I will disclose just as like a personal

bias that I hate ads. Um I think I think

ads were important to get to give the

early internet a business model. I don't

like them in general and I think that uh

ads plus AI is sort of uniquely

unsettling to me. You know, when I when

I think of like GPT writing me a

response, if I had to go figure out, you

know, exactly how much was who paying

here to influence what I'm being shown,

>> I don't think I would like that. And as

like things go,

>> oh, you're my boy, Sam. You're my boy. I

loved it. Yeah, that's that's how I

feel. I wouldn't like that either. It

feel really unsettling to me. What's

this one?

>> Advertising, things like that. What's

open's approach towards that? How are

you going to handle that responsibility?

>> We know what it is, Sam. done any

advertising product yet. Um, they have I

kind of

>> I mean I'm not totally against it. I can

point to areas where I like ads. I

>> think ads on Instagram kind of cool. I

bought a bunch of stuff from them.

>> What the hell's going on here? Yes, you

saw it here correctly. Sam Jippity Alman

doing a 180 over the course of two years

and now we're going to start seeing ads

in chat GPT. But honestly, how serious

are they? What are the ads going to look

like? Do they really need the ads and

how bad can it actually get? So, how

serious is OpenAI? Well, I think to

really understand how serious they are,

you first need to understand their

situation that they're in. Now, I've

seen a lot of numbers. There's some

great videos that I'll link in the

description that talk about kind of the

liabilities that Open AAI has. The

circular investing weird oraoris that is

Oracle, Nvidia, AMD, Open AI, and just

whatever that is. But just to kind of

give you the raw numbers, they need to

raise effectively a lot of money by

2030. They're going to have to pay even

more money just to be able to have their

data centers. The amount of loss they're

going to have is somewhere around $500

billion, and they need to come up with

$1.4 trillion by 2033. Now, these

numbers have varied depending on the

source. So, I don't really mind using an

AI to actually calculate how much the AI

is going to cost because I've just seen

numbers all in that ballpark. So, the

the clear motivation is that Open AI

needs money. And I think that this

document right here released by OpenAI

really just kind of highlights the

position that they're currently in. And

that is their compute versus their

revenue. If you look at this graph in

2023, they had some amount of AI. All

the way to 2025, they had a whole bunch

of AI, but notice that the growth in

compute has perfectly matched their

growth in revenue. That means their

efficiency of output is equal to their

efficiency of revenue. Even though

they've been able to produce a lot more

AI, they have a lot more data centers, a

lot more power, their margins haven't

changed. Even though Sam keeps telling

me that AI is going to be dropping 10x

every single year. And even at a recent

town hall, he said that chat jippity 5.2

in 2 years will be 100x cheaper.

>> GPT 5.2x high. I would say at least 100x

less.

>> So despite these claims, it does

currently seem that their uh their

revenue is directly tied to their power

output. And I'm positive that will

change. But nonetheless, that is the

state of affairs today. They owe a lot

of money. They're currently losing $8

billion a year, projected to be

somewhere between 12 to 16 this year,

and they have literally $1.4 trillion on

the line over the course of the next 7

years. But nothing really says how

serious they are than this hire right

here. Now, many of you may not know who

this is, this goth chick. Well, it turns

out she's kind of a heavy hitter when it

comes to advertising. For those that do

not know her, her name is Fidget Simo.

She worked at Facebook now Meta from

2011 to 2021 and she is the one that is

credited for leading the team that built

Facebook's mobile advertising business

and she is largely accredited to a lot

of the advertising of Facebook of which

during those years went from 3.5 billion

in revenue all the way up to 114 billion

dollars in revenue. So in other words,

she presided over a 30x increase of

advertisement at Facebook. She's going

to be the Nolan Sarrento of Open AI.

>> We estimate we can sell up to 80% of an

individual's visual field before

inducing seizures.

>> And really, this is part of the dirty

little secret of the Silicon Valley. If

you want to be one of the big boys in

the Silicon Valley, you either need to

make hardware or you got to sell ads.

Apple hardware, Facebook ads, Nvidia

hardware, Google ads, Open AI. You want

to take a guess which one they're going

to do? So, the real question is how much

money are they going to make from

advertising? Well, we're going to do a

couple kind of exercises here because

this really builds into the final point,

which is how uniquely unsettling ads are

in chat GPT. Sam, when he was being

honest at Harvard in 2024, he was just

he was dead on the money. They're

they're going to be uniquely unsettling.

But for now, let's just go with the

projections that we know. Open AAI

projects by 2030 that 8.5% of the

estimated 2.6 billion weekly users or

around 220 million people will be

subscribing to Chat GPT either on their

$20 or their $200 model. That will

probably change by the time things come

through. Who knows what the subscription

tiers will be by that point. So, I made

this friendly little graph right here

that just does some estimations. So, how

many daily active users? I don't know.

Let's just pretend that uh about 1.28 28

uh uses per week from the average person

using Chat Chippity. And let's also

suppose based off some tweets I've seen

that OpenAI is actually able to charge

$60 CPM, which by the way is three times

higher than what Facebook currently

charges. Let's also pretend for every

single prompt or query you make, they

can show one ad. With those numbers, if

that were true, they would be able to

make approximately 52 billion all the

way up to $313 billion per year just on

ads. Now, we can all agree that there

are some pretty serious handwaving that

just went in on those calculations, but

I'd say it's a fairly reasonable range

to assume somewhere between 50 to 300

billion in revenue from advertising over

the course of the next four to five

years. Now, the ads they're going to be

displaying, this is kind of their

current proof of concept right here. You

can kind of see how it's sectioned off

from the actual response. And at the

bottom, it's just like, hey, here's like

an advertisement for you. It's very

clear. It's demarcated. In fact, you

know what it reminds me of? It reminds

me a lot of how Google used to have

their ads. You see this? Like, at one

point, this was Google ads like 15 years

ago. Very clearly demarcated. It was

very very obvious you were looking at an

ad. But something weird kind of happens

because with today's Google, I don't

know when I'm looking at an ad. Like

look at this result right here. Learn to

code. Search it up. Code Academy. Is

that an ad or is that not an ad? Now I

would like to say it's not an ad, but

it's also appearing higher than the AI

overview, which makes me think that is

probably an ad, but it's also not, you

know, it doesn't even have the little

sponsor block or any sort of indication

that it is an ad. But that's the world

we live in today with Google. It has

significantly come a long distance since

the old days where it was very clear

what an ad was. And this is the uniquely

unsettling part about ads in AI. All

right, I have to bust on in here. I had

an entire ending to this video, but it

turns out so many things have changed.

And I think Anthropic actually nailed

this perfectly. If you did not see,

Anthropic released some ads kind of

showing how Chat GPT's ads are going to

be.

>> Hey, can I get a six-pack quickly?

>> Perfect. That is a clear and achievable

goal. Would you like me to tailor a

personalized workout plan?

>> Yes.

>> Perfect. Let me personalize this for

you. Let's start with your age, weight,

and height. Whenever you're ready. 5'7,

23 years old, 140 lbs.

>> Got it. I'll create a plan that focuses

on aesthetic strength training. But

confidence isn't just built in the gym.

Tricep Boost Max, the insoles that add

one vertical inch of height and help

short kings stand tall.

>> What?

>> So, they actually ended up releasing

four of those ads. out of all the ads,

that one to me was the most unsettling

because it uses somebody's kind of

vulnerable state. Hey, I want to look

better. Hey, I want to try to better my

life. It's just like, hey, I know you're

trying to go through those things, but a

lot of people also would say height is

important since you can't change that.

Maybe you could get this thing right

here. But the thing that it also fails

to convey is that people dump an

incredible amount of information into

chat GPT. People are telling their life

stories to Sam Alman and then now

there's going to be ads. And I get it.

Ads right now are going to be this nice

little sectioned off thing. But we all

know that ads are hill climbing

algorithms. They're going to start off

not making as much, but as they add in

more and more information, it's going to

get more and more accurate and people

are just going to keep on buying more

and more stuff. And even if OpenAI has

good intentions, you know, let's just

pretend that Sam Jippity Alman is

actually telling the truth, there's

going to be no way in which they won't

simply let the money talk. Because

unlike most things in life, running ads

comes with a concrete number at the end.

How much money did I make? Oh, we use a

little bit more information about a

user. We earn this much more money. Hey,

we need to increase our revenue by 15%.

What dial can we turn up to achieve

those numbers? You can clearly see why

this is so dangerous. And right now, it

appears to be very simple. And right

now, we're being promised that the ads

will be very sanitized and whatever. But

we all know that open AI is sitting on

probably the largest treasure trove of

personally identifiable information

about somebody. And if social media has

taught us anything, using people's

emotions as a means to engage, produces

more of a desired outcome, which for

social media is time on the platform and

for Open AI it's going to be slinging

ads. Like they're just going to fall

into it. There's just no way that that

won't happen. But there is actually a

more sinister version of AI advertising

that we're not really talking about

which is when does an ad actually start.

Clearly that anthropic one where it was

so obviously an ad being like yo

confidence is not built on just by the

gym itself. What happened if it's like

oh you're building a product okay you

want to launch a product we definitely

want to be using react because you know

if you're not using react you're not

going to be a successful business.

Obviously react works best with forcell.

Hey you're going to want to use this for

a login provider. And what you don't

even realize is you're being sold to on

each one of these little teeny tiny

decisions you're making for your

technology. Like why did it pick

Verscell? Maybe somewhere down the road

there was a deal behind closed doors and

you had no idea. Now obviously I'm not

saying Verscell is doing that now. I'm

just saying you could see a world where

everything you're doing is in

advertising. And even more so now that

they have Claudebot and all these

autonomous agents making decisions for

people and people are just uploading

their entire life to these automated

agents. Like why is it buying Sweet Baby

Race? Do you know? Why is your personal

agent always buying, you know, Horizon

organic milk for you? Is it because it

wants you to be healthy? Or maybe

there's a deal going on for these API

requests? I don't know. And that's

really the uniquely unsettling aspect of

this is that this starts off with like

this really nice principle. Hey, we want

to be able to bring AI to everybody. And

you know what? Maybe they will. Maybe

this will be a perfectly great endeavor

in which everybody gets a chance to use

it. They all get to produce slop and

then realize that they have tech coming

out of their eyeballs. But maybe it

doesn't go that way. Maybe the

advertisements become super particular

to that person, able to use their

vulnerability in certain situations to

be able to sell one of the most

effective sales pitches of all time and

then even further just using simple

output as a means to sell people because

there was some sort of premium training

data they exchanged but twixed hands. I

don't know the world is going to become

very confusing here in the next few

years and I think that this just might

be the moment where it started. The name

is I'm I think ads are uniquely

unsettling with AI. But Jen,

hey, you're probably wondering why am I

in San Francisco? I'm here for a big

event and I'm going to stream the whole

thing. It's going to be live on my

channel for the next five days. So, if

you're watching this video, it's

probably live right