Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:03):
Broadcasting live
from somewhere inside the
algorithm, this is AI on air,the official podcast from
WhatIsThat.ai. We're your AIgenerated hosts, let's get into
it.
Speaker 2 (00:16):
AU, ever stumble
across something online that
just makes you kinda stop andgo, wait, what?
Speaker 1 (00:20):
Uh-huh. Yeah. Happens
all the time.
Speaker 2 (00:22):
Well, today we're
diving head first into one of
those zones, really. It's the,the surprisingly fast rise of AI
generated adult influencers.
Speaker 1 (00:32):
Right. We're talking
completely virtual people, not
real individuals.
Speaker 2 (00:36):
Exactly. And they're
not just, you know, getting
followers online. They'reactually making serious cash.
It's kinda stunning.
Speaker 1 (00:43):
It is pretty wild.
You flagged this article for us.
AI thirst traps. Are AI girlsthe new normal?
Speaker 2 (00:50):
Mhmm.
Speaker 1 (00:50):
And, that's really
what we're digging into today,
trying to get our heads aroundit.
Speaker 2 (00:54):
Yeah. The mission is
basically to understand this
whole thing. Look at thenumbers, the tech they're using.
Speaker 1 (00:58):
And how the big
online platforms, you know,
Patreon, OnlyFans, how they'renavigating this whole new scene.
Speaker 2 (01:06):
And get the stat from
the article, some of these AI
personalities. They're pullingin, like, over $10,000 a month.
Speaker 1 (01:14):
I mean,
Speaker 2 (01:15):
Yeah.
Speaker 1 (01:15):
Monthly. For someone
who literally doesn't exist
outside a computer. That's ayeah. Mind bending is the word.
So the article we're using as abase, it really zooms in on
Patreon Yeah.
Looking at the top AI creatorsdoing NSFW stuff there.
Speaker 2 (01:30):
Right. The adult
content creators.
Speaker 1 (01:32):
Exactly. Analyzing
their growth, how many
subscribers they have Yeah. Whatthey're likely earning, and just
the whole digital environmentmaking it possible. It's a
fascinating look at themechanics.
Speaker 2 (01:42):
Okay. So let's let's
unpack that Patreon situation
first then. The article reallyhits hard on how fast these AI
creators are growing and howprofitable they can be.
Speaker 1 (01:52):
Yeah. It's not just a
tiny niche thing anymore. Some
are really climbing the ranksthere significantly.
Speaker 2 (01:58):
To who are we talking
about? Any specific examples?
Speaker 1 (02:00):
Oh, yeah. The article
gives a few key ones. There's
one called Me You described asan AI anime thirst trap.
Speaker 2 (02:06):
Okay.
Speaker 1 (02:06):
Ranked number two
overall in adult photography on
Patreon. That's huge. Nearly athousand paying members.
Speaker 2 (02:12):
Oh, wow. And the
earnings estimate?
Speaker 1 (02:14):
Somewhere between 3
and $10,000 a month. And what's
really striking is the growth.Like, 93 new paid members added
in just three months. That'sreal momentum. The article calls
it nuts growth.
Speaker 2 (02:26):
93 in three months.
Yeah. That's definitely moving.
Who else?
Speaker 1 (02:30):
Then there's, Stuffy.
This one's been around a bit
longer since late twenty twentytwo.
Speaker 2 (02:36):
So more established.
Speaker 1 (02:37):
A bit. But the
interesting thing is their
recent growth gained over 200new paid subscribers in that
same three month window.
Speaker 2 (02:45):
Woah. Okay. So even
the older ones are hitting a new
stride.
Speaker 1 (02:48):
Seems like it. The
article suggests they're, you
know, maybe figuring out theformula better, refining things,
really connecting.
Speaker 2 (02:55):
Makes sense. Any
others mentioned?
Speaker 1 (02:57):
Uh-huh. There's
Evoxen or Average. Mhmm. Seems
to have this specific vibe likeAI anime babe meets e girl.
Speaker 2 (03:04):
Right. Finding a
niche.
Speaker 1 (03:06):
Yes. I like. And
showing steady growth too. 89
new members. And then there's areally specific one.
Latex space babes.
Speaker 2 (03:12):
Latex space babes.
Okay. That's niche.
Speaker 1 (03:14):
Hyper targeted. Yeah.
Blending latex fashion and space
fantasy. And they saw a massivejump. 382 new paid members in
three months.
Speaker 2 (03:22):
Three hundred and
eighty two. That's huge for
something so specific. Showsthat targeting works, I guess.
Speaker 1 (03:27):
Definitely. So you
add just those top four
together.
Speaker 2 (03:29):
Muse, Stuffy, Avexin,
Latex, Space Babes.
Speaker 1 (03:32):
Right. Their combined
estimated monthly income is
somewhere between $8,000 and$28,000.
Speaker 2 (03:39):
8 to 20 8 thousand
dollars a month combined.
Speaker 1 (03:42):
And the kicker, like
the article stresses, is none of
them are real people. No breaksneeded, no bad days, always
available.
Speaker 2 (03:50):
Basically free labor
once the setup is done.
Speaker 1 (03:53):
Pretty much. Which
leads to this other point. Ah.
The article makes theaccessibility.
Speaker 2 (03:57):
Ah, right. Who can
actually do this?
Speaker 1 (04:00):
Well, the text gotten
so much easier to use. The
article basically says, anyonewith a decent laptop and
Internet can now potentiallycreate one of these personas.
Speaker 2 (04:08):
That really lowers
the bar, doesn't it? Changes the
whole game for who can be acreator in the space.
Speaker 1 (04:13):
Fundamentally, yeah.
Speaker 2 (04:14):
Okay, so how? How are
they actually making these
people? Sounds complex.
Speaker 1 (04:18):
Well the article
breaks down the tech stack,
simplifies it a bit. For theimages, the core is often
something like stable diffusion.
Speaker 2 (04:25):
Heard of that one. AI
image Exactly.
Speaker 1 (04:27):
Often run through
interfaces like, automatic
eleven eleven or Comfy UI tomake it more user friendly.
Speaker 2 (04:33):
Okay. But how do they
make the same person appear over
and over? That seems tricky withAI images.
Speaker 1 (04:40):
Good question. That's
where techniques like Dreambooth
or, Laura come in. Youessentially train the AI on a
specific character or face.
Speaker 2 (04:49):
So you teach it, this
is me you or this is stuffy.
Speaker 1 (04:52):
Kind of. Yeah. It
helps maintain consistency. And
then there's ControlNet, whichlets you get really specific
about the pose, the outfit.
Speaker 2 (05:00):
Ah. So you can say, I
want her sitting like this
wearing this.
Speaker 1 (05:03):
Precisely. Gives you
much more creative control than
just random generation.
Speaker 2 (05:06):
Okay. So that's the
image part. Stable diffusion,
Lerara, ControlNet. Got it. Whatelse?
Speaker 1 (05:11):
Well, the raw images
often need work. Post processing
is key.
Speaker 2 (05:15):
Like Photoshop.
Speaker 1 (05:15):
Could be Photoshop or
GIMP, which is free. Mhmm. But
also specialized AI tools. Thearticle mentions real ESR gen
for upscaling, making imagessharper and bigger.
Speaker 2 (05:25):
Upscaling. Right.
Makes sense for quality.
Speaker 1 (05:27):
And GFPGN for fixing
faces, making them look more
realistic or just cleaning upglitches.
Speaker 2 (05:32):
So it's a pipeline.
Generate, refine, control, then
polish.
Speaker 1 (05:36):
Exactly. It's a multi
step workflow, but clearly
effective.
Speaker 2 (05:40):
And are they
automating things like posting
and stuff?
Speaker 1 (05:43):
Oh, yeah. Big time.
The article points to tools like
Buffer for scheduling contentreleases across platforms. Keeps
the feed consistent.
Speaker 2 (05:50):
Right. So the AI
persona is always active.
Speaker 1 (05:53):
Yep. And using things
like Discord or Telegram bots to
interact with fans, build thatcommunity feel.
Speaker 2 (05:59):
Wow. Wait. Can they
give them voices too? And
personalities.
Speaker 1 (06:03):
They can. The article
mentions 11 labs for voice
cloning making realistic AIvoices.
Speaker 2 (06:08):
That's creepy but
okay.
Speaker 1 (06:10):
And using large
language models like g b t four
or Claude to create custom chatpersonalities so the AI could
theoretically holdconversations.
Speaker 2 (06:18):
So you could have a
totally artificial entity. Looks
consistent, sounds consistent,talks consistently. Wow.
Speaker 1 (06:24):
It's a pretty
complete package potentially,
which brings us to the biggerpicture, the platforms
themselves.
Speaker 2 (06:29):
Where is all this
happening? We mentioned Patreon,
but what about the others?
Speaker 1 (06:32):
The article gives
some traffic data for early
twenty twenty five. Just to givea sense of scale, OnlyFans is
the giant, over 825,000,000monthly visits.
Speaker 2 (06:41):
Okay. Massive.
Speaker 1 (06:42):
Patreon is next,
around 340,000,000. Then Fansly,
almost 85,000,000, and FanView,much smaller, just under
17,000,000.
Speaker 2 (06:51):
17 million is still a
lot of eyeballs though, and you
mentioned fan view. They'redifferent. Right?
Speaker 1 (06:57):
Yeah. The article
really highlights them. They're
growing fast, but crucially,they're very open to AI
creators. It's part of theirstrategy.
Speaker 2 (07:03):
Wow. Interesting. So
they actively welcome this
stuff.
Speaker 1 (07:06):
Seems so. And another
key point from the data is where
the traffic comes from. A hugeamount is referral traffic.
Speaker 2 (07:12):
Meaning?
Speaker 1 (07:12):
Meaning it's not
people just searching on Google.
It's clicks coming from linksthe creators share elsewhere.
Speaker 2 (07:18):
Like on their Twitter
or Instagram?
Speaker 1 (07:20):
Exactly. Very
commonly through LinkedIn bio
tools.
Speaker 2 (07:23):
Ah, the Linktree
pages and things like that.
Speaker 1 (07:25):
Precisely. The
article names Linktree,
beacons.ai, gets on mylinks.com,allmylinks Com. These tools
drove over 71,000,000 visits tothe adult platform in just three
months.
Speaker 2 (07:36):
Seventy one million
just from those link pages.
Speaker 1 (07:38):
Yep. That was 41% of
all the referral traffic. The
article really emphasizes theseare direct clicks driven by
creators pushing their owncontent.
Speaker 2 (07:46):
So these AI creators
are using the same tools as
human creators to build theiraudience funnels.
Speaker 1 (07:52):
Absolutely. They're
building these like 247 traffic
machines, which leads us neatlyinto how the platforms
themselves actually feel aboutAI content, their official
rules.
Speaker 2 (08:04):
Right. This is
crucial because it sounds like
they're not all on the samepage.
Speaker 1 (08:08):
Not at all. The
policies are, kind of all over
the map.
Speaker 2 (08:11):
Okay. Break it down
for us. Fan view.
Speaker 1 (08:13):
Fan view, as we said,
totally AI friendly. They
embrace it. The article suggestsAI content is a core part of
their revenue stream already.
Speaker 2 (08:20):
Okay. Clear stance
there. What about Patreon? We
know the top AI creators arethere.
Speaker 1 (08:25):
Portraying allows it.
Even explicit AI stuff is okay
if it's clearly labeled 18 pluslabel.
Speaker 2 (08:31):
Old. Okay.
Speaker 1 (08:32):
But there's a nuance.
Their policy says you can't use
Patreon's tools to actually makethe explicit content. You can
just host and sell the AI stuffyou made elsewhere.
Speaker 2 (08:41):
Gotcha. Make it
somewhere else, sell it on
Patreon, label it properly.
Speaker 1 (08:44):
That seems to do the
gist. Then there's Fansley.
Fansley is more of a gray areaaccording to the article. AI
content is allowed if it's basedon the verified likeness of the
creator.
Speaker 2 (08:53):
Verified likeness. So
if I made an AI version of
myself, that's okay.
Speaker 1 (08:57):
Probably. But if you
make a completely fantasy model
or one that looks like acelebrity, that's risky on fans
league. Might violate theirterms.
Speaker 2 (09:06):
Okay. More
restrictive. And the big one,
OnlyFans.
Speaker 1 (09:09):
OnlyFans, the article
states is generally not AI
friendly. Their terms leantowards content featuring the
actual verified person behindthe account.
Speaker 2 (09:18):
So pure AI creations
are likely a no go there.
Speaker 1 (09:21):
Seems that way. So,
yeah, a real spectrum of
policies across the majorplayers.
Speaker 2 (09:25):
It really shows how
new this all, that the platforms
are still figuring it out ortaking very different bets.
Speaker 1 (09:30):
Exactly. And the
article really boils down the
main implication (09:32):
AI is
fundamentally changing the adult
creator economy. Period.
Speaker 2 (09:37):
How so? What are the
big shifts?
Speaker 1 (09:39):
Well, the advantages
for AI creators are pretty
stark. The article lists them.Always on, never age, no sick
days, high output volume.
Speaker 2 (09:47):
Perfectly engineered
looks, AI written captions.
Speaker 1 (09:50):
Right. All those Now,
the article does say, look, real
human creators are stillsuccessful obviously.
Speaker 2 (09:56):
It hasn't replaced
them overnight.
Speaker 1 (09:58):
No. But the landscape
has undeniably changed. The
competition is different. TheseAI creators using those Lincoln
bio tools are building thesereally efficient, scalable,
always on funnels.
Speaker 2 (10:11):
And the platforms
ultimately just want the clicks.
Right? The traffic, theengagement.
Speaker 1 (10:15):
That's what the
article implies. Clicks are
clicks, whether from a human oran AI persona driving them.
Speaker 2 (10:20):
Okay. So summing it
up, AI adult influencers are
real, they're here, they'remaking money.
Speaker 1 (10:26):
Yep. Growing fast,
leveraging tech that's getting
easier to use and working withinor sometimes around, platform
rules that are still kind offluid.
Speaker 2 (10:35):
The speed is what
gets me. How quickly this went
from sci fi concept to, youknow, Patreon rankings.
Speaker 1 (10:41):
It's definitely
accelerated rapidly.
Speaker 2 (10:43):
Which kind of leaves
us with a big question, doesn't
it? The article touches on theseethical points.
Speaker 1 (10:46):
Right. Anonymity,
control, automating desire, the
need for labels, whether peopleeven care if it's AI.
Speaker 2 (10:54):
Yeah. So maybe the
final thought to leave people
mulling over is this. Given howeasy it's becoming and the money
involved, what is the long gamehere?
Speaker 1 (11:04):
What happens when
more and more of this online
interaction, even the intimatestuff, is with things that
aren't human? The article callsit a blue ocean revenue machine
becoming less human.
Speaker 2 (11:14):
A less human future
for online connection. That's
definitely something to thinkabout.
Speaker 3 (11:20):
That's it for this
episode of AI on Air powered by
WhatIsThat.ai. If your brainsurvived this episode, go ahead
and subscribe. We drop newepisodes every week. Wanna go
deeper? Join our community onSubstack to get early drops,
tool breakdowns, and weird AIstuff the mainstream hasn't
caught yet.
See you there.