OpenAI Sora Breaks Internet! Society At Risk? (Ep. 110) - Counter Thought

Episode 110

full
Published on:

23rd Feb 2024

OpenAI Sora Breaks Internet! Society At Risk? (Ep. 110)

OpenAI broke the internet with the release of Sora, its video AI model. Sora is not the first video AI model, but will video AI break society and be another cautionary tale like social media platforms?

#sora #artificialintelligence #openai #counterthought #podcast #news #society #goodvsevil

----


Remember to Like, Comment and Share this video.

Remember to SUBSCRIBE to the Counter Thought channel (@counterthoughtpodcast).


----


AUDIO versions of podcast episodes are available on your podcast app.


VIDEO versions of podcast episodes are available on YouTube and Rumble.


  • YouTube: https://youtu.be/vLIpNiS-MVQ


  • Rumble: https://rumble.com/v4ew6oo-openai-sora-breaks-internet-society-at-risk-ep.-110.html


----


FOLLOW Counter Thought on social media:

Instagram: @counter_thought

Instagram: @counterthoughtceo

X (Twitter): @counterceo

TikTok: @counterthought

Facebook: "Counter Thought Podcast" page

Transcript
Speaker:

Welcome to Counter Thought,

2

:

a podcast conserving America's

freedom, culture and values.

3

:

This is Brian Kletter,

the creator and host of the podcast.

4

:

You can engage with the podcast

on Instagram @counter_thought

5

:

or @counterthoughtceo,

6

:

and on our Facebook page “Counter

Thought Podcast”.

7

:

For audio versions of the podcast,

you can find us on

8

:

Apple Podcasts,

Google Podcasts, Spotify, and more.

9

:

And for video versions of the podcast,

10

:

join us on YouTube at the Counter

Thought channel.

11

:

Let's go!

12

:

Open A.I..

13

:

The creators of GPT released

their video A.I.

14

:

model called Sora

and it broke the Internet.

15

:

And while its capabilities are incredible

and while the intentions are good,

16

:

I can't help but wonder, even though

it's breaking the Internet now,

17

:

our video A.I.

18

:

model is going to break society.

19

:

Welcome to Counterattack

20

:

Openai AI.

21

:

The creators of GPT released

a testing version of their video A.I.

22

:

model called Sora.

23

:

And when it did so, all of social media,

24

:

YouTube rumble, other different streaming

platforms

25

:

just went berserk over the capabilities

that Sora has.

26

:

Now, while Sora is not the first A.I.

27

:

video model, there are others out there.

28

:

There are many, actually,

29

:

who could create clips of podcasts

just like this in other YouTube videos.

30

:

Just text to video editing.

31

:

Text to video creation.

32

:

So open the AI in.

33

:

Sora is not necessarily breaking

new ground, but what they are doing,

34

:

like what they do with their GPT model

is they are,

35

:

I guess, just going above

where everyone else already is,

36

:

or at least that is the expectation.

37

:

Now you

38

:

might be wondering, okay, what is Sora?

39

:

Yes, it is a video model,

40

:

but what does that actually mean?

41

:

So what open A.I.

42

:

says, and this is from from their website,

43

:

Openai says this about Sora,

44

:

quote, We're teaching A.I.

45

:

to understand

and simulate the physical world in motion.

46

:

While the goal of training models

that help people

47

:

solve problems

that require real world interaction.

48

:

Introducing Sora Our text A video model.

49

:

Sora can generate videos.

50

:

As of right

now, up to a minute long while maintaining

51

:

visual quality

and adherence to the user's prompt.

52

:

So there's a lot there.

53

:

Basically, what they're saying

is that they're teaching it to understand

54

:

and simulate the physical world in motion,

55

:

and they have multiple videos

demonstrating the current capabilities.

56

:

And I'll share one or two of those

with you here in a few minutes.

57

:

But with the goal of training models

to help solve problems

58

:

that require real world interaction,

you could essentially say, okay, I need

59

:

I need a video of horses running across

an open prairie or something like that.

60

:

And instead of you

having to go hunt down on the Internet,

61

:

maybe some stock footage

and pay for that footage,

62

:

or get your own production

crew to go out and film that footage

63

:

and the horses and the land

and everything else.

64

:

You would just type it into Sora.

65

:

Video of horses

66

:

running across a prairie.

67

:

Video generated right.

68

:

Let me show you what exactly

69

:

I'm talking about

70

:

is. Isn't that incredible?

71

:

man, it's.

72

:

It is. It is.

73

:

Honestly, it is incredible.

74

:

The tech person in me just, like, loves

that this is even possible.

75

:

And I'm thinking, okay, for this podcast,

how can I use A.I.

76

:

video editing in So is not the first A.I.

77

:

video video tool.

78

:

There are others.

79

:

There are others out there.

80

:

One of them being a synthesizer,

I believe, is how you pronounce it.

81

:

And that is something that you can just

create clips that could take this

82

:

long form video of the full episode,

and it would just automatically generate

83

:

clips to be used

on, you know, in vertical format,

84

:

which on TikTok

and Instagram and Facebook.

85

:

Or it could create smaller clips and stay

in this landscape horizontal position

86

:

that I could then upload for, you know,

87

:

future feature clips for the channel

here on YouTube or on Rumble.

88

:

So the

capabilities are are very impressive.

89

:

And one of the things like like

90

:

Openai says is that it will allow you

the efficiency and the speed and cost

91

:

saving capabilities of not having to get

your own production crew or to contract

92

:

work with a production crew

to get what exactly you are trying to get.

93

:

Like a plane flying over,

snow capped mountains or something

94

:

like that on a crystal clear,

crystal clear blue winter sky.

95

:

Right.

96

:

You know that we won't have to go

and outsource that work

97

:

or pay people in-house to go do that,

which is could save hundreds of thousands

98

:

to millions of dollars, depending

on what exactly you're looking for.

99

:

And this is only going to be refined.

100

:

Open air says that themselves.

101

:

They talk about how,

you know this is a read testing

102

:

red team red teamers are testing

to see the capabilities and yes right now

103

:

not everything is going to be perfect

like you may say something

104

:

to text in there

have a woman biting a cookie.

105

:

Well, at this point,

the cookie may not show bite marks after

106

:

the woman bites into it.

107

:

But you know that will be improved

upon and then you will see

108

:

the bite mark in the cookie.

109

:

Right.

110

:

It'll look like a real woman or the woman

that was used to create this video

111

:

who signed some type of consent

form to use her likeness

112

:

eating a cookie.

113

:

But that didn't necessarily happen.

114

:

That didn't necessarily happen.

115

:

So there are definitely good uses

to video tools.

116

:

I just mention one synthesizer

I could record myself,

117

:

and then I think I could

118

:

then take the same video and automatically

make it make myself speak in Spanish.

119

:

Right.

120

:

Reach a larger audience with this podcast.

121

:

There are others out there

that can do automatic video clips.

122

:

They will go through

and read through, right?

123

:

Just give it the time

to read through the full podcast episode.

124

:

The full length, say it's 20 minute video

125

:

and then it'll determine

what are like the the wow moments.

126

:

You know that

the eye catching moments or takes

127

:

within the video

and go ahead and produce and recommend

128

:

clips for me then to utilize

like I said on other social media

129

:

platforms or even just shorter

clips here on YouTube or rumble.

130

:

Efficient video production, right?

131

:

Cost saving capabilities.

132

:

Now that also rubs people the wrong way.

133

:

Right.

134

:

And a lot of people are worried

about A.I., not just video A.I.

135

:

tools, but other other A.I.

136

:

tools. People using A.I.

137

:

right now going through through resumes

and creating better resumes and

138

:

and then also using AI to analyze

139

:

data and spit out,

you know, like an analysis

140

:

and so on and so forth, and synthesize

all of this information.

141

:

And I've I've heard and been told.

142

:

Right.

143

:

That we we as humans

should not be necessarily scared of A.I.,

144

:

but we should do is embrace A.I.,

145

:

embrace the tools,

and use it to your advantage.

146

:

And that eventually there's

going to be the separation because it's

147

:

going to become so prevalent across

society and across businesses

148

:

that there will be a

149

:

divide of those who adopted A.I.

150

:

and those who didn't.

151

:

Those of us individuals

who know how to use A.I.

152

:

and those who don't.

153

:

So there's going to be this separation

like either you're going to

154

:

to move forward with the rest of us,

or you are going to get left behind.

155

:

Now, specifically for video A.I.,

when you think about it, you're like,

156

:

okay, well,

157

:

if I don't have to then get a production

crew to go create this content.

158

:

Is that just going to replace everything?

159

:

Well, not entirely.

160

:

And open A.I.

161

:

and a synthesis say say similar things.

162

:

They're like, okay, well,

you would still need the production crew

163

:

initially to go create the stock footage.

164

:

Now, will that always be the case

165

:

as more and more stock

footage is created as A.I.

166

:

is able to generate

167

:

more scenes and everything by itself,

168

:

utilize the information

in the stock footage that it has.

169

:

Will that eventually eliminate an

individual actually going out and filming

170

:

something to then be used or replicated

or iterations or iterate from?

171

:

That remains to be seen.

172

:

And that is what scares

people like in the workforce.

173

:

Are you trying to take my job?

174

:

Is I going to replace humans

175

:

or is I going to improve

176

:

the lives of humans?

177

:

That is that is the question.

178

:

So on the flip side of the good

179

:

is obviously the bad.

180

:

The bad would be the usage

181

:

for disinformation

182

:

creating social hermits

183

:

because people are scared that

something is going to be created of them

184

:

or they will take down their social media

platforms in hopes that,

185

:

you know, the public profile

that they already had,

186

:

the images

that were already had, the content

187

:

that was already shared on

the social media platforms

188

:

are not going to be manipulated

and used to create disinformation

189

:

campaigns,

you know, used to smear someone.

190

:

We recently heard about Taylor Swift

and the video or the pictures and videos

191

:

that were created of her that, you know,

that she came out and vehemently

192

:

denied saying that those are not you know,

those are not real.

193

:

Those are fake.

194

:

Deep fakes have been

going on on for years now

195

:

in the form of video

196

:

and pictures.

197

:

And then what are we going to do

to fight against it?

198

:

You know,

199

:

we don't we don't have

200

:

millions of dollars, right,

to hire lawyers and go after companies

201

:

or individuals who could use this

this tool,

202

:

the video and air tools do

to create disinformation,

203

:

to make it seem

as if I said something that I never said.

204

:

One of the videos on on the Synthesis site

205

:

was with the CEO

and he was on Bloomberg UK, I think.

206

:

And the video segment

207

:

for that broadcast started with

one of the hosts of the program talking,

208

:

and then it cut to at the end of

the video, it cut back to the actual host.

209

:

And he's like, That wasn't even me.

210

:

That was an air generated video.

211

:

I didn't even say those things,

but it made it look like exactly me

212

:

and me saying those exact things.

213

:

So that is what we have to

214

:

to wrestle with an open air

215

:

in the release on their on their website

216

:

for sorry

they address a lot of these things they

217

:

they address and categorize it

into three different categories.

218

:

They have safety capabilities

in in research

219

:

and under safety

they state quote, we'll be taking several

220

:

important steps ahead of making PSAs

available in open air products right.

221

:

Again, they're going through red testing.

222

:

We are working with Red Teamers

domain experts

223

:

in areas like misinformation,

hateful content and bias,

224

:

who will be adversarial,

testing the model,

225

:

and then it continues.

226

:

Quote, We're also building tools

to help detect misleading content such

227

:

as the detection classifier that can tell

when a video is generated by Sora.

228

:

We plan to use

229

:

include C to make metadata in the future.

230

:

If we deploy the model in an open

air product and it goes on,

231

:

that's great.

232

:

But I'm not sure if you've noticed or not.

233

:

Things tend to go viral

234

:

before any type of fact checking occurs,

235

:

right?

236

:

That is one of the things

that has come from

237

:

maybe what was originally intended

to be good on

238

:

on social media platforms,

the sharing of information

239

:

that is now seen as as a negative, Right.

240

:

That's the definition of viral virals.

241

:

Viruses, Right.

242

:

Just just spread. They go everywhere.

243

:

And it's hard to contain

if you most likely can't even contain it.

244

:

And if you did try to contain it,

the amount of resources

245

:

that it's going to take for you

to contain it, which goes back to my point

246

:

about we're not all millionaires

hiring lawyers for people who,

247

:

you know, who libel

248

:

and smears

and things like that, defamation.

249

:

How are we actually going to contain

a video,

250

:

a deepfake video,

something created through open air,

251

:

even if it does have these parameters,

if opening air is able to

252

:

to infuse this within their tools

253

:

in the same with other video companies,

254

:

who's to say

that others are not going to use it

255

:

for malicious intent?

256

:

Right.

257

:

Where there's good,

there's also going to be to be bad.

258

:

That's true across all of society

with with everything.

259

:

Right.

260

:

So what exactly is going to happen

when it is you

261

:

when a video, a tool like Sora

262

:

is used to

to do harm to another individual?

263

:

I mean, think about it.

264

:

We're we're in peak

election season right now, right

265

:

about to get through

266

:

the primaries

and get into the general election, heading

267

:

looking ahead to November,

where we're going to be deciding between,

268

:

you know, Biden or Trump,

269

:

but who's to say that the

270

:

the campaigns

271

:

will create a video

that depicts the candidate

272

:

saying something that they never said

in an environment they were never in,

273

:

and how that how quickly that can go

viral, especially

274

:

especially in today's society,

right in today's media, where

275

:

where we see that

those in the left media pick and choose

276

:

and this is media

as a whole, pick and choose

277

:

what information it is

that you get to know about.

278

:

You turn on the news.

279

:

You only hear about stories

that they tell you about.

280

:

And we already see that

281

:

there has been

nefarious acts taking place in the media,

282

:

especially towards towards Trump,

283

:

but beyond Trump and going again

five, ten, 15, 20 years into the future,

284

:

who's to say that campaigns aren't

just going to be deepfakes

285

:

and we already have the

the commercials that come out

286

:

that are attacking

the other candidate, right?

287

:

Those

288

:

have like truth to them,

289

:

you know, more truth

than what a deepfake would be.

290

:

Then just creating, again,

291

:

placing a candidate in an environment,

seeing things that they never said.

292

:

And then what you and me

are going to sit there and try to analyze

293

:

the video and detect it's it's metadata

to see if this is or it's markers

294

:

and indicators to see if this is actually

a real video. No,

295

:

it's going to go viral.

296

:

It's going to contribute

to myths and disinformation.

297

:

Not only that, but

we've already have issues on social media

298

:

platforms of girls being depicted

299

:

doing sexual acts and just smearing them.

300

:

Right.

301

:

The high schoolers

we're talking about here

302

:

making it seem like a girl

is posted in nude videos

303

:

and everything and then just, you know,

trying to completely wreck her life.

304

:

Present day and her life in the future.

305

:

So call

306

:

me, call me cynical, call me negative,

call me pessimistic, but

307

:

I am very concerned about

308

:

where this technology could lead.

309

:

And the CEOs of these companies,

especially the one for

310

:

synthesis, said that

we're just going to have to be vigilant.

311

:

Yes, that is true.

312

:

I agree with that.

313

:

We will have to be to be vigilant.

314

:

But how exactly

315

:

is our vigilance going to stack up against

316

:

against what would be required

to actually combat the information?

317

:

And again, these the definition of viral

is it spreads.

318

:

Right. And you're not going to be able to

319

:

to correct whatever disinformation,

320

:

whatever harm was caused,

321

:

because you're not going to be able to get

everybody's eyes on that.

322

:

Saw the original malicious content

323

:

to be able to see that the update.

324

:

Right.

325

:

That says that

that was actually disinformation.

326

:

That was misinformation

that was false. That was not real.

327

:

And going back to social media,

328

:

social media right now should be,

I believe, a cautionary tale.

329

:

When Facebook began

and I mean, I was there.

330

:

It started in 2003.

331

:

I started college in the summer of 2004

when I first joined Facebook.

332

:

Facebook, you know the story.

333

:

If you haven't, go see the movie

The Social Network, the Facebook

334

:

began,

you know, off the of the existence of that

335

:

the yearbook, so to speak, at at Harvard

336

:

and we end with only being open

once it was released to the public

337

:

only being open

338

:

to college students

because you had to supply a edu email,

339

:

you had to have a dot

edu domain for your email.

340

:

And I was at the University of Florida

and I had that right so I could join

341

:

and it stayed like that for I want to say

that at least the first two years.

342

:

And then Facebook saw the potential

and it said,

343

:

okay, we're going to expand

this, We're going to go from

344

:

and dot edu domain

345

:

to open it up to like community colleges.

346

:

Right.

347

:

So you you know, it wasn't

just a four year school, the university.

348

:

It was also

then open to to two year colleges.

349

:

And then that expanded

to just the general public.

350

:

And now all hell broke loose.

351

:

Right.

352

:

As originally intended to be able

to to chat with each other,

353

:

to post pictures, to interact, you know,

the social aspect of social media,

354

:

but as well intended it was Facebook

355

:

and then Instagram and the Twitter

and then others.

356

:

We have seen

the negative effects of social media.

357

:

We have seen everything

from misinformation to disinformation

358

:

to smearing individuals to the

what is said to be,

359

:

you could argue, strongly causation,

but definitely correlation

360

:

between social media

and mental health issues in today's youth

361

:

youth ranging from, you know, ages

ten through, you know, early

362

:

to early twenties or you can stop it at 18

and then you get into early adulthood.

363

:

Right?

364

:

We have seen the negativity, the

365

:

the negative aspect of social media

366

:

and social media should be,

I think, a cautionary tale

367

:

for this.

368

:

This video I and other air tools.

369

:

The intentions were great

370

:

for Facebook when it started,

371

:

but as it expanded

372

:

and as more and more people got onto it,

373

:

the good intentions also brought

374

:

bad intentions.

375

:

And I fear that this is going to be

the same

376

:

for these different platforms,

these different A.I.

377

:

tools, especially video.

378

:

I've already getting comments right

379

:

on some of our most recent videos

because I've got larger reach

380

:

saying that

because I'm using a green screen

381

:

that my background looks fake,

which technically it is.

382

:

But saying that I'm a I right,

that I don't look real sitting here

383

:

in front of the screen screen.

384

:

Maybe I should use an A.I.

385

:

video tool to actually give me better,

better lighting and enhance my background

386

:

to where it's not just a flat image,

to where it has a 3D aspect to it,

387

:

but I'm already being said always

just age.

388

:

Just spouting things off was like a robot.

Just just fake, right?

389

:

Not even real. I'm real.

390

:

I'm real.

391

:

And yeah, those people are being stupid

and all of that.

392

:

But I can't help but see in 510,

15, 20 years

393

:

the cautionary tale

that we should have taken the lessons

394

:

learned from social media and apply it

to Openai, AI and other video A.I.

395

:

video tools

396

:

and much to their credit,

397

:

Openai and the other companies

talk about ethics and safety

398

:

because I think they do

see what has happened with social media

399

:

and they do see the negative things

that could be that

400

:

that these tools,

these models could be used for.

401

:

So I give them credit for that.

402

:

But much like a virus,

403

:

I think once the bad actors get involved

404

:

that there is going to be no stopping

those

405

:

those bad actions.

406

:

And that is why I think maybe in five,

407

:

ten, 15, 20 years,

we may not be saying that open A.I.

408

:

and then the other A.I.

409

:

video editing platforms

like Sora and Cynthia, that's easier.

410

:

And in those out there

that can be utilized for good,

411

:

that those weren't necessarily

a net good for society,

412

:

and instead they have broken society,

413

:

they have caused irreparable harm.

414

:

But maybe I'm just a negative, Nancy.

415

:

Maybe I am just being cynical.

416

:

Maybe I am being shortsighted.

417

:

Maybe I do not have enough faith in the

the creators

418

:

and the engineers and the geniuses

that are creating these different

419

:

tools within the air space.

420

:

But tell me what you think.

421

:

Leave a comment in this video.

422

:

Educate me.

423

:

I am new to the air space.

424

:

Am I being too negative?

425

:

Am I not seeing the big picture?

426

:

Is my view of the big picture.

427

:

You know,

428

:

looking through the negative lens

when I should be

429

:

looking through the positive lens, is

the positive out going away the negative?

430

:

Let me know

431

:

and I will

I will revisit this topic at a later date.

432

:

But as of right now,

433

:

as of right now,

434

:

Sora,

435

:

incredible as it is

436

:

in other air video

tools, helpful as they are,

437

:

if the technology is there for it

to be used for good,

438

:

the technology is also there for it

to be used for bad.

439

:

And I am going to air

on the side of caution and say that

440

:

I think

441

:

looking back in five, ten, 15 years,

442

:

as much good that was brought to society.

443

:

With these tools,

444

:

we will be seeing that

there was much more negative

445

:

that has actually broken society.

446

:

Thank you for listening

to Counter Thought.

447

:

A podcast conserving America's

freedom, culture and values.

448

:

Remember to subscribe and like or rate

449

:

the podcast on your podcast app

or on YouTube, and engage

450

:

with the podcast on Instagram

@counter_thought,

451

:

@counterthoughtceo or on Facebook

at “Counter Thought Podcast”.

Next Episode All Episodes Previous Episode

Listen for free

Show artwork for Counter Thought

About the Podcast

Counter Thought
Conservative commentary about news, politics and culture
A podcast conserving America's freedom, culture and values.

As a Conservative, I believe America needs to be restored to Conservative principles. This podcast is my dog in the fight.

About your host

Profile picture for Brian Kletter

Brian Kletter

I was born and raised in Florida, and currently reside in Orlando! I grew up in a Christian household, and accepted Christ as my Lord and Savior when I was 8 years old then renewed my relationship with Him when I moved to Orlando almost 10 years ago. I am a 2x graduate of the University of Florida--there was never a question if I was going to attend anywhere else! I bleed Orange and Blue. Sports has always been a passion of mine, however, I made room for this podcast to be a new passion as I fight for the freedom, culture and values of America. I hope you enjoy!