The recent episode of MC Fireside Chats, hosted by Brian Searl, brought together a panel of experts to discuss the rapidly evolving role of artificial intelligence in the outdoor hospitality industry. Searl opened the broadcast by emphasizing that while technology is advancing at an unprecedented pace, the fundamental goal for campground and RV park operators remains unchanged. The core objective of adopting these new tools is not to replace the human element, but to weave technology seamlessly beneath the surface to enhance the overall guest experience.

Joining the conversation was a diverse group of industry leaders, each bringing a unique perspective on technology and operations. The panel included Matt Whitermore of Climb Capital and Unhitched Management, Ravi Parikh from RoverPass, Mychele Bisson of Wayhaven Resorts, Caleb Cook from CampLife, and Nick Purslow of Posh Outdoors. Together, they represented a broad cross-section of the industry, ranging from property management software developers and marketing specialists to national portfolio operators and modular lodging providers.

Ravi Parikh, a self-described AI enthusiast who uses the technology for hours each day, kicked off the technical discussion by highlighting the latest advancements in AI personal assistants. He detailed the rapid succession of tools like OpenClaw, Hermes, and Elon Musk's GrokBot, explaining how these agents can integrate directly into communication platforms like Slack, iMessage, and Google Workspace. Parikh painted a picture of a near future where these AI agents act as comprehensive executive assistants, capable of reading emails, summarizing team chats, scheduling meetings, and executing complex workflows autonomously.

The conversation then transitioned into the frustrations and global dynamics of the current AI landscape. Matt Whitermore expressed a common grievance among power users, noting that flagship models from companies like Anthropic sometimes seem to degrade in performance just before a new version is released. Parikh contextualized this by explaining the intense international competition, noting that highly efficient Chinese models like Kimi 3 are driving down the global cost of computing, while tech giants like Apple are simultaneously pushing to enable users to run powerful AI models locally on their own devices for free.

Grounding the high-tech discussion in day-to-day park operations, Mychele Bisson shared how she utilizes AI at Wayhaven Resorts. Having spent the summer actively managing her properties, she admitted to falling slightly behind the bleeding edge of AI news, but emphasized that her operations team heavily relies on the technology for practical applications. By using AI to draft standard operating procedures, build training modules, and instantly pull complex reservation reports, Bisson ensures her back-end operations are incredibly streamlined so her staff can remain fully present for their guests.

Nick Purslow echoed the sentiment of practical, targeted AI use, specifically within the realm of marketing for Posh Outdoors. Rather than relying on AI to completely generate content, Purslow uses platforms like ChatGPT as a creative sparring partner to rapidly ideate variations of marketing copy and social media captions. He firmly believes that while AI is an incredible tool for overcoming writer's block and scaling creative output, human taste and editing remain absolutely essential to prevent the brand's voice from feeling robotic or inauthentic.

Caleb Cook from CampLife brought a philosophical lens to the discussion, balancing his excitement for new software features with a cautious approach to consumer trust. He revealed that CampLife is developing new tools to help smaller parks navigate staffing shortages and optimize email marketing campaigns. However, Cook stressed that the foundation of true hospitality is making people feel at home, which requires a baseline of trust that operators must carefully protect when deciding how and where to deploy automated systems in front of their campers.

The panel then engaged in a nuanced debate regarding the ethics and efficacy of AI-generated imagery in campground marketing. Brian Searl and Matt Whitermore argued that AI can be a highly practical, budget-friendly solution for small businesses, such as using generative fill to add lifelike models to an authentic, empty photograph of a resort's pool or fire pit. They posited that as long as the underlying property depicted is entirely real and accurate to the guest's eventual experience, utilizing AI to enhance the photo's lifestyle appeal without hiring expensive professional models is simply a modern evolution of traditional photo editing.

Conversely, Nick Purslow and Caleb Cook warned that consumers, particularly Gen Z, are becoming increasingly hyper-vigilant and skeptical of anything that appears artificial. They noted that some brands have faced significant backlash for using AI-generated models, as modern audiences are quick to scrutinize and call out synthetic content in comment sections. Mychele Bisson validated this cautious approach by sharing her own strategy, noting that Wayhaven Resorts relies entirely on organic, guest-generated content—like GoPro footage of kids on water slides—which has fostered deep authenticity and contributed to a highly successful season with significant year-over-year revenue growth.

As the broadcast concluded, the speakers offered their final thoughts on how the industry should navigate the AI revolution. Ravi Parikh strongly encouraged all operators to invest just a few weeks into deeply learning how to prompt and utilize AI, promising that the resulting time savings would yield an immediate and massive return on investment. Ultimately, Brian Searl, Mychele Bisson, and Caleb Cook reached a unanimous consensus: AI is a permanent fixture in the business landscape, but its greatest value in outdoor hospitality lies in automating the back office so that humans can deliver unparalleled, face-to-face service.

Welcome, everybody, to another episode of MC Fireside Chats.

2

My name is Brian Searl with Insider

Perks and Modern Campground.

3

Excited to be here with you.

4

That was close, guys.

5

I was two seconds, I forgot my hat.

6

I had to go find it during the intro.

7

I was like, almost didn't make

it, sat down half a second before.

8

But I'm here.

9

I made it, so not that anybody's excited.

10

They're all here to see you guys.

11

But excited to be here for another

week four episode of MC Fireside Chats.

12

We're gonna talk about AI technology.

13

We've got a couple special

guests, some new people here.

14

Let's go around the room and briefly

introduce ourselves 'cause some of

15

us haven't been together for a while,

and then we got some new people too.

16

Matt, you wanna start us off?

17

Matt Whitermore: Yeah.

18

Hey, everybody.

19

Matt Whittemore here, director

of market expansion at Climb

20

Capital and Unhitched Management.

21

National platform, growing quickly, adding

parks through both third-party management

22

and through our investment platform.

23

Always enjoy these chats, Brian,

so appreciate you having me

24

on and excited to dig in here.

25

Brian Searl: Yeah.

26

Thanks for being here, Matt, as always.

27

Ravi, our new recurring

guest for week four.

28

Ravi Parikh: Hey, everybody.

29

I'm Ravi.

30

I'm the CEO of RoverPass, which is a

property management software company

31

Brian Searl: That's it?

32

You're just gonna undersell

yourself like that, man?

33

Ravi Parikh: I use AI about 10 hours a day

personally, and our whole team uses it, so

34

I'm excited about, I can s- speak to the

gospel of AI for hours and hours on end.

35

Yeah, we're we're a very AI-pilled

organization over here at RoverPass.

36

Brian Searl: Awesome.

37

Glad to have you here, Ravi.

38

Mychele?

39

Mychele Bisson: I'm Mychele Bisson.

40

I am the founder of Wayhaven Resorts.

41

We own campgrounds and marinas

from Alaska down to Florida.

42

And I'm always trying to be here,

but I'm usually in my campgrounds

43

during the summer, so I'm glad to

be back today to be able to have

44

these conversations with you guys.

45

Brian Searl: Yeah.

46

Thanks for being here.

47

Appreciate it.

48

And our two special guests.

49

Caleb, you wanna go first?

50

Caleb Cook: Sure, yeah.

51

I'm Caleb Cook.

52

Good to be with you guys.

53

And I'm from CampLife.

54

I was formerly the marketing and creative

team lead, but actually just made a leap

55

into the product design side, so I'm kinda

joining you from both perspectives today.

56

And yeah, CampLife is like RoverPass, a,

a reservation and park management system.

57

And now a marketing system too.

58

We recently released some email marketing

tools too that we're pretty excited about.

59

So yeah, that's that's

the 40,000 foot view.

60

Brian Searl: You know

what we should do, Matt?

61

Matt, you and I should have a project

where we bring back MTV's Celebrity

62

Deathmatch with AI, and then the CampLife

and RoverPass guys can fight it out.

63

Matt Whitermore: All right.

64

I'm in.

65

Ravi Parikh: Yeah, let's do it.

66

Matt Whitermore: All right.

67

Ravi Parikh: We'll, I'll build a,

I'll build a little robot where

68

me and Caleb can go head-to-head.

69

Brian Searl: All right.

70

Last but not least, Mr.

71

Nick Purslow, who, Nick, are you

still three hours north of me

72

and still haven't come see me?

73

It's raining up there.

74

Nick Purslow: Ah, no, totally.

75

You haven't come see me.

76

Yeah, Edmonton, you've upgraded

the the intro music since the

77

last time I was been on here.

78

It's great.

79

Brian Searl: Yeah

80

… Nick Purslow: but yeah,

and- You- Sorry, go

81

yeah.

82

I'm Nick Bisson.

83

I'm a co-founder at Posh Outdoors.

84

We partner with landowners through

revenue share partnerships, so they want

85

to add unique year-round kind of modular

lodging structures on their property.

86

We'll install them, provide marketing

services to ensure that they're filled

87

and take a percentage of the revenue.

88

So we've got one project up near

where Brian and I Brian and I

89

are in the Canadian Rockies and

we've got a partnership in Texas

90

as well, and we're hoping to

scale heading into the next year.

91

Brian Searl: Awesome.

92

Excited to hear what you've been up to.

93

Yeah, I've only been to Edmonton

once in my life, man, so don't like-

94

Nick Purslow: Yeah

95

… Brian Searl: take offense.

96

I went up there with-

97

Nick Purslow: At the same time

98

… Brian Searl: I went up there

to catch a train last year.

99

We took one of the sleeper

trains across to…

100

it wasn't a sleeper for us, but we

took it across to Winnipeg and then

101

went up north to Churchill, so-

102

Nick Purslow: Nice

103

Brian Searl: And I went to a really

good brewery and pizza place that

104

I gotta remember to tell you about.

105

I'll f- I'll…

106

I don't remember the name

of it off the top of my

107

Nick Purslow: head, but- Isn't isn't

Churchill where they have the polar bears?

108

Brian Searl: Yes.

109

Yeah.

110

Nick Purslow: Oh, nice.

111

Yeah, that's

112

Brian Searl: lovely.

113

Over the years, like

10,000 beluga whales- Nice

114

so you can…

115

It's crazy.

116

Nick Purslow: Yeah.

117

Brian Searl: Yeah.

118

Really good place to visit.

119

Okay, so to our recurring guests who

are typically here with us, Mychele,

120

Matt Ravi, you're a recurring guest

now, so you can weigh in if you want.

121

Is there anything that's come across

your desk in the last month or so

122

that's related to AI technology, stuff

like that, that we should be talking

123

about or you guys wanna talk about?

124

Matt Whitermore: I would

125

Brian Searl: say my-

126

Mychele Bisson: Got so much

127

Brian Searl: Oh, we all went at once.

128

Okay, go ahead.

129

Ravi Parikh: I'll

130

go.

131

Mychele Bisson: I was

gonna say, go ahead, Ravi.

132

You said that you could

talk about this for hours,

133

Ravi Parikh: there's so much.

134

Every week there's something new.

135

I'm a subscriber of x.com,

136

formerly Twitter, and the Twitterverse is

popping off about the new Grok bot which

137

is a competitor to Hermes and OpenCloud.

138

That's the latest.

139

And then Claude released a better

version of Claude Remote, so you can

140

basically code while you're walking

your dog or, wherever, like hanging

141

out with your friends, just pop

open your phone and start shipping

142

code while you're doing that.

143

So I would say those are like the latest

small things that I've seen come up.

144

P- and I don't know, GrokBot

might be big, but I haven't

145

messed with it yet personally.

146

Brian Searl: Maybe set the stage for

our audience and I don't know, Matt,

147

if you've played with any of this

or Mychele, but set the stage for

148

our audience for who th- who don't

know about OpenClaw or Hermes or-

149

Ravi Parikh: Yeah, yeah.

150

Brian Searl: So- Because GrokBot is,

I think the consumer geared version

151

of like your easy way into that world.

152

Ravi Parikh: Yeah.

153

So the first version of this was OpenClaw

that came out, I don't know, it's probably

154

been like seven, eight months now.

155

A guy named Peter Steinberg

released it for free on x.com

156

and then OpenAI ended up hiring him

as their I don't know exactly his

157

position, but it's essentially like

he's building what everybody's gonna

158

use as their personal assistant.

159

So Iron Man's…

160

What's the name of Iron Man's assistant?

161

The voice- I don't know.

162

Brian Searl: I don't know

what you're talking about.

163

I can't remember.

164

'Cause it was played by

Gwyneth Paltrow, right?

165

Ravi Parikh: I don't remember the name.

166

Not Cortana.

167

That's, Man, I can't

remember what it's called.

168

But anyways, the idea was that everyone's

gonna have a personal assistant,

169

OpenAI got excited about it and

then hired this guy Peter Steinberg.

170

He released this product called OpenClaw,

and OpenClaw basically connects into

171

like everything that you've got.

172

Google Calendar, your email Google Drive.

173

It can make phone calls for you.

174

It can do all these different things.

175

So like you just tell AI "Hey, schedule me

this meeting," or, "Call this restaurant

176

and make this reservation," or whatever.

177

Whatever you want it to do it can do.

178

Then the next one was Hermes

that came out, which was like…

179

I don't know if it's Hermes or Hermes like

the designer bag brand, but Hermes maybe.

180

They were the next iteration on OpenClaw.

181

And then the Twitter sphere

was going crazy about that.

182

And then the next iteration i- of

that is GrokBot, which is Elon's

183

competitor to Hermes and OpenClaw.

184

It seems like OpenClaw has died off.

185

Like I don't really know

what happened to it.

186

And so like in seven months time,

we've released three different

187

versions of your personal assistant.

188

At some point one of

them's gonna take off.

189

I don't know which one yet.

190

I kinda gave up after OpenClaw

and haven't messed with Hermes

191

and Grok Bot, but I need to.

192

I need to spend some time

kinda trying to set them up.

193

Brian Searl: Yeah, for clarity of our

audience, like I think an easier way

194

to understand this is and you tell

me if I'm wrong, Ravi, with how I'm

195

perspecting or perceiving this or

explaining this or whatever, right?

196

Is that, your regular Claws, your

Chat GPTs, your Geminis are…

197

While they can connect and search

your inbox for Gmail and do things

198

like that, they're mostly like

conversation in, conversation out.

199

Search, give me ideas, write

me social media posts, write

200

me code, do those things.

201

Whereas these tools that we're talking

about, OpenClaw that came out in

202

January, Hermes, the Hermes that

followed, whatever we're pronouncing

203

it as, Grok Bot now is the idea of

more taking control of your computer

204

and operating it as a human would.

205

Ravi Parikh: Yeah, like the big ways

that you use these tools you can connect

206

it into your iMessage, you can connect

it into WhatsApp, you can connect

207

it into Slack or Microsoft Teams,

you can connect it into your email.

208

So every which way that you

could possibly talk to it in

209

all of the different channels.

210

If you're on your phone you might

not have the Slack or Microsoft

211

Teams app, so you wanna send it a

voice note, right from iMessage.

212

That's the first part is capturing

all that data, and then from

213

there it's connected to all of

the data sources that you have.

214

So it could be like work data sources, it

could be personal data sources, anything,

215

and it can access those data sources and

then do certain actions based off them,

216

based off the tools that you give it.

217

If you wanted to connect it to something

like Twilio, which is used for like

218

placing phone calls or sending text

messages, it can place a phone call

219

or it can send a text message for you.

220

If you're like having a chat

with somebody via text message…

221

This always happens to me, like I'm

like, I get an introduction to someone

222

via text message and I'm like, "Man,

like I can't connect my Google Cal

223

to my text message to schedule this."

224

So you just add your OpenClaw or

your Hermes or in this case I guess

225

GrokBot to that text thread, and

it can be like, "Hey, here's all

226

the availability," and then you…

227

it can handle scheduling for you.

228

So it's like really like your personal

assistant for everything, right?

229

Every morning I had it send me

a Slack message that's "Hey, you

230

forgot to answer all these emails.

231

You need to look at these conversation

threads inside of your Slack account

232

because your team is like talking

about all these different things

233

and they need your weigh in."

234

And then it sends me like a, a

series of metrics on RoverPass every

235

morning or every week or whatever

fixed schedule that I told it to.

236

So it can consolidate all of that into one

and then present that information to you

237

in which if you wanna text with it, if you

want an email, if you want a Slack message

238

it can give you all of that information.

239

Brian Searl: And I think the intention

behind GrokBot, and we'll have to see

240

if it goes there is to make it less

head explody for all the people who are

241

watching this, like what's happening.

242

Ravi Parikh: Yeah.

243

Yeah.

244

Yeah.

245

OpenClaw was made clearly by a developer

that was building it for himself and it

246

like feels like you're on the command line

when you're trying to use it, which is

247

not super accessible for like everybody.

248

Whereas GrokBot is "Hey, I'm gonna

build this like nice consumer slick

249

interface," the Elon way of doing things

and make it easy for people, right?

250

Like it's the Apple-ified

version of of OpenClaw.

251

And it's more stable from what I hear.

252

Like I always had problems with

OpenClaw where like it would just

253

kept breaking and that's why I stopped

using it, 'cause I got tired of trying

254

to fix it and I was like, "It'd just

be quicker for me to do this myself

255

than trying to fix you constantly."

256

So hopefully, hope- hopefully

that one is more stable

257

Brian Searl: Matt, have you

played with any of this new stuff?

258

Matt Whitermore: I haven't yet, but

I'm on the edge because, I was gonna

259

jump in and say personally, I'd say

the last month or so, and I don't know

260

if this is like AI psychosis or not,

but I've been disappointed with Opus

261

5 and Fable 5 on Anthropic models.

262

I feel like there's been this cycle

of whenever they're about to release

263

a new model The performance just

deteriorates, the model gets dumber and

264

just it gets it, being trying to keep

yourself on the cutting edge of AI is

265

one of the most frustrating things.

266

It's amazing and it's incredible and it's

mind-blowing and exciting, but it's also

267

insanely frustrating for a lot of reasons.

268

There's, like Ravi was saying, there's a

new feature every week, every day, right?

269

I've tried to quiet my mind with that and

focus on agent skills and automations and

270

just try and stick to the things that will

translate from provider to provider, agent

271

harness to agent harness, model to model.

272

Because I've been anticipating this.

273

It just feels different

for me for the last month.

274

I don't know if anybody

agrees with anthropic models.

275

Ravi Parikh: I-I think anthropic's

freaking out about Kimi 3 coming out.

276

So Kimi 3-

277

Matt Whitermore: Yeah

278

… Ravi Parikh: for those that don't know,

is the Chinese equivalent of Opus 4.8,

279

which w- which I guess Fable 5, which is

Anthropic's latest, most advanced model.

280

A Chinese company released something

that is like one-twentieth the cost

281

that's just as powerful, and so they're

like, it's this like interesting

282

tit-for-tat war because the Chinese

people the, the Chinese are distilling

283

their models from Anthropic and OpenAI.

284

So they're basically, what distillation

is, it's basically like at, they like

285

at scale send a bunch of questions

to Claude or OpenAI and read the

286

outputs, and then they train their

models based off of the output.

287

So it's like kind of this like

warfare that's happening between

288

the two countries, and like China

is like usually three to six

289

months behind the latest model.

290

So then if the US is supposed to compete

against China, and all the US businesses

291

are supposed to use a model that's 20

times more expensive than a model that is

292

just only three months behind the latest

model, wouldn't all the US companies

293

just switch to the Chinese model?

294

Which is we did do some

of that because it is-

295

Brian Searl: No, I do, I use

Chinese models in everything.

296

It's open source

297

Ravi Parikh: It's one-twentieth

of the cost, right?

298

So it's like- Yeah … not everything

you do needs to be the most cutting edge.

299

Cutting edge.

300

Right?

301

Like- Yes … so it's just this

like, it's interesting how these

302

like war dynamics are playing out.

303

But for us consumers I love it because it

keeps bringing the cost of AI down, right?

304

It's like a real fierce competitor-

in the race and it makes compute

305

cheaper for all of us over time.

306

So I think it's a good thing.

307

Brian Searl: Yeah, we saved

almost $15,000 a month in our AI

308

bills by moving to this stuff.

309

And they're all open source, and you

can s- you can use providers that

310

don't log your data, that don't save

your stuff, that have the privacy in,

311

in, everybody's logging it anyway.

312

Claude's logging what you do.

313

OpenAI's logging what you do.

314

Whether it's the Chinese or the Americans,

someone's keeping all your data, putting

315

it somewhere, and doing something with it.

316

It probably isn't something like this.

317

Ravi Parikh: And the n- the latest

thing is that just two days ago, Apple

318

released their MacBook Studios because

I think everyone's been wondering, like,

319

where is Apple in this whole thing?

320

Like, why do they not ha-

like, why is Siri so dumb?

321

Really.

322

You c- it barely can set an alarm for me.

323

And I think Apple's response to all

of this is that they want people to

324

run models locally on their computers.

325

So they actually released these super

powerful studio desktops where you

326

can lease them for 200 bucks a month,

and then there's another company that

327

you can rent your compute power to, so

then they'll pay you 250 bucks a month.

328

So you can actually make $50 a month

running a local model on your computer.

329

And it's this really

interesting, supply chain.

330

I don't know how to, I don't

even know how to explain it.

331

But you went

332

from having to pay Anthropic to now the

Chinese are competing against Anthropic

333

to Apple just coming in and being like,

"Hey, everybody gets free compute.

334

Just run like Quinn," or which is

like another one of the models,

335

"On your computer for free.

336

Like, all you gotta do

is pay for your power."

337

So then it's man, is everyone just gonna

have free compute constantly because

338

they're just running their models-

339

Brian Searl: Yes

340

… Ravi Parikh: on their local computer.

341

Brian Searl: Eventually, and that's

what's gonna break the AI bubble

342

and bring it all crashing down.

343

Not because AI is- Yeah … in itself

a bubble, but because the companies

344

can never actually make the trillions

of dollars that they're borrowing.

345

But that's a whole other topic.

346

Mychele, what's new in your world?

347

Ravi Parikh: Sorry, I digress.

348

Mychele Bisson: Oh, God, I can

tell you what, it's not that stuff

349

'cause, … that went way over my head.

350

So no, I think the AI stuff is great.

351

I was playing around with it

at the beginning of the season.

352

And then we got into the season, and

I'm just one of those weird people

353

who kinda wanna be in my parks.

354

So I was actually in the

parks for a lot of the season.

355

Plus, like I was mentioning earlier,

getting my daughter ready to go off

356

to college and becoming empty nesters.

357

And I kinda passed it off to my operations

director, and he's been kinda playing

358

around with it, and I don't really know

where he's gone, but I'm pretty sure

359

he's probably launching spaceships out

of one of the marinas at this point.

360

It just moves so quick,

I can't even keep up.

361

And just taking a couple months

off to focus on what I was

362

doing, I feel so left behind

363

Brian Searl: I think that's one

of the important things that, that

364

we, that I wanna balance on this

show as we go forward is, right?

365

Like technology is moving

faster than it ever has before.

366

Nobody on this show is gonna be able to

keep up with it, even me who has no life.

367

Or maybe Ravi has beaten me, I don't know.

368

Ravi …" likelier.

369

But so we're not gonna be able to keep up

with it, but I think it's very important

370

to look at where we are as an industry

and how this is gonna benefit us.

371

Because I think there's a, a push and pull

here between what you just said, Mychele,

372

and where I am with AI in trying to

develop products, and Ravi is, and where

373

Matt's playing with it, and he's also

trying to improve the guest experience

374

at Unhitched and everywhere else, right?

375

Is that I think outdoor hospitality

is perfectly positioned for this.

376

I think regardless of how much da- and

I say this to clients all the time on

377

calls, I, there's gonna be a ton of AI

slop out of there, out of h- out here.

378

There's gonna be, like, you're gonna,

we're gonna quickly go toward a Ready

379

Player One where people are putting on

headsets and living in virtual worlds.

380

But the one thing that AI will never

be able to duplicate is sitting by

381

a river, near the mountains, under

the trees, listening to the bir-

382

maybe listening to the birds, we'll

do that, but not the same thing.

383

And so I think we're perfectly

positioned to take advantage of that.

384

So the way we, as for you guys as

outdoor hospitality operators, 'cause

385

I'm not an operator, needs to think

about this, and the people watching

386

this show too, is how do I use this

technology to weave it under the surface

387

so it's not visible to the guest, but

makes their stay and experience better?

388

Is that fair?

389

Mychele Bisson: No, I

completely agree with that.

390

I know that for us we've used it as

in building a chat model so that our

391

managers can get information quickly

so that it's really easily trainable

392

so that they can build out training

SOP models when we bring new people in.

393

I know that we're working with a

reservation system right now to help

394

build out a new program where they

actually, everything is a little less, a

395

little easier on the reservation front.

396

Everything is kinda being AI-driven

on that end reporting so that we're

397

not searching for reports anymore.

398

Now we can just type in what we want.

399

Just things that we know will make

things easier on our side as operators,

400

making sure that we're being able

to streamline certain things on the

401

back end so that we can really focus

on the guest experience so when

402

they come in, we know them by name.

403

We can da- whatever purpose they

have in there, whether it's buying

404

something or, booking the reservation,

phone calls, things like that.

405

We're trying to streamline all

the back end so that our people

406

can be completely present.

407

And so that's the nice thing that I

think AI is starting to do for all of

408

us is that a lot of people are afraid

of it because it's not something that

409

they are used to, and they feel like it

takes away from the guest experience,

410

and I think it's completely the opposite.

411

I think it's streamlining everything on

the back so that we can be fully present

412

Brian Searl: Nick, do you guys have AI in

Edmonton or has it not made it that far?

413

Nick Purslow: Not made it that

far in the border yet, no.

414

We-- my, my job is basically we're

supplying these structures and

415

providing the marketing services.

416

We're not in the weeds, on the

ground, in the operations, in

417

the back end kind of thing.

418

So my most of my use cases for it

have been on the marketing side.

419

So for example, at our second location

out in the Texas Hill Country, we're

420

putting together, Instagram carousels

where it's 10, 10 slides each with a

421

very small kind of piece of text to

make it as readable as possible to

422

make the, the reader want to read on.

423

And I just find it a really useful tool

to basically as a creative sparring

424

partner, and it's just simple Chat GPT.

425

I'm not as clever as you, Brian,

when it comes to setting up all

426

the automations and workflows.

427

But just as a creative sparring

partner to help ideate.

428

For example, if I'm looking for some

marketing copy and I've got one sentence

429

in mind, I might ask it to, I might

feed it that and then ask it to spit

430

out 10 potential variations of that.

431

And I hate kind of AI written

copy, but it can be a great kind of

432

creative tool to just spit out lots of

variations and then relying on, human

433

taste to refine that a little bit.

434

So it just opens up kind of the

possibilities for me when I'm crafting

435

copy and I enjoy that, and I'd

never want to outsource that to AI.

436

But it's very good for, helping

take it to the next level.

437

Brian Searl: That's a good thing to,

to-- for the people who are listening

438

to this and wanna ground themselves

and are hearing what Ravi's saying

439

and hearing what I've said before

and hearing what Mychele said and

440

what Matt said and every- Caleb I'm

sure we'll get to you in one second.

441

You'll say some really

profound things too.

442

But just I think it's important

for the audience to realize that,

443

that this is not something where

you have to jump in and understand

444

immediately tomorrow what Ravi's saying.

445

Otherwise, you will get

completely overwhelmed.

446

Your goal should be to start small,

just like and I'm not saying you're

447

s- you're small, Nick, right?

448

With your- Yeah … AI.

449

But start with one thing and say, "What do

I need help with in my business that would

450

make me more efficient, that would allow

me to go out and serve my guests better?"

451

If that's analyzing your accounting,

if that's improving your SOPs, if

452

that's doing your Instagram reels, if

that's whatever, fill in the blank.

453

I think that's the place

you begin and you start.

454

And then just like we, I told people

two, three years ago who asked me

455

how do I start with AI, you go to AI

and you ask, "How do I start with AI?

456

This is my struggle.

457

This is where all my hours

are going," whatever.

458

And just the more you do with it,

the more you touch and play with it.

459

That's been Matt's growth

journey too, right?

460

I think, Matt

461

Like the, it- Totally … yeah,

the more you play with it, the more

462

you'll understand oh, I can do this.

463

Oh, I can do that.

464

Oh, I can do…

465

And the goal is not all time

consuming, it's to save you time.

466

So you can go down the rabbit hole like

me and Ravi apparently have, or you

467

can kinda go down the rabbit hole and

peek and see what's down there and then

468

come back up and realize that there's

sometimes it's nice to see what we've got.

469

Ravi Parikh: We trained our entire team.

470

They may not be quite at the level of

depth that I'm at right now, but i-

471

it's all about just thinking about,

like if you come across a problem, how

472

does AI solve that problem for you?

473

And like, when should you go

to AI versus not go to AI?

474

And the best analogy that I give to

people is if you just start asking

475

Claude how to build an airplane, it's

gonna start telling you how to build

476

an airplane and where to procure the

parts to do it for and start teaching

477

you about fluid dynamics, right?

478

So it's like the, the people who are

the most curious about what AI can

479

do for them are just gonna ask AI an

insane amount of questions, and then

480

it's going to give you answers back and

teach you exactly what you need to know.

481

It's like having the world's

greatest tutor in your pocket.

482

You just ask it anything that you

don't know about and pretty soon you're

483

gonna know everything about that thing.

484

And that's the best way to learn with AI.

485

It does take time to figure that out,

but you'll start to spot patterns

486

and see "Oh, okay, can I like

automate my bookkeeping with this?

487

How do we do that?

488

Oh, you need an API into

QuickBooks for that.

489

Okay, how do I set that up?

490

Oh, okay, go to this link, click on this.

491

I don't know how this works."

492

Then you can just take a screenshot

of that page and upload it into AI

493

and be like, "Click on that link."

494

And it can like spoon feed it to you

pretty easily and then you'll, your

495

level will start to advance, right?

496

And you'll start to understand

how all these things work and

497

flow and, just go from there

498

Brian Searl: Do you recommend

people build their own airplanes?

499

Ravi Parikh: Yeah.

500

Just definitely just build

a, the W- the, the Wright

501

brothers started one way, right?

502

Now you guys have Cl-

no, I'm just kidding.

503

Don't do that.

504

Definitely don't do that.

505

Brian Searl: I don't…

506

it interests me 'cause I've always

just wanted to have my own private

507

jet, but I realize I'll probably never

have enough money, but also I then

508

wouldn't wanna fly in the jet that

I built 'cause I don't wanna die.

509

So it's a quick no for me.

510

Mychele Bisson: It's probably not

a good idea to have a jet anyway,

511

because if you have one, then you have

to have two because the other one's

512

usually in maintenance all day long.

513

Yeah.

514

Plus you have to have two staffs for

both of them, and they have to be

515

able to revo- trust me, I've talked

to people who do this, and they keep

516

telling me it's the worst thing ever.

517

That's what they tell me.

518

I think it's to make me feel better.

519

But I'm going with it.

520

Brian Searl: Makes sense.

521

We'll go with it too, 'cause I don't think

I'm ever gonna have a private jet either.

522

Caleb, how do you use AI at Camplife?

523

Caleb Cook: Yeah.

524

It's something that we're really pursuing

from a perspective of asking, where can

525

we eliminate some of those day-to-day,

rough spots where people are just,

526

yeah, throwing time down a barrel.

527

And so we do have some tools that

we're working on in the back end

528

that are really gonna streamline some

of the staffing side of things too.

529

That's gonna be pretty

big here coming up soon.

530

Where we're at with AI, obviously

we're using it in the back end with

531

some of the development processes

as well to speed up our, go to

532

market and our velocity as well.

533

And yeah, I'm in that same boat of

just, asking AI a bit and learning

534

from AI and then just diving in too.

535

It's interesting 'cause I'm, this is

gonna be a little bit more personal

536

rather than a, a company statement per

se, but I'm a person that I, in some

537

ways I hate to love AI because there

are a lot of questions that I have about

538

it still from a societal perspective.

539

But looking at what it has enabled,

especially for, for some of these

540

smaller park operators and what we can

provide for them it's pretty exciting.

541

And yeah, definitely having made the

leap to the product side, I see that

542

a lot more clearly with what we're

able to start building out here.

543

It's exciting stuff.

544

But yeah.

545

Nick Purslow: I think we

share similar guilt, Caleb.

546

I'm the exact same.

547

I'm like, I don't like all of

this, but it's really good."

548

Yeah.

549

Caleb Cook: True.

550

True.

551

Nick Purslow: Yeah.

552

Caleb Cook: And I think I, I really

resonate with Brian, what you were

553

saying, and Mychele as, as well, that,

we're using this as the underlying

554

support to providing good hospitality.

555

And this is actually something that

came up in a conversation that we had

556

some pa- with some parks yesterday, was-

557

At the root of hospitality is

making people feel at home, and

558

you wouldn't feel at home with

somebody that you don't trust.

559

And so as we're using this AI,

toolbox we have trust to maintain

560

with our customers, with our campers.

561

And so making sure that, yeah, we're

not just using AI sight unseen, that

562

yeah, we utilize it, but we do run it

through our own human filters to say,

563

"Hey, does this sound like my park?"

564

"Does this preserve

trust with my customers?"

565

I think that's gonna be the big

question that comes up is yeah,

566

we know how to use this now.

567

Now how do we preserve trust

and preserve hospitality here?

568

Brian Searl: That's

with everything, right?

569

That's the same thing with your automated

phone systems that you wanna do- True

570

… with a branch and a tree and everything.

571

Does this preserve trust or

does this irritate my customer?

572

Yeah.

573

I think there's gonna be a lot of

people who do it wrong, and there's

574

gonna be some people who do it right.

575

Caleb Cook: Yeah.

576

Yeah.

577

Brian Searl: So hopefully we

can encourage the hospitality

578

of this, the outdoor hospitality

industry to mostly do it right.

579

Caleb Cook: Absolutely.

580

Yeah.

581

Yeah, 'cause

582

Mychele Bisson: At the end of

the day, it's a relationship.

583

I also think that it's part of the

presentation also, though, right?

584

Like if you have something, and I'm

just, this is just from my own personal

585

experience of what we've experienced

in coming into new parks, but it's

586

all in how you present whatever it is.

587

When we presented an AI answering system

for emergency, everybody freaked out at

588

first, and we were like this is to ensure

that, one, if it's really an emergency,

589

that we get the, to the right person.

590

But two, if you just need the code,

it's gonna answer a question without

591

waking up the manager who will then be

able to better serve you the next day."

592

And then it was, "Oh, that makes sense."

593

I think it's just all in presentation.

594

I think it's all in how

people perceive things.

595

When you get something brand

new, everybody freaks out.

596

Nobody wants change.

597

Nobody wants different.

598

Everybody goes to the worst case

scenario of, "Oh my God, robots

599

are taking over the world, and now

human-to-human contact is out."

600

And it's like, no, we're actually

doing this so that you can have

601

better human-to-human contact.

602

Ravi Parikh: Yeah.

603

Mychele Bisson: And that is the

story of, how you work with people.

604

It's just how you present it to

them, because if you're not telling

605

them the right story, look at

the Grimm Bror- Brother stories.

606

The originals suck.

607

They are horrible.

608

Disney remakes them, and

everybody falls in love with them.

609

So it's, that's what I think AI is about,

is it's just you've got to be able to

610

present it in a way that they can relate

to it and not think that they're taking

611

over the world and we're all gonna

be out of jobs and, subject to being

612

slaves to a machine all of the sudden.

613

Brian Searl: Yeah, I think it

goes back to the example like I,

614

I've cited on this show before.

615

First time we released our AI phone

agent, like years ago now, somebody

616

commented on a LinkedIn post and

said, "I will never wanna talk to

617

an AI ever on the phone, ever."

618

And my response to them was, "As soon as

you call your cable company and you aren't

619

have, you don't have to be transferred

for two hours to eight different people

620

to solve your problem, and you're off

the phone in 30 seconds, you'll never

621

wanna talk to a human being again."

622

Ravi Parikh: And actually SpaceX

80% of their customer service

623

is done through voice AI.

624

They're able to resolve 80%

of tickets through voice AI,

625

and that's a huge company.

626

I don't know how good it is,

but we implemented phone AI for

627

our support team, and we just…

628

If you wanna dial zero, it just

transfers you to a human, right?

629

So if you just give the customer a way to,

opt out until most people are generally

630

comfortable with these types of AI tools.

631

Eventually, I do think that, you

know- industry will get to the point

632

where for example, voice is on par or

as good as anything else out there.

633

Like I went to Taco Bell like a month ago

and ordered a Chalupa, and I was talking

634

to voice AI, and I was like, "Holy crap,

that was the most efficient Chalupa

635

ordering experience I've ever had."

636

And it didn't mess up my order.

637

And I was like, "That was awesome,"

because they always mess my order up.

638

And like there will be little aha moments

that people start to have around it

639

as the technology advances, and it's

getting better like insanely fast.

640

At this point, the AI is making itself

better because it's like writing itself.

641

Yeah I think over time it's gonna

improve all of these interactions, and

642

humans will start to understand like

what the right cases of it are and what

643

the wrong ca- use cases of it are for.

644

Brian Searl: Yeah, it's all

perception, like Mychele said.

645

But before we get to that, like you

didn't order lettuce at Taco Bell, right?

646

There was a problem with that.

647

Ravi Parikh: No.

648

Yeah, I didn't order lettuce.

649

Yeah.

650

Brian Searl: Okay.

651

All right.

652

So so the, yeah, it's all

perception with the guests though.

653

Like right now people have a perception

of chatbots as the stupid, horrible

654

things that you would yell and curse

at and just say human on the bottom

655

of a website since 2013 or '14.

656

They were never good.

657

Yeah.

658

And that's what most

people think a chatbot is.

659

The same way as most people think

an automated robot answering

660

system is the press one, press two.

661

It can barely understand you,

it can't transcribe your words

662

correctly, and you just want a human.

663

Eventually that's gonna flip,

and it's gonna happen, I think,

664

sooner than most technology does.

665

Yeah.

666

And when that does, then

that's gonna be better.

667

And again, to Mychele's point then

allow us to go back to focusing

668

more on the guest experience.

669

Ravi Parikh: We- And the- … we as

humans can't do everything, right?

670

There's just only so many of us, but the

AI is on 24/7, 365, and doesn't sleep.

671

So for, for simple things,

hopefully it can get pretty, pretty

672

effective at helping people out

673

Brian Searl: How do you think

it changes social in the future?

674

Like I know what you're using it for right

now, but as we see more of this, let's

675

call it AI slop for what it is, right?

676

As the attention gets pushed and pulled,

as people become more used to things

677

not being real or not being able to tell

they're real, how does that impact how a

678

property does social media in the future?

679

Nick Purslow: I honestly don't know.

680

I've been noticing more and more

Instagram reels with AI models.

681

And my first thought when I see it as

a consumer is like, "That looks crap."

682

But the numbers are doing pretty well.

683

I don't know, people don't seem to care

or, but the engagement's pretty good.

684

So it wouldn't surprise me if, for

example, properties just, and obviously

685

the technology's gonna get better,

so it wouldn't surprise me if, for

686

example, you don't need models anymore

for a photo shoot or a content shoot.

687

At the minute I just wouldn't do

it just 'cause I think it looks

688

so much better with real people.

689

But given we're so early and

people … i've seen people using

690

it and they're engaging pretty well.

691

That's one immediate thing

where I think, and then-

692

Brian Searl: My worry, my worry is the

depth and authenticity of it, right?

693

Yeah.

694

So this is my worry.

695

Like right now it's not fun for me, but

apparently it's fun for some people to

696

scroll through the AI slop and leave

and react and do all the things, right?

697

But when we get to the point where people

are starting to use this with their

698

campground, and I saw KOA do this, like

AKOA, not corporate KOA, but I saw AKOA

699

do this the other day where they posted

a picture of their train and it was

700

clearly AI 'cause there's neon lights

on the cur- on the thing at night and

701

they made it look like this really cool

experience and I don't think they were

702

intentionally trying to mislead anybody,

but there were a lot of people that were

703

like, "Hey, I live in the local area.

704

Can I bring my son or

daughter to come to that?"

705

And they're expecting something

I think- Yeah … rightfully

706

or wrongfully that what isn't

actually what they're going to get.

707

Even if it's a small difference-

Yeah … like neon lights.

708

So that's what worries me.

709

Ravi Parikh: I think there's a, … I

listened to a podcast called The All In

710

Pod and a couple months ago there was

a topic where they were talking about

711

Congress is Potentially thinking about

passing a law that identifies on social

712

or graphically generated content that

you have to say that this is AI, right?

713

Or like there's like a watermark

or something so people know

714

that this is AI generated versus

if it's real or versus AI.

715

And I think that's a good thing

to know whether what you're

716

looking at is AI generated or not.

717

Because it's starting to get

kinda hard to figure that out.

718

Like Nick said, when you're scrolling

through L- Instagram reels, it's

719

"Is this real or is this not real?"

720

And that makes fake news kinda scary.

721

Yeah.

722

Nick Purslow: I think

it is tricky as well.

723

Say if you're, t- using AI to create

models like, human models, that's not…

724

You're not giving a false impression of…

725

It's not like you're, generating a, a

brand new unit that does it all like that.

726

It's just literally just

filling that unit with people.

727

But yeah, I'd, I don't know.

728

I'd be interested to see how it goes.

729

I'm su- I'm sure we're gonna

see it getting a lot better

730

and a lot harder to tell.

731

Ravi Parikh: A- Anthropic actually

started putting watermarks

732

inside, like hidden watermarks in

anything that's AI generated text.

733

So if you use Claude to, write any copy on

your website, there's some hidden metadata

734

in there that says, "This was generated

by AI," which I do think Google's gonna

735

start picking that up and maybe giving

SEO penalty for AI written content,

736

but TB- TBD, like we don't know yet.

737

Brian Searl: Yeah.

738

I come down a little bit in the

middle on both of this stuff, right?

739

Because I don't think anybody should

be misleading their guests, but

740

I also think there's a benefit to

specifically what you were talking

741

about, Nick, like the model thing.

742

A small campground can't afford

to pay models to come into their

743

thing and do a photo shoot.

744

And so if you can take a real photo,

and we've done this like years

745

ago with we did it as an example

for one of the CRR properties.

746

We never used it online, we just, you

played around with Gemini and showed

747

it, where like you had a, a beautiful

picture that maybe a professional

748

photographer had taken years past

of their fire at night around the

749

pool, and there's just nobody there.

750

And so you just insert

a couple people there.

751

I don't know if that's misleading.

752

I think it's a fine line, because

like the, the models that you

753

would normally shoot aren't gonna

act that way for the guest anyway.

754

So is it a benefit or is it a harm?

755

I don't know.

756

I think I'm in the middle.

757

Nick Purslow: Yeah.

758

I think it's-

759

Caleb Cook: So a couple of different stor-

760

Nick Purslow: Go ahead.

761

Caleb Cook: Oh, sorry.

762

Go for it, Nick.

763

Nick Purslow: No, you go.

764

You go.

765

Caleb Cook: A, a couple of

different stories on this note.

766

There was actually a a retail company

that used AI generated product

767

photography with models and everything.

768

And they found that their social

media interaction rate and their

769

purchases tanked after using it.

770

Now on the flip side of

this coin there was a-

771

Brian Searl: Hold on.

772

Hold on one second.

773

'Cause did their customers know it was AI?

774

Caleb Cook: They were they did not,

it was not called out to them, but-

775

Brian Searl: Okay

776

… Caleb Cook: the customers were starting to

call it out as, "Oh, I see something that

777

makes me believe this is AI generated."

778

Brian Searl: But that, but

that's the reason though.

779

It's not if you do quality good AI, right?

780

I don't know.

781

Caleb Cook: Yeah.

782

And presumably, like I'll

give them the benefit of the

783

doubt and say it was good AI.

784

But I think we have a- A market

now that is on alert for AI.

785

And on the flip side of the story is

there's a professional photographer

786

that posted some ph- some food

photography that they did for a

787

client, and somebody on their socials

was like, "That's definitely AI."

788

And so she was, like, posting all

these process photos of saying,

789

"No, this is how I made this."

790

And so we're in this weird space of

everybody's, everybody's guard is up

791

for AI content for good and for worse,

so I tend to, yeah, fall in that,

792

that midground with you too, Brian

from the perspective of saying let's

793

do what," you, in a sense Photoshop

did to photography, years ago.

794

"Let's use reality and help people to

see themselves in that scene more."

795

And some people will like that and…

796

Finish.

797

Sorry, go ahead.

798

Please finish … the question always

is too, is yeah what's your market?

799

And what's interesting with the

retail brand is it was primarily

800

Gen Z and late millennials that

were the target market for that.

801

And these ones are like, they're

like hardcore calling it out.

802

And so that's a question for

us too, is who are we trying

803

to reach with our AI content?

804

And that may answer

some big questions too.

805

Nick Purslow: There, there's a, a

certainly a real anti-AI movement, I think

806

particularly kind of Gen Z in particular.

807

For example, like one of the

things that I would just never,

808

ever use it for is music.

809

I just would never listen

to AI-generated music.

810

I think that's a pretty hard line for

me, and there are so many other people

811

that are saying if you go on, if you're

looking for a YouTube kind of DJ set

812

or whatever you'll pretty quickly

find s- in the titles, a lot of them

813

have, no AI or no AI slap, whatever.

814

And there's a clear market for

people who are actively against AI.

815

And so I'm wondering, even almost as like

a counter-positioning marketing tactic,

816

particularly if you are seeking kind of

younger demographics is, almost be like,

817

we will ne- yeah, we will never use AI.

818

I think that can, that will really

resonate with a select group of people,

819

but I'd be interested to see whether

that grows, over time with, tech

820

bros getting more and more influence

and power, whether that, whether

821

that, that movement will grow or

whether it'd be a flash in the pan.

822

But I would…

823

Yeah I don't think it's necessarily

gonna be automatically accepted by

824

all generations that this is just

the way it is, even if that's the way

825

it's been painted a, a little bit.

826

So I'm curious to see how that plays out.

827

Brian Searl: Yeah, I think you're always

gonna have to find a balance, right?

828

'Cause like I, this is not a new argument.

829

There's always going to be

people who will, for AI, and then

830

we'll go historically, right?

831

There will always be

people who are anti-AI.

832

There will always be

people who are for AI.

833

I think most people end up

landing somewhere in the middle.

834

But this is an age-old argument.

835

I remember working…

836

We used to work on trade in

837

Daytona Beach called Dream Inn.

838

And back then was my first exposure,

like we were doing, like we were

839

actually going around taking pictures

of campgrounds and hotels at that point

840

with, I had a DSLR and whatever else.

841

And he was the first person who

brought to my attention "I have 50%

842

of my guests who don't like seeing

anybody in the pool photos because

843

they can't envision themselves there

with somebody else in the pool scene."

844

But that, but it, but then there's

the 50% of the other people who are

845

like, "I want people in the photos

to see how happy they are and how

846

they interact and everything else."

847

So there's always going to be a push

and pull with everything, I think.

848

I don't know that ever gets solved.

849

It maybe gets solved for your

specific park, your demographic, who

850

are your users, what do they want?

851

How do you make them happy?

852

Caleb Cook: It

853

all comes back to that

knowing your camper well

854

Brian Searl: Matt, how do you do this?

855

'Cause you've probably got a bunch

of different types of campers at all

856

your, in all your portfolio, right?

857

Mychele, you, if you wanna

take that after this.

858

Matt Whitermore: Yeah.

859

It's a really interesting question.

860

I am on the side of good content is good

content, whether it's written content

861

or whether it's an AI-generated image.

862

We're, I think we're at the point

where the quality of the image models

863

and the video models out there if

someone's gonna take the care to, to

864

generate AI im- AI-generated videos and

photos, w- we can all be fooled, right?

865

I- if they're really taking the time to do

it well, none of us would have any idea.

866

What gets me is the lazy use of AI, right?

867

It's so funny, I, there's this I'm

in Syracuse, New York, and there's

868

this local Instagram account called

Stop the Slop 315, which is, … is

869

our area code here, and they just

they're, like, kinda mean, right?

870

They they single out local restaurants

that have the classic AI posters, right?

871

And they're horrible.

872

But that's a mom-and-pop restaurant

that doesn't have high margins,

873

probably doesn't have a lot of revenue.

874

They don't have a marketing agency.

875

They don't have a graphic designer.

876

And they just go on their free

version of Chat GPT, and they give a

877

one-sentence prompt that says, that's

probably copying another AI poster,

878

and says, "Make me a version of this.

879

Here's my menu, and here's my website."

880

And, it's, there's levels to this and

there's t- everybody's in a different

881

position, and that's probably somebody

who's, I'm gonna, stereotype a little

882

bit here, but a little bit older,

a little bit less tech inclined.

883

Brian Searl: But even that,

but even that person there's a

884

benefit to that, isn't there?

885

Because I can imagine…

886

I remember traveling, I used to travel

300 days a year when we were building

887

this company, and we used to go to so

many restaurants who had no menus, who

888

had no online booking, who had no…

889

And so even if it's AI looking

to us, doesn't, isn't that

890

helpful to their guests?

891

Matt Whitermore: In that case,

the alternative is they probably

892

weren't gonna post anything, right?

893

'Cause they weren't gonna

have the time to do it.

894

When I look at Unhitched in our

portfolio, this, it's funny, this

895

has been an internal debate, right?

896

The, the internal stance was never any AI

generated images, and we've never done it.

897

I am on the other side of that argument

though, 'cause the way I look at it is

898

the w- like the way I would approach

that is let's go have a professional

899

photographer go photograph and video and

do the 360 video, and then you give that

900

to your a- your image model, and then

you have an unlimited real loop of this

901

is what my property actually looks like.

902

This is what our guests actually

look like, and you iterate it.

903

And the benefit there is that I don't have

to go send a professional photographer

904

every season and every year and every

time the leaves change or whatever, right?

905

You're not totally cheating, you're

not skipping all the hard steps, but

906

you're making it a lot more scalable.

907

And right, I would approach

it from the standpoint…

908

And it's not to fool people, right?

909

It's it's that good content

is good content and authentic

910

content is authentic content.

911

Whether or not that first draft was

written by AI, I could care less.

912

Ravi Parikh: I think if you identify

that it's AI, people will be a

913

little more disarmed by that.

914

It's like it's when you try

to- Totally … fool them that

915

people get frustrated about it.

916

But if you're like, "Hey, this is

AI generated," and we acknowledge

917

it, then it's like, okay we know

that this is AI, so we kinda get it.

918

Matt Whitermore: I mean, we

all see the, we all see the

919

content on LinkedIn and, right?

920

Yeah.

921

People screenshot it and make

fun of it, and I laugh at that.

922

And it- it's right, it's not about the

use of AI, it's about the lazy use of AI.

923

Yeah.

924

The short prompts- Yeah, because-

… the not giving it the context.

925

And I think we're in a world where

a few year- may- even today, maybe

926

a few r- year- a few, years from

now, 95% of the content you read

927

online is gonna be, there's gonna

be AI involved at some point, right?

928

Most of the stuff that I write starts

off as an AI draft, and I, the way I

929

look at this is it gives me unlimited

capacity to generate that first draft,

930

and then I can spend all my time on

refining it and making it authentic and

931

putting the real human stuff into it.

932

The alternative was I was gonna wake

up at 5:00 AM and stare at a blank

933

screen and a blinking cursor, and

it was gonna take me an hour to just

934

type those fir- first, two sentences.

935

Now, have unlimited drafts and un-

unlimited material I can read through,

936

and I can pull that, and then I can

say, "Okay, this is halfway there.

937

This is directionally I like this, but…"

938

And then you're an editor, right?

939

You're an orchestrator and

an editor, not a drafter.

940

Same thing with, right- editing photos.

941

I laugh, right?

942

Do you do you get offended when,

you're looking at somebody's Instagram

943

photos and then it's edited, right?

944

Is that not authentic, right?

945

Oh, it's not just the raw photo?

946

Like where do we draw

the line with this stuff?

947

I think it's really super

interesting conversation.

948

Brian Searl: Yeah, I think we're all

gonna figure out that line, I think

949

we're all gonna figure out that line

together as society moves forward.

950

I don't know if there's a right or

wrong answer, because I'm in Ravi's camp

951

partially, where our AI phone agents

that we have, we tell everybody, "You

952

should have your AI phone agent answer

the phone and say that they're an AI."

953

We should disclose that.

954

And then I'm with you on some of the s-

the slop on LinkedIn should be disclosed

955

as AI or regulated or however we wanna

do it, whether that's by LinkedIn

956

or the government or whatever else.

957

Probably by LinkedIn would be my

preference, smaller government and all.

958

But but then there's some things where

I'm like, "Do I need to disclose this?"

959

If I'm not intentionally misleading and/or

I'm stepping back, maybe this is a one-two

960

punch and I'm stepping back and saying,

"Could this potentially be misleading?"

961

Maybe I even run that through an AI

and ask "Hey, I edited this AI photo.

962

Is this misleading to my guest?"

963

If I'm just adding those people, like

I was talking about, near a pool in

964

a, at their ranch, and again, they

never use this photo publicly, right?

965

But if I'm doing that and the pool

furniture's the same, the colors are

966

the same, the fire didn't change, the,

the nighttime didn't change, the moon is

967

still aligned, the trees are the same,

everything is exactly how that customer's

968

gonna see it except for that lady is not

gonna be there when they get there, which

969

wouldn't be there in a model shoot anyway.

970

Should I be disclosing that's AI?

971

I don't know.

972

I don't think so, but that's me.

973

Mychele Bisson: I think if everything

is not misleading and everything

974

is exactly the same, I don't

think that's misleading at all.

975

We do a lot of organic content, so

everything we do is shot in the park

976

with our people with activities going on.

977

So all of our social media, pictures,

videos, all of that stuff is organic.

978

And we just felt like it was better

that way because we wanted people to

979

actually get the real experience and to

feel what our guests were feel- Like,

980

we did a shoot with some of our guests

a couple weekends ago in one of our

981

parks where they, the kids took the

GoPro and went down the water slide

982

with it, and they did an amazing job,

and we're gonna reuse that forever

983

because it's just a- amazing shots.

984

So I think, if I could…

985

I guess I would reuse that in AI.

986

If I could redo it in there, I would use

it if it was the same exact information

987

and the same exact everything except

the people, 'cause those people, like

988

you said, aren't gonna be there anyway.

989

But we do a lot of organic

stuff, and it does really well.

990

Our social media my social media alone

is, does really well just because I feel

991

like it is organic and I feel like I'm

taking everybody on a journey with me.

992

Brian Searl: And I'm not making an

argument where we should default to AI.

993

We should default to human, but

if you don't have a budget or you

994

wouldn't otherwise post in the case

of a restaurant, or you have a great

995

photo and there was no people in

it and you need people in it for…

996

Then that's where I'm

saying there's a little bit-

997

Mychele Bisson: Yeah

998

… Brian Searl: of flexibility.

999

Caleb Cook: 100%.

00:49:03,020 --> 00:49:05,390

Mychele Bisson: No, I think- I mean-

I think I agree with you in the side

00:49:05,390 --> 00:49:09,280

of if nothing is changing and it's not

misleading and it's exactly the way that

00:49:09,280 --> 00:49:11,110

it would look if they were there, I…

00:49:11,280 --> 00:49:12,260

Totally for it.

00:49:12,470 --> 00:49:15,540

Like I said, I would take that same

footage that we have of the kids and

00:49:15,830 --> 00:49:19,270

redo it for years with different people

in it or different things as long as

00:49:19,270 --> 00:49:23,010

everything looked exactly the same,

just change out maybe the people.

00:49:23,400 --> 00:49:27,180

I just- I'm on the side of I just

don't want anybody to look at it

00:49:27,180 --> 00:49:28,910

and be like, "That is not there."

00:49:29,390 --> 00:49:30,570

Because they call it out.

00:49:30,620 --> 00:49:30,650

Yes.

00:49:30,920 --> 00:49:34,300

We had a weird shot of one

of our one of our marinas.

00:49:34,630 --> 00:49:37,780

We had a shot that people don't usually

see 'cause it was from the water, and it

00:49:37,780 --> 00:49:39,510

was just the way that it looked at it.

00:49:39,630 --> 00:49:42,710

It even fooled me, 'cause I was

like, "That's not our marina."

00:49:43,060 --> 00:49:46,860

And so I'm looking at it trying to

figure out if they maybe took one of

00:49:46,860 --> 00:49:50,510

our other marinas and put it into the

picture and they messed it up, and

00:49:50,510 --> 00:49:53,880

I'm looking at it and I'm looking at

it and I was like, "That's not it."

00:49:54,190 --> 00:49:56,180

And so then somebody else was

like, "No, that's that…"

00:49:56,250 --> 00:50:00,400

And then somebody else posted, a guest

posted a picture underneath the same thing

00:50:00,410 --> 00:50:01,900

and they were like, "This is the marina.

00:50:01,900 --> 00:50:03,470

This is where the angle came from."

00:50:03,470 --> 00:50:05,290

And I was like, "Okay, they

saved me," 'cause I didn't even

00:50:05,290 --> 00:50:06,480

realize it was my own marina.

00:50:06,850 --> 00:50:10,840

So I think, like that I…

00:50:10,910 --> 00:50:12,070

Like, when people get…

00:50:12,140 --> 00:50:14,710

'Cause they do, they

nitpick, and they're…

00:50:14,950 --> 00:50:16,320

'cause they want the same experience.

00:50:16,320 --> 00:50:19,080

They wanna know "This is what I'm gonna

see," and they will pick it apart if

00:50:19,080 --> 00:50:21,330

they realize it's not the same picture.

00:50:21,540 --> 00:50:24,840

'Cause there were like 20 people on this

thread going, "This is not that marina.

00:50:24,860 --> 00:50:26,190

I don't know where this came from."

00:50:26,430 --> 00:50:29,390

And even I was fooled by it because

of the angle that it came from,

00:50:29,390 --> 00:50:31,760

but then somebody was like, "I'm

literally sitting in the spot that

00:50:31,760 --> 00:50:32,990

they shot the picture and here it is."

00:50:32,990 --> 00:50:38,760

And so that, that saved that post, but

it's just if you change anything, I

00:50:38,780 --> 00:50:42,810

think everybody is on high alert because

everybody wants to be the detective

00:50:42,810 --> 00:50:44,500

and be like, "That's not real," right?

00:50:44,930 --> 00:50:48,980

And so- Yep … I think that's where

the problem lies is just when it

00:50:48,980 --> 00:50:52,610

doesn't look the way that it should or

if you add something or make it look

00:50:52,630 --> 00:50:57,430

different, I can go out and AI myself

into looking like Christie Brinkley.

00:50:57,500 --> 00:51:00,320

Everybody's gonna be like, "That's not

her," 'cause they're gonna see me in real

00:51:00,320 --> 00:51:02,250

life and realize that's definitely not me.

00:51:02,260 --> 00:51:03,370

I wish, but no.

00:51:03,710 --> 00:51:06,480

It's just, it's those things that

people nitpick because they're

00:51:06,480 --> 00:51:07,430

gonna say, "That's not what it is."

00:51:09,174 --> 00:51:10,394

Nick Purslow: I think you've

always gotta make allowance for-

00:51:10,394 --> 00:51:12,604

Mychele Bisson: I may have just aged

myself because I don't know if everybody

00:51:12,604 --> 00:51:13,894

knows who Christie Brinkley is.

00:51:13,894 --> 00:51:15,057

I just realized that.

00:51:15,057 --> 00:51:16,034

Brian Searl: I know Christie Brinkley.

00:51:17,694 --> 00:51:19,864

Nick Purslow: I, you gotta make

an allowance for there's always

00:51:19,864 --> 00:51:22,214

gonna be crazy people in the

comments regardless as well.

00:51:22,264 --> 00:51:24,674

Like we did a pre-sale campaign

where we just used like

00:51:24,814 --> 00:51:26,434

artist impression renderings.

00:51:26,574 --> 00:51:29,954

It wasn't AI, it was just,

traditional kind of animation.

00:51:29,954 --> 00:51:32,804

And people were like

beating us up for that.

00:51:32,804 --> 00:51:35,034

And it's like right now the

property's an, an empty lot.

00:51:35,134 --> 00:51:38,154

And this is what we're gonna build

and it's quite clearly, animation.

00:51:38,644 --> 00:51:41,574

And so some people you just, I think you

gotta accept that, that there's gonna

00:51:41,574 --> 00:51:44,374

be no convincing them and that there'll

always be crazy people out there,

00:51:44,474 --> 00:51:45,664

Mychele Bisson: oh, there's

always crazy people-

00:51:45,674 --> 00:51:45,784

Brian Searl: All right.

00:51:45,784 --> 00:51:45,824

Let's-

00:51:45,824 --> 00:51:47,064

… Mychele Bisson: out there

trusting all the time.

00:51:49,274 --> 00:51:50,041

Nick Purslow: I'm just gonna

give you a hundred percent-

00:51:50,041 --> 00:51:51,584

Brian Searl: Let's spend the last

few minutes of the, let's spend

00:51:51,584 --> 00:51:53,354

the last few minutes of the show

asking each other questions.

00:51:53,404 --> 00:51:55,144

Matt, do you have any

questions for anybody here?

00:51:55,824 --> 00:51:58,204

Just a que- question for one

person that you wanna know?

00:52:00,734 --> 00:52:01,444

Matt Whitermore: Oh, that's a good one.

00:52:01,514 --> 00:52:03,154

Nick, we haven't caught up in a while.

00:52:03,254 --> 00:52:05,744

What's new in the land of Posh Outdoors?

00:52:05,744 --> 00:52:07,014

What's what's been going on?

00:52:07,444 --> 00:52:08,214

Nick Purslow: Lots going on.

00:52:08,364 --> 00:52:11,704

We're always in cat raise mode,

but we're really ramping up now.

00:52:11,704 --> 00:52:15,874

We wanna get to 2, 3, 4

locations by mid next year.

00:52:16,404 --> 00:52:19,414

And I've just been deep in the weeds

of social media marketing and really

00:52:19,464 --> 00:52:23,024

figuring out that code and how to

speak to people through social media.

00:52:23,024 --> 00:52:25,184

So lots going on on, on my end.

00:52:25,184 --> 00:52:27,024

Curious to know what's going

on, on, on your end too.

00:52:29,752 --> 00:52:30,752

Matt Whitermore: Yeah, no, appreciate that

00:52:30,802 --> 00:52:31,642

Nick Purslow: what's going on in yours?

00:52:32,332 --> 00:52:33,092

Matt Whitermore: Just growing fast.

00:52:33,122 --> 00:52:35,832

Yeah … we're ho- hopefully some

exciting announcements soon with

00:52:35,832 --> 00:52:38,562

some, some big adds to the portfolio.

00:52:38,612 --> 00:52:40,272

Management side, investment side.

00:52:40,322 --> 00:52:41,402

Been keeping busy.

00:52:41,452 --> 00:52:46,672

Try- trying to scale, with a

corporate crew of about 20 in

00:52:46,692 --> 00:52:48,012

headquarters has been tough.

00:52:48,112 --> 00:52:51,392

Doing a lot of AI stuff to- … get

more efficient behind the scenes.

00:52:51,392 --> 00:52:53,892

So that's been my my day-to-day focus.

00:52:54,632 --> 00:52:54,922

Nick Purslow: Nice.

00:52:56,222 --> 00:52:57,402

Brian Searl: Nick, do you

have a question for anybody?

00:52:58,112 --> 00:52:59,012

Nick Purslow: That was my question.

00:52:59,502 --> 00:53:00,209

Brian Searl: Oh, you're gonna take it

00:53:00,209 --> 00:53:00,906

Nick Purslow: That's what I was gonna ask

00:53:00,906 --> 00:53:01,552

… Brian Searl: take it easy on them?

00:53:01,552 --> 00:53:01,912

Okay, all right.

00:53:02,962 --> 00:53:05,622

Nick Purslow: I, r- it'd be good

to know how, what Ravi's best use

00:53:05,622 --> 00:53:08,022

case is for AI given, your position.

00:53:08,022 --> 00:53:08,892

Ravi Parikh: Oh, best use case.

00:53:09,452 --> 00:53:17,832

I don't know what the best use case

is, but I pretty much wrapped our AI

00:53:17,832 --> 00:53:26,212

around, like every problem I face,

I try to think of it AI first now.

00:53:26,772 --> 00:53:31,572

Something cool we've built at RoverPass

is this concept called SETT Com.

00:53:32,402 --> 00:53:35,472

We call it SETT Com because I

think central command sounds cool.

00:53:36,012 --> 00:53:42,312

But it's basically connected into

every part of our business HR, finance,

00:53:42,322 --> 00:53:46,082

support, product, engineering, everything.

00:53:46,132 --> 00:53:53,142

And so like our whole team now essentially

can just ask questions and it can look

00:53:53,142 --> 00:53:56,532

at our code base and answer questions.

00:53:56,532 --> 00:54:01,522

It can generate FAQ articles on the

fly based off how features work.

00:54:01,522 --> 00:54:06,112

So I don't know that I have the

best use case but there are just so

00:54:06,122 --> 00:54:12,572

many unlimited use cases that we're

constantly finding every single day.

00:54:12,692 --> 00:54:17,372

But like one thing that we've

started doing is identifying

00:54:17,382 --> 00:54:19,232

loops within our workflows.

00:54:19,242 --> 00:54:23,852

So it's where, let's say like we

implemented voice AI in our support.

00:54:24,082 --> 00:54:28,432

When a campground owner calls us, if for

some reason we can't pick up the phone or

00:54:28,432 --> 00:54:31,302

it's after hours, it defaults to voice AI.

00:54:32,332 --> 00:54:37,332

And the problem was is that we didn't

have training data to train the voice

00:54:37,382 --> 00:54:42,272

AI on like what to do and what not to do

because we didn't use to record our calls.

00:54:42,282 --> 00:54:47,832

So it's how do you train AI on an

infinite number of possibilities?

00:54:47,832 --> 00:54:52,202

If you're like, somebody asks you what

the Cowboys score is on a call, it's

00:54:52,202 --> 00:54:54,372

like like what do you do in that case?

00:54:54,382 --> 00:55:01,432

So we've got like AI agents that grade

the call after the call based off how

00:55:01,432 --> 00:55:08,772

it's supposed to run and then it fixes

our prompt based off of what that other

00:55:08,802 --> 00:55:13,402

AI agent thinks the voice AI agent

screwed up and then if it doesn't have

00:55:13,402 --> 00:55:18,442

an FAQ article to answer the client's

question, it automatically creates the

00:55:18,442 --> 00:55:24,772

FAQ article and queues it up for our

support team to approve that FAQ article.

00:55:24,782 --> 00:55:29,382

So looping is the next

frontier of AI, I think.

00:55:29,382 --> 00:55:33,872

It's like identifying these like

recursive loops in your workflow and

00:55:33,872 --> 00:55:38,492

then having AI improve those so you

can start to compound from there.

00:55:38,512 --> 00:55:43,330

So just I don't know if that answers your

question directly, Nick, but there's just

00:55:43,330 --> 00:55:47,270

so many use cases that we're finding,

so I don't know if there's any one best

00:55:47,900 --> 00:55:50,960

Brian Searl: The best is the best for you,

and th- and that's what you gotta start

00:55:50,960 --> 00:55:52,328

your journey like we talked about, right?

00:55:52,328 --> 00:55:52,485

Yeah.

00:55:52,485 --> 00:55:52,643

Yeah.

00:55:52,643 --> 00:55:54,060

Start your journey with something small.

00:55:54,460 --> 00:55:54,660

Yeah.

00:55:54,660 --> 00:55:57,100

And you'll figure your best will be

something today, and then it'll be

00:55:57,100 --> 00:55:59,150

different tomorrow, it'll be different

next week, it'll be different next

00:55:59,150 --> 00:56:00,100

month, it'll be different, right?

00:56:01,030 --> 00:56:01,360

So-

00:56:01,970 --> 00:56:04,460

Ravi Parikh: And the crazy thing is

you couldn't even do half of this

00:56:04,460 --> 00:56:10,060

stuff until December of last year

when Anthropic released Opus 4.5.

00:56:10,190 --> 00:56:15,310

AI wasn't as anywhere near as good

as it was until like late last year.

00:56:15,360 --> 00:56:17,900

So it's crazy how fast this is all moving.

00:56:19,280 --> 00:56:20,510

Brian Searl: Caleb, do you

have any questions for anybody?

00:56:20,790 --> 00:56:23,340

And if anybody needs to drop off,

I know we're a little bit over, so-

00:56:23,690 --> 00:56:24,780

Nick Purslow: I've gotta

dash to another call.

00:56:24,810 --> 00:56:25,390

Thank you, Brian.

00:56:25,400 --> 00:56:25,430

Okay.

00:56:25,440 --> 00:56:26,650

Nice to meet you all.

00:56:26,650 --> 00:56:26,770

Thanks for your time.

00:56:26,770 --> 00:56:26,860

Thank you.

00:56:26,860 --> 00:56:26,880

Yeah.

00:56:26,880 --> 00:56:28,810

Caleb Cook: All right.

00:56:28,810 --> 00:56:29,140

Yeah.

00:56:29,350 --> 00:56:33,130

I'd be curious to Mychele here, just

how, … I know it's been a little bit

00:56:33,130 --> 00:56:37,330

of a unique season across the industry,

and so I'm curious how marketing has

00:56:37,330 --> 00:56:42,790

gone to recurring guests or past guests

and how they've responded to that.

00:56:44,130 --> 00:56:46,070

Mychele Bisson: Actually,

we had a great season.

00:56:46,100 --> 00:56:48,490

We started off really slow because

of the, I think it was mainly

00:56:48,490 --> 00:56:49,800

the weather and the gas prices.

00:56:50,100 --> 00:56:53,400

But we pivoted to marketing closer

to home and actually booked out

00:56:53,400 --> 00:56:54,690

our parks most of the season.

00:56:54,950 --> 00:56:58,210

I think for July 4th we were

like at 106% across the board.

00:56:58,720 --> 00:56:58,750

Nice.

00:56:58,760 --> 00:56:59,100

Yeah.

00:56:59,100 --> 00:57:00,360

So we did really well.

00:57:00,360 --> 00:57:03,330

August is always probably our slowest

month because everybody goes back to

00:57:03,510 --> 00:57:07,840

school, and so I was just talking to

my team right before this about, "Okay

00:57:07,840 --> 00:57:10,390

let's give all the locals a bigger

discount just to get them all in."

00:57:10,390 --> 00:57:11,890

And we're doing like f- first responders.

00:57:11,890 --> 00:57:15,060

Like this is when we do all of that

stuff so that we can just get people and

00:57:15,060 --> 00:57:18,230

heads in beds because we know August and

September everybody goes back to school.

00:57:18,530 --> 00:57:22,500

So I think it's always just understanding

who is where and who's coming to

00:57:22,500 --> 00:57:26,360

the park at what times to understand

how to switch your marketing up.

00:57:26,390 --> 00:57:31,010

But I think because we are a smaller

group we're able to pivot really quickly.

00:57:33,926 --> 00:57:34,336

Caleb Cook: That's awesome

00:57:34,336 --> 00:57:34,556

Brian Searl: All right

00:57:35,786 --> 00:57:36,046

Mychele Bisson: Yeah

00:57:36,046 --> 00:57:36,776

Brian Searl: Mychele,

you have any questions?

00:57:37,376 --> 00:57:39,154

Mychele Bisson: I think, I think

we're probably done for the day

00:57:39,154 --> 00:57:39,536

Brian Searl: You lost everybody

00:57:39,536 --> 00:57:39,736

Mychele Bisson: Yeah

00:57:39,736 --> 00:57:41,466

Brian Searl: You, you lost everybody

but like the other day was the day-

00:57:41,466 --> 00:57:43,516

Mychele Bisson: Actually, my question

was for Caleb anyway because I wanted

00:57:43,516 --> 00:57:47,216

to know, so with CampLife, how are you

guys integrating AI into your systems?

00:57:48,766 --> 00:57:49,116

Caleb Cook: Yeah.

00:57:49,556 --> 00:57:51,816

So there, there are a couple

of things we're working on

00:57:51,816 --> 00:57:53,136

the backg- in the background.

00:57:53,236 --> 00:57:57,006

I don't actually know how much I

can divulge on a couple of the-

00:57:57,006 --> 00:57:59,706

Brian Searl: Nobody watches the show,

Caleb, so just say whatever you want.