peter Yagecic (20:52):
Fandom Unpacked is brought to you by Situation,
an award-winning marketingagency built for live
entertainment that champions thepower of unforgettable shared
experiences around the world.
We offer full marketing andcreative services for
experience-based brands in liveentertainment attractions,
theater sports, arts and cultureand more.
Damian Bazadona (21:20):
Check us out at situationinteractivecom.
Now back to our Q&A.
There's one of the common.
I'm going to switch gears alittle bit.
I want to talk a little bitabout authenticity in terms of,
and content development, and soI feel like one of the more
common questions I've seen andI'd love to get your take on
this is how does AI run the riskof making content feel less
authentic?
And what I've heard from contentdevelopers oftentimes is like
hey, look, I use it for a lot ofstuff, but my voice I'm so
niche that it would never reallycapture what I do.
(21:44):
Now I'm like you're wrong,because if you train it enough,
it's going most likely to be youif you want it to be Now.
I personally don't want to livein a world like that.
So there's something about youhave your own personal judgment
of things you want to hold on to.
But it's interesting I hearthis a lot that the belief of
like oh, it'll never get me, Ithink is missing the point of
the power of what you're dealingwith.
(22:05):
What's your take?
Because this is a big one.
A lot of people on this webinarright now, on the live stream
these are content creators, andso what's your take?
Janette Roush (22:18):
I'm just curious on AI and development.
There's so many pieces to it,right?
Because part of it I'm like.
I think, yes, not today.
Can it perfectly replicate ourown unique voices?
I don't think that's very faraway and I think, from an
artistic perspective, totallynot there today.
I think there's a lot of funstuff you can do with Midjourney
(22:41):
or the new image tool inside ofChatGPT.
It's not human-replacing today,but but I think it could get
much closer pretty soon.
And that you know, in my worldthat creates a lot of
consternation because, you know,destination marketing
organizations are promotingdestinations and there is like a
(23:04):
big push to make sure thateverything we do to promote the
destinations is authentic.
We wouldn't want a, you know,phony version of the Empire
State Building to be used, evenif we knew in every single way
it was perfect.
That's also today.
And so if you think 20 yearsago when Photoshop was new and
(23:27):
everybody's like, oh God, I wantto on the cover of our brochure
or for this trade show booth, Iwant to make sure that the blue
sky behind the city skylinereally pops, and there's a
little bit of hand-wringingabout how much Photoshop is okay
and how much is too much.
But we don't have thatconversation anymore and I'd
argue people completely expect,if they see a sky in the photo,
(23:51):
that somebody touched that skyup and if the sky were kind of
hazy and dreary that peoplewould kind of be upset that you
didn't do that work.
So what we consider to be ournorms are going to subtly shift
over time.
Damian Bazadona (24:07):
If you think about it's almost like how do
you measure authenticity?
That's like a big one, right?
Because it's like how, like,how, like I don't know.
You know, I would imagine someof the pushback because, I agree
, I I'm hard, how far technologyhas come, how fast.
And yeah, you can like, rightnow you can tell when
something's a little bit off,you're a little bit off here.
But if you call this trainingwheels and go 20 years from now,
oh, the technology itself isgoing to figure it out.
(24:30):
The question, I suppose, reallyis it's consumer push, it's
consumer pushback, right, if Ifound out that I was talking to
someone who I think I'm amachine, I think I'm talking to
a human, so that really goes tolike, what's the role of brands
and being transparent of this isjust, you're dealing with my AI
self, but also the role ofgovernment and compliance, which
I think is that's.
The government is clearly goingto step in at some point and
(24:55):
try and put guardrails on this,because the knock-on effects of
this are going to be incredibleif this just remains unleashed
on our economy, on our society.
This is a hard, this is atricky question.
The crystal ball of like, whatis government?
What is compliance orgovernment?
What do you feel like?
Where do you see that maybegoing?
Janette Roush (25:10):
A lot of the compliance stuff, like I don't
know how much the US governmentis going to step in from a
regulation standpoint.
Certainly other governmentswill.
So as long as we're selling ourproducts to European citizens,
the EU AI Act will dictate a lotof kind of the bigger things
that all of us can use AI to do.
(25:31):
But ultimately, I think it doescome down to the brand level to
be very clear about how do youuse AI and where are your
company's personal limits,because nobody is going to tell
us what to do.
There is no real right or wrong.
It's just what do wecollectively agree to do
(25:54):
together?
And so for a brand that canlook like AI guidelines.
So that's something you know werolled out when I was pretty
early on here at Brand USA andit's something that I would
suggest every organization needs, because it's the opportunity
for you to come together and sayhow comfortable is your
(26:14):
organization with pushing thelimits when it comes to AI use
and maybe collectively agree.
Or leadership comes together andagrees oh, we're gung-ho on
this and we want to provide thetools for everybody to use it
safely.
Or you may say no, we want tokeep AI out of the workplace.
I'd suggest you're at risk forshadow AI or BYO AI if you did
(26:39):
that, but that could be aposition to take, and then it
allows you to say how muchalteration are we okay with
before we want to tell people,you know, with a note on our
website or a footnote to apicture, to some way be clear,
because I think that is like.
The number one reputationalrisk is that people don't want
(27:02):
to be tricked, and so it's notthat using AI has a moral
implication, but you want tomake sure that people understand
what the AI did and what thehuman did.
Damian Bazadona (27:16):
Yeah, well, let me okay.
So let me switch gears a littlebit.
Talk about I'd love for you totalk about your must-do list for
marketers, right?
I think it'd be helpful for thefolks on here to see what
you're playing with and sometools.
Janette Roush (27:28):
Great.
So for me, chatgpt Team is kindof the go-to account.
If you're only going to do onething for your team, get ChatGPT
Team.
It's going to be $25 a user amonth if you purchase monthly, I
believe, or, yeah, if youpurchase annually, that's the
monthly rate and your ChatGPTcomes out with the newest
(27:50):
additions, like the bestfeatures, faster.
I don't think it out with thenewest additions like the best
features, faster.
I don't think it's always thegreatest model.
I'd say today it's probably thegreatest model, but then Claude
will leapfrog over it.
So if you're looking for justone answer, looking at this
entire list, it's chat GPT, apaid version.
(28:11):
But then on top of that Ireally like Claude, which also
offers kind of a safer teamaccount because it just has a
better writing style.
And if you're somebody who getsinto a long conversation with
something like ChatGPT, you'llfind that ChatGPT will start
spitting out the same ideas overand over and it kind of just
(28:34):
gets stuck in a rut.
Cloud is much less likely to dothat in my experience.
And then if you look at theGoogle suite, google's you know
their new model that just cameout is currently top of the
leaderboards.
That is not typically the casefor Google, so it would be lower
on my list to buy unless Ialready had Google Workspace.
(28:54):
But I will say some of theother products that I have on
there, like Google Notebook LMterrific research tool.
For that one you just uploadPDFs or other files and then it
will answer your questions onlylooking at your files and it
will open up your files and itwill highlight where it found
the answer to your question, sothat it's a way to work through
(29:18):
hallucinations and be reallygrounded in truth.
And then, just like picking afew from the other categories in
the image and voice tools,midjourney takes some training
to understand how to createimages with it, but it's well
worth it.
Training to understand how tocreate images with it, but it's
well worth it.
Descript is a fantastic tool forediting short videos or audio
(29:38):
content.
So you could take this podcast,feed it into Descript and then
they have what they call an AIunderlord.
So not your overlord but yourunderlord, and it will find all
of you're like I'm looking forfive 30-second segments and
we'll find the five mostinteresting segments from the
podcast.
It will make it vertical orhorizontal.
It'll put whatever style ofcaptions you want on it.
(30:01):
You can do more editing tricksif you're not.
You know I'm not an editor, butif you already know how to do
that, it's easy to do it withinthe tool and even so, somebody
like me I could fake somebodyinto thinking I was a podcast
producer by just having this one$20 a month subscription.
I also under the everythingelse bucket.
(30:22):
Beautifulai is a terrificpresentation tool and I think
it's worth trying out.
Napkinai, if you are evermaking infographics for an ad
meeting.
Damian Bazadona (30:32):
Great See, I love it.
I'd sometimes I know a lot ofpeople.
I asked Well, what actual toshould I tool should I use?
I mean, there's a lot ofdifferent and I know a lot of
people on cloud, by the way, andI've tested notebook LM to me
as one of my favorite.
It was one of those things whenyou, when you see, I put my own
medical stuff to it and it gaveme better.
It gave me, but with betterbedside manner.
Peter, I know there's more.
(30:54):
I want to be respectful of allthe questions we have.
Peter Yagecic (31:01):
Sure, yeah Well, I think I know the answer to
this one, but as it relates toresearch, how can we specify
where we want AI to pull answersfrom to avoid research
information bias?
You mentioned that notebookwill just scam what you give to
it.
But is it as easy as saying?
Because most of these areconversational interfaces, you
can just say that to them.
Janette Roush (31:16):
Sometimes I mean, if you're depending on what the
research question is, sometimesit will go, it will lead you
astray, and I think the problemwith asking it is that it wants
to make you happy and so it willlie.
So there is no asking AI tofact-check or proofread itself.
Peter Yagecic (31:37):
We do use AI tools not Descript, which I'm
very interested in, but we useit.
We feed all the transcriptsfrom this very series into it,
asking for great quotes, and ithas hallucinated those quotes.
It's come up with great stuff,but stuff that nobody actually
said.
So yeah, take it with a grainof salt.
Janette Roush (31:53):
Yeah, that's why the notebook tool is so great,
because it will pull.
You will look at the actualquote in real time.
Peter Yagecic (32:00):
Yeah, and you know, I think, as Damien
mentioned, this is just thebeginning of probably a
continuing conversation.
One question that I would wantto ask on behalf of our audience
is how can people best stay inyour orbit, jeanette?
I mean, you are staying on theforefront of this, it's in your
job title.
What's the best way for peopleto continue to make sure that
(32:22):
they get continued wisdom fromyou in this topic?
Janette Roush (32:25):
I write a lot on LinkedIn about AI actually, so
LinkedIn is the best spot tolearn more from me.
So LinkedIn is the best spot tolearn more from me.
And then I follow so manydifferent newsletters that I had
to open up a separate emailinbox to track all of them, so
it doesn't just overwhelm mywork inbox.
Damian Bazadona (32:44):
Listen to podcasts I am heavily consuming
this information so that youdon't have to work quite so hard
at it.
That's how I saw I mean, I'veknown Jeff for years, but that's
what kind of triggered thisconversation her LinkedIn feed
of watching all the stuff thatshe's doing, which I just think
is just awesome.
Peter, can I ask the lastquestion?
I just want to ask Go for it Onthe horizon that you're excited
(33:07):
about.
And obviously there's a millionthings on the horizon in the AI
space.
It's like sky's the limit.
But, like you know, I'd love toask kind of, what's on the
horizon that you're excitedabout that?
Or you know, if you have adream up, an AI experience in
your head of like, oh, here'swhat I can see happening.
Like I'd love to talk a littlebit more about the future.
Janette Roush (33:26):
I really see the idea of our one personal AI
device being the way that weinteract with AI in the future,
and so I think, from a liveevent perspective, what that
means like, let's say, mypersonal AI ends up being
Google's Gemini.
Gemini, like Google, alreadyreads all of my emails and it
(33:48):
has, you know, parts of mycalendar, and you know so much
of our personal lives is alreadyreally deeply known by Google.
So if that were to become mypersonal AI, it would know when
do I have free time coming up inAugust, and so does my husband
and our stepkids, and it's beena long time since we've gone to
(34:11):
see a Broadway show been a longtime since we've gone, you know,
to see a Broadway show, andthey know that there was some
chat message that the kids hadthat I didn't see, that talked
about some or other show, and atthe point that I am ready where
I would say, oh, I need tostart planning a family activity
, gemini will open up on myphone a I don't know if I would
(34:33):
call it a website, but it won'tnecessarily have to look like a
text chat interface.
I think, based on what we aredoing, it will create an
interface on the fly thatmatches the activity that we're
talking about and that I woulduse that interface to choose a
(34:57):
vacation destination or to picka Broadway show.
I think, if it were somethinglike picking a Broadway show,
that if I am a short-form videoperson, that it would create
three 15-second video clips fromthree different Broadway shows
to help give me a taste of whatit is, from three different
Broadway shows, to help give mea taste of what it is, and it
may steal those images you knowfrom YouTube.
Or maybe we keep on ourwebsites for our products like a
(35:23):
library, like a capsule ofhere's all the video and image
content that you can pull in toyour LLM to create these spun-up
websites.
Here's all the descriptionsthat you would need to know, and
it wouldn't be something aperson might want to read.
But let's say, the websiteincludes here's all the target
(35:44):
audience profile segments whoare currently buying tickets to
this show, and then that's a waythat the data that we already
have can help us find morepeople who fit into those you
know audience target segments,without them needing to raise
their hand first, becausethey're just doing it
(36:06):
intuitively through using theirdevice.
Damian Bazadona (36:09):
Yeah, it's exciting.
I think that it's just anextremely exciting time to be in
the marketing space for liveexperiences and the whole
intersection of phantom.
So I think to me I'm like it'sscary to some, exciting to
others.
The reality is it's a massivechange and we haven't even
started seeing the impact of itand the knock-on effects.
And with change is opportunity,and so I I've never been more
(36:32):
excited to be in this space.
I think it's actually very cool.
Peter Yagecic (36:36):
That's going to do it for this episode of Fandom
Unpacked the podcast.
If you liked what you heard,please be sure to leave us a
review on Apple Podcasts.
Find out how to join us livefor an upcoming recording at
SituationLivecom slash fan.
We'll see you next time, truebelievers.
You're listening to Fandom Unpacked the podcast,
an audio version of our regularlive stream series where we
unpack modern fandom with someof the brightest minds in sports
and entertainment.
I'm producer Peter Yagisic andI'm joined by our host for
today's Q&A, Situation's CEO andfounder, Damian Bazadana.
Damian Bazadana Our guest isJeanette Rausch, Chief AI
(00:27):
Officer and SVP of Innovation atBrand USA, the destination
marketing organization for theUS, dedicated to driving
international inbound travel.
Today, we'll be talking allabout the ways generative AI is
impacting live events and fandom.
Here's Damian, with our firstprompt.
Let's roll.
Damian Bazadona (00:40):
You have to curb me and Peter.
Talking about AI for 30 minutesis going to be a challenge.
It's going to be tough, butwe'll do it.
We'll do it for hours.
So AI is at the center of fandom.
As you know, on our Livestreamseries, we spend a lot of time
talking about fandom.
That's the whole thing, and AIis literally at the center of it
, of how it's going to evolve itand change it, and I think it's
really one of the most excitingunlocks I've seen in my career
(01:02):
in the digital space.
That is exciting.
So I know on the call today Isaw the registration list.
We've got sports folks here,theater folks, attractions,
everyone from the liveexperience economy at large.
So it's exciting.
So I want to just dive right in.
Jeanette, you spent a lot oftime with marketers across
hospitality, travel, liveentertainment all these live
experiences.
How would you describe thegeneral mood around the broader
(01:27):
topic of AI?
Because you know, I'm assuming,if you, for the people on this
live stream right now, if youhad to guesstimate how they're
feeling right now when they'rethinking about AI, how would you
describe that?
Janette Roush (01:37):
I'm going to guess that it's a combination of
curiosity overwhelm, somenervousness.
I'll say probably 20% of theaudience is currently using paid
tools of some kind so that yourdata stays a little more secure
and you have access, kind of,to better capabilities.
And then I think the other 80%most of you are probably using
(02:01):
free tools and playing aroundwith it every once in a while
and maybe kind of asking well,what's the big deal, what's the
point?
Or this is the thing that'sgoing to steal my job and you're
not quite sure what to do withit or what the next step is.
Damian Bazadona (02:16):
From where you're at, it's funny, I think,
of curiosity, anxiety.
Curiosity and anxiety are likemy eating habits at night too,
I'm like shouldn't do this, butit looks so good, and I've seen
you speak at a lot of differentevents and what would you say is
the biggest misconception?
You're hearing, you're like inlarge rooms, all these types of
marketers, everyone's asking youa question, given that's a
(02:37):
significant part of your role,what do you feel like is the
biggest misconception thatpeople are getting on it right
now?
And, by the way, let me justqualify really quickly.
I'm using the broadest term ofAI.
It's like air at this point.
It means everything.
It's been around everywhere.
We're talking a lot, I think, alittle bit more in the genitor
space, the idea we're going tobe leaning a little bit into
chat, gbt, that ecosystem, justto qualify a little bit, because
(02:57):
it's a massive topic.
Janette Roush (02:59):
But talk about the misconceptions you're
hearing that it can either solveall of your problems or that it
cannot measurably help yourproblems in any way.
So it's kind of living too muchat the extremes.
There is a certain fallacy ofworking with AI that everything
that it gives you is going to beperfect and amazing and almost
(03:22):
a mandate.
You have to use this becauselook at how much language you
know.
This was trained on.
Like, how much smarter is thisthan me?
What do I know?
Whatever it said, let's do, orthe total opposite, which, and
I'd say, neither version ishelpful.
Damian Bazadona (03:39):
Yeah, I feel like particularly in
misconception.
It gets a lot wrong.
And I'm going to a lot of thecontext.
What I'm speaking is I'm sortof in the chat gpt lane, uh, for
a whole range of reasons, butit's probably the most
frictionless to sort of jumpinto and I now use it as an
unlock for pretty mucheverything.
But it's fascinating when Ispeak with other people about it
.
I see those two camps often and, um, it gets a lot wrong.
(04:00):
It's we're like on the trainingwheels right now, like Like
we're just getting.
You can see the opportunity toit.
How do you envision AIreshaping the way fans discover
live experiences and thinkinglike travel planning or finding
that next concert?
And when I'm poking at, thekeyword in that sentence is
discover, because we're going toget into other stuff later.
But the discovery is rapidlychanging with the relationship
(04:23):
of search and how that's looking.
What's your take?
Janette Roush (04:27):
You know somebody I was running a workshop with a
friend of mine at the Oregon'sGovernor's Conference on Tourism
last week gone to Medium andwrote this ode to a toothbrush
that they had found throughChatGPT.
That is like this small Germanbrand and I guess they had put
(04:52):
in a chat query something thatwas you know, not just what's
the best toothbrush in the world, but also with some of their
own personal specifications forwhat they were looking for.
And it came back and said thisthis is the toothbrush that you
want.
And they're like well, that'swho am I to argue that I'll give
that a try?
And then they were like thiswas the best toothbrush I have
(05:13):
ever seen in my entire life.
This is changing toothbrushingfor me.
And then my friend discoveredthat article through ChatGPT.
So it kind of became thisvirtuous discovery cycle and I
think what it shows us is it's alittle unpredictable what these
large language models?
What?
Because it's not arecommendation engine, it's a
(05:36):
next word predictor, and so Ihave heard from people they're
really nervous that it's goingto Latin.
All of the recommendations thatcome out of AI that we're all
going to be sent to the same top10 restaurants near me that
would show up in a Google searchresult, and I think what that
(05:57):
misses is this personalizationaspect that our own AI is going
to add in.
So where I see all of this goingis and I you know there's going
to be a number of differentpaths for how people use AI but
I think the primary way is thatwe're going to have on our
phones the one chatbot.
(06:17):
That's kind of our main source.
So if it's chat GPT, then it'schat GPT and that it will.
Chatgpt just turned onpersistent memory a week ago
across all of your chats, so nowit knows everything about you
that you've ever had aconversation with it about.
(06:39):
More targeted to me, jeanetteRauch, than just here is, you
know, the top ranking toothbrushon Amazon, which you know.
Great, that's volume.
But do I want to make everydecision based on volume of
other people purchasing?
Not necessarily, not as a fan,and so figuring out and kind of,
(07:02):
and so figuring out and kind ofunlocking.
Damian Bazadona (07:20):
How do we get our productions, our sports
teams, our live events tosurface in these large language
models?
I think that's like the bighill for all of us to climb
right now, so let's dive into it, because that's, I think, the
huge question In the marketingspace.
I would imagine a lot of peopleon this call.
A significant amount of theirmoney probably goes to well,
definitely goes to some Googleproduct at large and or a meta
product, but let's just usesearch.
That's like the gateway in, andwe all know how to play that
game.
It's a paid model.
Yes, there's some organic playand there's different placements
(07:40):
.
You can play in math, but,generally speaking, and so you
can now look.
Well, at least you used to beable to fully understand that.
Google now is clearly, I thinkit's in a clear identity crisis,
trying to identify what is itgoing to be with this Google AI
overview.
How do you think, though,brands should be preparing for
it?
Like, what are the steps Againin the past?
Well, I'll pause there, becauseI'll just keep asking questions
(08:02):
.
What do you think they shouldbe?
How do they should be preparingfor discovery?
What would you be?
What do you think?
What are you recommendingpeople should?
Janette Roush (08:08):
do.
The best advice that I haveheard is just do what you ought
to be doing anyway and creatingthe good quality content that
through your research andthrough all of the work that
everyone does to determinewhat's going to be the most
moving.
To do it, because right now welive in this weirdly gamified
(08:33):
world of content where, you know, for Google search, everything
is written with keywords in mindand everything you know we have
, all we've done this toourselves, because the internet
is full of recipe websites thatyou have to scroll through a
life story to get to the recipe.
And that has happened becauseGoogle rewards those websites
(08:54):
because they've chucked in morekeywords.
But it's not the actual bestanswer to the question that
somebody might be trying to askyou know, how do I cook this
particular kind of food?
So what large language modelsif they continue to center, like
the person and the experienceof the person using it?
(09:17):
You know, and that's probablyan open question that might not
always be true, but let's assumethat's pretty true right now
then it's not going to force meto read all of the junk before
the recipe.
It's just going to give me therecipe as succinctly as possible
and so, like that becomes theopportunity is don't try to
(09:40):
trick or confuse the largelanguage models by, I don't know
, having content on your websitein every single possible
language, because large languagemodels like to go to
in-language examples before theygo to other language examples,
depending on what you speak whenyou're typing into the engine
and like you can waste a lot oftime gamifying, or you can just
(10:05):
do the right thing create thecontent that users need, answer
the questions that they need toknow before they purchase the
ticket.
Damian Bazadona (10:14):
Yeah, no, I think we're about to see the
biggest shift I think I've seenin a long time, as it relates to
consumers connecting with thethings that they want.
And there's Google, and I loveGoogle as a company and I think
they're going to figure it out,it's no question.
But the current search ontravel sucks.
It just sucks.
I'm sorry what it does.
It gives you every choice, andevery choice that's paid for,
(10:37):
and you're kind of going am Igetting anything of value here?
Until you start digging in?
And there's too much work.
Until you start digging in, andthere's too much work.
And again, a long hill to go,even on the ChatGPT side.
But I don't know from personalexperience.
I just find it a lot better,way better experience.
Let's talk about how AI willactually change the actual
(10:57):
experience for fans.
Let's think pre, during, post.
I'm a Knicks fan.
That was not a great game lastnight.
Peter Yagecic (11:04):
I'm so sorry, but that was not a great game last
night.
Damian Bazadona (11:05):
But like, how might AI help me, whether that's
wayfinding, gamifying, likewhat are you seeing, what are
you thinking, even in sports ortravel or any other just
vertical, just curious how youthink it's going to impact the
actual experience itself.
Janette Roush (11:18):
Wayfinding.
I think there are hugeopportunities there Wayfinding
and translation.
So again, that personal devicethat lives on your phone.
Imagine you're a wheelchairuser.
My device, my Google account,would already know all of the
dimensions of my uniquewheelchair, because this is an
(11:40):
issue across travel right nowthat there's a number of
companies working to solve insome way, because a hotel might
say the room is accessible andthat is a absolutely meaningless
adjective.
I don't know if my you know ifit's a roll-in shower or, you
know, do I need to have twopeople with me to help transfer
(12:00):
to the bed or is that equipmentalready there?
And so for a future version ofGoogle Maps to understand where
is my hotel and what is the pathto the theater or to the arena,
that includes curb cuts atevery curb so that my wheelchair
can make that trip.
I think that's like the level ofaccessibility help that we can
(12:24):
expect in the future.
You know, we already seeamazing things for people who
are blind or have low vision andthe ability of you know.
Imagine a pair of meta glassesthat can do audio descriptions
for the user as they're outnavigating the real world.
So I think some of the greatestopportunities are in that arena
(12:47):
.
And then I think it comes down,like for the fan experience
I've kind of been thinking about, like what would it be like to
be able to experience the gamefrom the perspectives of
actually the players on thecourt?
Damian Bazadona (13:02):
Yeah.
Janette Roush (13:03):
And so, like that's, I'm cheating now and
like moving ahead of like what'sthe ultimate dream 10 years
from now?
But imagine being able to see amusical, but now you're seeing
it through the eyes of Elphabainstead of through the eyes of
the audience member.
Like experiences that wouldnever be available to us.
I think that's the kind ofthing AI can make real us.
Damian Bazadona (13:28):
I think that's the kind of thing AI can make
real.
Yeah, you know, just buildingon.
This is really not necessarilyabout the experience itself, but
almost the transaction processof being able to map out where I
want to sit.
I will tell you, if I I am anaisle snob, I will pay a premium
for an aisle.
I will sit on the roof ifthere's an aisle seat, and so
any.
It does not take a genius to beable to watch me for like 10
minutes and go oh, this guywants an ILC, and so pretty much
no matter what and I knowthere's a lot more automation
(13:49):
happening on that front of botchbeing able to shop within as
agents, excuse me, which wouldbe incredible, which I know,
peter, you've been testing with.
I just think there's obviouslya massive opportunity on that
front.
Speaking of Peter, peter, arethere any questions coming?
I just want to make sure I'mbeing mindful of any questions
that might have come in.
Peter Yagecic (14:09):
Yeah, we do have some audience questions coming
in.
Going back to your toothbrushexample, jeanette, deep research
gen AI seems like a much bettercomparison shopper than earlier
LLM models.
Just ask deep research to helpfind the cheapest tickets for
any given Broadway show thisweekend and see what you get.
I think the question is a yearfrom now, will any humans be
doing this research or will itall be our personal AI agents?
(14:32):
But maybe could you take aminute to kind of let the
audience who may not be familiarwith deep research we were
talking about it right before wewent on camera the difference
of kind of what deep research isdoing, and then you know will
it take over all of the researchthat we're doing?
Janette Roush (14:54):
So deep research, and this is annoying because
the people at these companiesare so terrible at naming their
products.
And so Google Gemini came outwith a deep research back in
December, which then OpenAI cameout with their own version of
deep research a month later.
I think Perplexity now offersdeep research.
I think Claude has rolled itout, but they've called it
something slightly different.
(15:15):
But what the technology isdoing is essentially I think of
it as going down an internetrabbit hole.
So all of these services thatall offer some version of this
type of work, it's you lay out aproblem for it and then it may
ask a couple of clarifyingquestions or let you see its
(15:37):
research plan, and then thelanguage model goes off and
spends up to 30 minutesresearching on the Internet to
see what are the best ways toanswer this challenge.
You have given it, and it isgreat for comparison shopping on
products.
It is really good if you'redoing a competitive analysis.
(15:57):
So you know, brand USA, we'relooking to move more into the
business event market, so that'sgetting more conventions to be
held in the United States andfor you know and this will be a
new piece of work for us Like,we're not going to have sales
(16:29):
teams that are literally bookingthe convention centers, because
that happens at the local level.
So what is our role?
And deep research does a greatjob of taking kind of a big,
needy, diffuse project like thatand coming up with a
McKinsey-style 50-page reportthat is based on 30 minutes of
(16:52):
research.
It is insanely valuable.
Peter Yagecic (16:55):
More are coming in if we want to do another
audience question.
Damian Bazadona (16:57):
Do it, do it, do it, do it.
I've got 400, but you go.
Peter Yagecic (17:00):
Starts out?
Security, security, security.
Should we operate under theassumption that anything we put
into chat GPT is not private andcould show up anywhere?
Is there a way to protectourselves?
Which I think the answer tothat is yes, but, Jeanette, I'll
let you take a whack at that.
Janette Roush (17:13):
Yes.
So I have done a lot ofstudying on AI governance and
kind of what are like how do weresponsibly work with AI in our
companies?
And so, like I'll say for mepersonally, I am a privacy
fatalist.
I have no assumption thatanything I ever do or say is
private anymore.
So I'll put my blood work inthe chat GPT so that I know what
(17:37):
my doctor is going to saybefore she says it.
But that is not how I wouldapproach privacy and security.
You know, on behalf of ourbusinesses or on behalf of Brand
USA, and so for both of those,the first step is use a paid
model, because that gives youthe opportunity to turn off a
toggle that says that thelanguage model can use anything
(18:02):
that you put into it as trainingfor future language models.
So that's the first step.
It's going to in most cases, ifyou're using a chat GPT team
account or an enterprise account, that's going to give you SOC 2
security.
So again, it keeps your datasafe from hackers when it is
moving to the cloud where it'sgoing to be analyzed, and then
(18:24):
while it is at rest in the cloud.
But then there's separatequestions about what is okay to
put in a large language model.
That isn't necessarily relatedto security.
So I'd say yes, I would feelpretty comfortable knowing that
I have a SOC 2 compliant closedsystem putting most items into a
(18:47):
language model.
I'm not going to put the schemafor a new iPhone into it if I
work at Apple and I'm usingChatGPT, which is heavily
invested in by Microsoft, butthey're not doing that anyway
and most of the stuff we'redoing is not going to be at the
level of state security but interms of what else is not okay
(19:10):
to put into a language model PII.
So you know personallyidentifiable information.
It is a violation of GDPR toput any you know names and phone
numbers, email addresses of anEU citizen into a tool like
ChatGPT, because the EU says youhave to give like affirmative
(19:32):
consent in order for somebody touse your information in a
language model.
So that's just a best practiceto not do that, because it's not
your information to give away.
And then you want to look atthe third-party permissions of
things like research documents.
So one of my favorite use casesfor AI personally is to create
(19:54):
a project inside of ChatGPT orinside of Clod.
It looks like little foldersand you can put background
information into those littlefolders but some background
information, like a researchreport from a third party.
If it's syndicated research,they may say in your contract
that you're not allowed to sharethis information with a third
(20:16):
party.
A language model is legallyconsidered a third party.
So it's time for all of us.
You know, as small businessowners I bet on this phone or
people who work for really,really big brands you want to
make sure that you are talkingto your vendors.
Like, rather than ooh I don'tknow if it's okay to do that
(20:37):
Like, let's get it out of.
You know the secret realm.
It's okay to say I'd like toput this stuff into a large
language model that has trainingturned off.
Can we add those terms to ourcontract?
Peter Yagecic (20:52):
Fandom Unpacked is brought to you by Situation,
an award-winning marketingagency built for live
entertainment that champions thepower of unforgettable shared
experiences around the world.
We offer full marketing andcreative services for
experience-based brands in liveentertainment attractions,
theater sports, arts and cultureand more.
Damian Bazadona (21:20):
Check us out at situationinteractivecom.
Now back to our Q&A.
There's one of the common.
I'm going to switch gears alittle bit.
I want to talk a little bitabout authenticity in terms of,
and content development, and soI feel like one of the more
common questions I've seen andI'd love to get your take on
this is how does AI run the riskof making content feel less
authentic?
And what I've heard from contentdevelopers oftentimes is like
hey, look, I use it for a lot ofstuff, but my voice I'm so
niche that it would never reallycapture what I do.
(21:44):
Now I'm like you're wrong,because if you train it enough,
it's going most likely to be youif you want it to be Now.
I personally don't want to livein a world like that.
So there's something about youhave your own personal judgment
of things you want to hold on to.
But it's interesting I hearthis a lot that the belief of
like oh, it'll never get me, Ithink is missing the point of
the power of what you're dealingwith.
(22:05):
What's your take?
Because this is a big one.
A lot of people on this webinarright now, on the live stream
these are content creators, andso what's your take?
Janette Roush (22:18):
I'm just curious on AI and development.
There's so many pieces to it,right?
Because part of it I'm like.
I think, yes, not today.
Can it perfectly replicate ourown unique voices?
I don't think that's very faraway and I think, from an
artistic perspective, totallynot there today.
I think there's a lot of funstuff you can do with Midjourney
(22:41):
or the new image tool inside ofChatGPT.
It's not human-replacing today,but but I think it could get
much closer pretty soon.
And that you know, in my worldthat creates a lot of
consternation because, you know,destination marketing
organizations are promotingdestinations and there is like a
(23:04):
big push to make sure thateverything we do to promote the
destinations is authentic.
We wouldn't want a, you know,phony version of the Empire
State Building to be used, evenif we knew in every single way
it was perfect.
That's also today.
And so if you think 20 yearsago when Photoshop was new and
(23:27):
everybody's like, oh God, I wantto on the cover of our brochure
or for this trade show booth, Iwant to make sure that the blue
sky behind the city skylinereally pops, and there's a
little bit of hand-wringingabout how much Photoshop is okay
and how much is too much.
But we don't have thatconversation anymore and I'd
argue people completely expect,if they see a sky in the photo,
(23:51):
that somebody touched that skyup and if the sky were kind of
hazy and dreary that peoplewould kind of be upset that you
didn't do that work.
So what we consider to be ournorms are going to subtly shift
over time.
Damian Bazadona (24:07):
If you think about it's almost like how do
you measure authenticity?
That's like a big one, right?
Because it's like how, like,how, like I don't know.
You know, I would imagine someof the pushback because, I agree
, I I'm hard, how far technologyhas come, how fast.
And yeah, you can like, rightnow you can tell when
something's a little bit off,you're a little bit off here.
But if you call this trainingwheels and go 20 years from now,
oh, the technology itself isgoing to figure it out.
(24:30):
The question, I suppose, reallyis it's consumer push, it's
consumer pushback, right, if Ifound out that I was talking to
someone who I think I'm amachine, I think I'm talking to
a human, so that really goes tolike, what's the role of brands
and being transparent of this isjust, you're dealing with my AI
self, but also the role ofgovernment and compliance, which
I think is that's.
The government is clearly goingto step in at some point and
(24:55):
try and put guardrails on this,because the knock-on effects of
this are going to be incredibleif this just remains unleashed
on our economy, on our society.
This is a hard, this is atricky question.
The crystal ball of like, whatis government?
What is compliance orgovernment?
What do you feel like?
Where do you see that maybegoing?
Janette Roush (25:10):
A lot of the compliance stuff, like I don't
know how much the US governmentis going to step in from a
regulation standpoint.
Certainly other governmentswill.
So as long as we're selling ourproducts to European citizens,
the EU AI Act will dictate a lotof kind of the bigger things
that all of us can use AI to do.
(25:31):
But ultimately, I think it doescome down to the brand level to
be very clear about how do youuse AI and where are your
company's personal limits,because nobody is going to tell
us what to do.
There is no real right or wrong.
It's just what do wecollectively agree to do
(25:54):
together?
And so for a brand that canlook like AI guidelines.
So that's something you know werolled out when I was pretty
early on here at Brand USA andit's something that I would
suggest every organization needs, because it's the opportunity
for you to come together and sayhow comfortable is your
(26:14):
organization with pushing thelimits when it comes to AI use
and maybe collectively agree.
Or leadership comes together andagrees oh, we're gung-ho on
this and we want to provide thetools for everybody to use it
safely.
Or you may say no, we want tokeep AI out of the workplace.
I'd suggest you're at risk forshadow AI or BYO AI if you did
(26:39):
that, but that could be aposition to take, and then it
allows you to say how muchalteration are we okay with
before we want to tell people,you know, with a note on our
website or a footnote to apicture, to some way be clear,
because I think that is like.
The number one reputationalrisk is that people don't want
(27:02):
to be tricked, and so it's notthat using AI has a moral
implication, but you want tomake sure that people understand
what the AI did and what thehuman did.
Damian Bazadona (27:16):
Yeah, well, let me okay.
So let me switch gears a littlebit.
Talk about I'd love for you totalk about your must-do list for
marketers, right?
I think it'd be helpful for thefolks on here to see what
you're playing with and sometools.
Janette Roush (27:28):
Great.
So for me, chatgpt Team is kindof the go-to account.
If you're only going to do onething for your team, get ChatGPT
Team.
It's going to be $25 a user amonth if you purchase monthly, I
believe, or, yeah, if youpurchase annually, that's the
monthly rate and your ChatGPTcomes out with the newest
(27:50):
additions, like the bestfeatures, faster.
I don't think it out with thenewest additions like the best
features, faster.
I don't think it's always thegreatest model.
I'd say today it's probably thegreatest model, but then Claude
will leapfrog over it.
So if you're looking for justone answer, looking at this
entire list, it's chat GPT, apaid version.
(28:11):
But then on top of that Ireally like Claude, which also
offers kind of a safer teamaccount because it just has a
better writing style.
And if you're somebody who getsinto a long conversation with
something like ChatGPT, you'llfind that ChatGPT will start
spitting out the same ideas overand over and it kind of just
(28:34):
gets stuck in a rut.
Cloud is much less likely to dothat in my experience.
And then if you look at theGoogle suite, google's you know
their new model that just cameout is currently top of the
leaderboards.
That is not typically the casefor Google, so it would be lower
on my list to buy unless Ialready had Google Workspace.
(28:54):
But I will say some of theother products that I have on
there, like Google Notebook LMterrific research tool.
For that one you just uploadPDFs or other files and then it
will answer your questions onlylooking at your files and it
will open up your files and itwill highlight where it found
the answer to your question, sothat it's a way to work through
(29:18):
hallucinations and be reallygrounded in truth.
And then, just like picking afew from the other categories in
the image and voice tools,midjourney takes some training
to understand how to createimages with it, but it's well
worth it.
Training to understand how tocreate images with it, but it's
well worth it.
Descript is a fantastic tool forediting short videos or audio
(29:38):
content.
So you could take this podcast,feed it into Descript and then
they have what they call an AIunderlord.
So not your overlord but yourunderlord, and it will find all
of you're like I'm looking forfive 30-second segments and
we'll find the five mostinteresting segments from the
podcast.
It will make it vertical orhorizontal.
It'll put whatever style ofcaptions you want on it.
(30:01):
You can do more editing tricksif you're not.
You know I'm not an editor, butif you already know how to do
that, it's easy to do it withinthe tool and even so, somebody
like me I could fake somebodyinto thinking I was a podcast
producer by just having this one$20 a month subscription.
I also under the everythingelse bucket.
(30:22):
Beautifulai is a terrificpresentation tool and I think
it's worth trying out.
Napkinai, if you are evermaking infographics for an ad
meeting.
Damian Bazadona (30:32):
Great See, I love it.
I'd sometimes I know a lot ofpeople.
I asked Well, what actual toshould I tool should I use?
I mean, there's a lot ofdifferent and I know a lot of
people on cloud, by the way, andI've tested notebook LM to me
as one of my favorite.
It was one of those things whenyou, when you see, I put my own
medical stuff to it and it gaveme better.
It gave me, but with betterbedside manner.
Peter, I know there's more.
(30:54):
I want to be respectful of allthe questions we have.
Peter Yagecic (31:01):
Sure, yeah Well, I think I know the answer to
this one, but as it relates toresearch, how can we specify
where we want AI to pull answersfrom to avoid research
information bias?
You mentioned that notebookwill just scam what you give to
it.
But is it as easy as saying?
Because most of these areconversational interfaces, you
can just say that to them.
Janette Roush (31:16):
Sometimes I mean, if you're depending on what the
research question is, sometimesit will go, it will lead you
astray, and I think the problemwith asking it is that it wants
to make you happy and so it willlie.
So there is no asking AI tofact-check or proofread itself.
Peter Yagecic (31:37):
We do use AI tools not Descript, which I'm
very interested in, but we useit.
We feed all the transcriptsfrom this very series into it,
asking for great quotes, and ithas hallucinated those quotes.
It's come up with great stuff,but stuff that nobody actually
said.
So yeah, take it with a grainof salt.
Janette Roush (31:53):
Yeah, that's why the notebook tool is so great,
because it will pull.
You will look at the actualquote in real time.
Peter Yagecic (32:00):
Yeah, and you know, I think, as Damien
mentioned, this is just thebeginning of probably a
continuing conversation.
One question that I would wantto ask on behalf of our audience
is how can people best stay inyour orbit, jeanette?
I mean, you are staying on theforefront of this, it's in your
job title.
What's the best way for peopleto continue to make sure that
(32:22):
they get continued wisdom fromyou in this topic?
Janette Roush (32:25):
I write a lot on LinkedIn about AI actually, so
LinkedIn is the best spot tolearn more from me.
So LinkedIn is the best spot tolearn more from me.
And then I follow so manydifferent newsletters that I had
to open up a separate emailinbox to track all of them, so
it doesn't just overwhelm mywork inbox.
Damian Bazadona (32:44):
Listen to podcasts I am heavily consuming
this information so that youdon't have to work quite so hard
at it.
That's how I saw I mean, I'veknown Jeff for years, but that's
what kind of triggered thisconversation her LinkedIn feed
of watching all the stuff thatshe's doing, which I just think
is just awesome.
Peter, can I ask the lastquestion?
I just want to ask Go for it Onthe horizon that you're excited
(33:07):
about.
And obviously there's a millionthings on the horizon in the AI
space.
It's like sky's the limit.
But, like you know, I'd love toask kind of, what's on the
horizon that you're excitedabout that?
Or you know, if you have adream up, an AI experience in
your head of like, oh, here'swhat I can see happening.
Like I'd love to talk a littlebit more about the future.
Janette Roush (33:26):
I really see the idea of our one personal AI
device being the way that weinteract with AI in the future,
and so I think, from a liveevent perspective, what that
means like, let's say, mypersonal AI ends up being
Google's Gemini.
Gemini, like Google, alreadyreads all of my emails and it
(33:48):
has, you know, parts of mycalendar, and you know so much
of our personal lives is alreadyreally deeply known by Google.
So if that were to become mypersonal AI, it would know when
do I have free time coming up inAugust, and so does my husband
and our stepkids, and it's beena long time since we've gone to
(34:11):
see a Broadway show been a longtime since we've gone, you know,
to see a Broadway show, andthey know that there was some
chat message that the kids hadthat I didn't see, that talked
about some or other show, and atthe point that I am ready where
I would say, oh, I need tostart planning a family activity
, gemini will open up on myphone a I don't know if I would
(34:33):
call it a website, but it won'tnecessarily have to look like a
text chat interface.
I think, based on what we aredoing, it will create an
interface on the fly thatmatches the activity that we're
talking about and that I woulduse that interface to choose a
(34:57):
vacation destination or to picka Broadway show.
I think, if it were somethinglike picking a Broadway show,
that if I am a short-form videoperson, that it would create
three 15-second video clips fromthree different Broadway shows
to help give me a taste of whatit is, from three different
Broadway shows, to help give mea taste of what it is, and it
may steal those images you knowfrom YouTube.
Or maybe we keep on ourwebsites for our products like a
(35:23):
library, like a capsule ofhere's all the video and image
content that you can pull in toyour LLM to create these spun-up
websites.
Here's all the descriptionsthat you would need to know, and
it wouldn't be something aperson might want to read.
But let's say, the websiteincludes here's all the target
(35:44):
audience profile segments whoare currently buying tickets to
this show, and then that's a waythat the data that we already
have can help us find morepeople who fit into those you
know audience target segments,without them needing to raise
their hand first, becausethey're just doing it
(36:06):
intuitively through using theirdevice.
Damian Bazadona (36:09):
Yeah, it's exciting.
I think that it's just anextremely exciting time to be in
the marketing space for liveexperiences and the whole
intersection of phantom.
So I think to me I'm like it'sscary to some, exciting to
others.
The reality is it's a massivechange and we haven't even
started seeing the impact of itand the knock-on effects.
And with change is opportunity,and so I I've never been more
(36:32):
excited to be in this space.
I think it's actually very cool.
Peter Yagecic (36:36):
That's going to do it for this episode of Fandom
Unpacked the podcast.
If you liked what you heard,please be sure to leave us a
review on Apple Podcasts.
Find out how to join us livefor an upcoming recording at
SituationLivecom slash fan.
We'll see you next time, truebelievers.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.
Ridiculous History
History is beautiful, brutal and, often, ridiculous. Join Ben Bowlin and Noel Brown as they dive into some of the weirdest stories from across the span of human civilization in Ridiculous History, a podcast by iHeartRadio.