319 comments

[ 2.0 ms ] story [ 279 ms ] thread
How long until GPT-4 plus virtual humans replace youtubers?
I think the growing popularity of vtubers is probably a sign of things to come. I think once (if it's possible) you can synthesize natural sounding speech and basic emotions/behaviors, it wouldn't surprise me if that didn't become a huge amount of the market for video content.
Why would anyone watch that? I thought that after all the infatuation with technology, power and wealth other humans are still what we need to connect with..

Why not shut oneself into a virtual world if virtual everything is what they want? I have the feeling that in the not too distant future there will be big labels on things stating “not virtual” like we’re getting now for organic/bio products

You would be amazed at the stupid shit people watch. There's a surprisingly enormous Twitch segment of people who stream playing slots. No gameplay at all, just watching someone else clicking "go again" on a random number generator. Here's a slots streamer who apparently has 30 thousand live viewers right now: https://www.twitch.tv/trainwreckstv
We already have tons of vtubers who make no secret of using voice changers. Some, like Zentreya, use speech-to-text-to-speech for an artificial voice.

If that sounds odd to you, I encourage you to check out her channel and compare with similar content from Twitch streamers; it's surprisingly easy to get used to and enjoy.

Hopefully a GPT-4 would improve the overall quality of the vlogs :-)
Very cyberpunk - something I would expect to read in the pages of Transmetropolitan.
This is shockingly realistic (and perhaps a bit disturbing that a virtual model can generate so much interest). How do people create such virtual models that look so accurate and human-like?
Body is real. The face is synthetic and augmented onto the actress.
They claim AI, but it's probably just 3d modeling and texturing. Possibly based on a real person - see also characters like Aloy, or the ones from Death Stranding. Or the attempts in recent Star Wars media to bring back deceased actors or make them younger.
I think this only works because of the novelty effect.

A big reason why influencers or any kind of celebrity are successful is because enough people aspire to be like them. That's pretty hard to achieve for an AI.

I think vtubers are a counterpoint to that argument. There's a lot of virtual avatar type celebrities on internet platforms in particular and they have had fanbases for a while.

I also don't think people aspire to be like influencers in a 'real' sense. Influencers are hyperreal versions of something fans know they won't actually be, and in that sense artificial characters might even be more interesting.

Yes 'codemiko' looks like a good example of how the future will play out.
I don’t know anything about influencers but why would people want to follow (more than for fun) someone that is AI generated? I mean an INFLUENCEr would be someone I aspire to be like. So if i follow a marathon athlete I’d like to be like her, do the same exercises etc. that would be impossible to mimic with a virtual marathon runner.

Replace marathon with ‘coding’ or whatever hobby you have, and I’m sure you’ll find it weird as well

If virtual influencers are travelling and taking pictures of themselves at cool places, you can certainly visit those places too. If they take pictures of food (very influencer-y thing to do), you can visit that restaurant and eat that food.
> I mean an INFLUENCEr would be someone I aspire to be like.

It seems not necessary to be.

You're reinterpreting what "influencer" means. An influencer isn't someone who influences you to be like them. That's called "role model". An influencer is someone whose job is to influence your purchasing decisions.

Before it became a career, originally "influencer" was a marketing designation for a person who's well-connected within target market's social graph, and thus is the optimal target to hire or manipulate into peddling crap to their followers.

There's nothing here preventing virtual personas from being influencers too. Being a real person who can be emulated is an asset for an influencer, but it's not a necessity - it doesn't matter why people follow you, it matters that they do, and that they are vulnerable to you manipulating them. Having such a network of vulnerable followers is what makes an influencer valuable to sponsors.

Vtubers are a really interesting side phenomenon to this. I think part of the appeal of a vtuber is that it seems more fair and effort-based than an influencer -- as far as I can see, you can become a successful influencer with nothing but a pretty face and maybe some good photography/photo editing skills, but in the vtuber world 99% of your success is just your ability to make entertaining content. There are vtubers with janky homemade live2d rigs that manage to be successful despite that, and there are vtubers with beautiful models commissioned by well-known illustrators that remain fairly obscure because the talent behind them is mediocre.
I do think that the novelty effect will quickly wear off.

What is the future for 'virtual humans' after that? Will it be mainly virtual assistants (that people will quickly grow to hate)? Will they come to dominate pornography or say cam sites, and be tarred by that association and then end up ignored in other spheres? on the other hand, if they avoid that fate, they could gradually become a pretty significant force in mainstream film; i think getting popularity there will be a hard road to climb if people first associate them as tacky.

And unless these virtual personalities end up engaging in scripted dramas and celebrity feuds and things like the way (i gather) k-pop bands do etc, they risk people coming to find them extremely boring. Being able to be in more than one place at once is in some ways a disadvantage if it makes them so ever-present that their audience no longer experiences them as part of some kind of narrative, which is what lets people develop feelings for them.

Remember how people used to say that our entertainment was going to become interactive - that movies and TV shows were going to become more like videogames? Well, it's happening, but from the opposite end.
The future looks like a hologram siri virtual assistant.
Arguably we've come far enough that people would be fine with the idea of a completely virtual image projected to them.

I see it two-folds:

- the image influencers are projecting is already detached from most people's reality, they know it's faked but accept it as an ideal to tend towards. Just as we all know movies are full of VFX and 1 in a million charismatic actors that are horrible people IRL, but we still aspire to the projected image.

- having no human being directly attached to the persona means no scandal, no medical issues, no miserable life lived in the shadow of the spotlight and it's beneficial to both producers and viewers. You'll never have your idol smeared by tabloids because they kicked homeless people.

RE your second point, I think there will be scandals. They'll just be 100% scripted[0]. If there's one thing we've learned about the Internet over the past 30 years, it's that people crave drama. Owning a few talking heads on YouTube is good, but having them get into fight is even better - you have an excuse for a nearly infinite stream of reaction videos, reaction-to-reaction videos, apology videos, apology-questioning videos, meta-analysis videos, etc. Viewers binge this, and influencers (and their owners) swim in money.

--

[0] - Though I think it's safe to assume the big ones today are also scripted. For example, every now and then my wife stumbles on some video from the cosmetics space on YouTube, and holy shit, the makeup-peddling influencers there have been running an ongoing personal drama for some good 3 years now. It will probably last another 3 - viewers seem glued to their seats, and producers just enjoy the viewership and opportunity to peddle some more name-brand makeup.

You might be right, drama is core component to entertainment. Hopefully with virtual representation, we'll have more virtual trash talking and less stalking/doxxing/career destruction.

We're having a taste of that with vtubers, where there's a lot happening, in particular as the money amount moving around are already pretty big, but comparatively way less dire than with actual idols.

Heh. Reminds me when League of Legends introduced the character Seraphine as the new member of their virtual kpop group, with a whole twitter account about her "daily life" and stuff.

I remember reading one reddit comment that went "Can't wait for the moment Seraphine gets canceled for something she said in an interview and has to put out a tearful apology video. Gotta hit those real-life influence milestones!"

(of course Seraphine would never actually get canceled; that would require that her writers give her any amount of personality whatsoever)

(comment deleted)
I don't think it is "to aspire to be like them", more of a "to aspire to be close to them". The same story for celebrities. You just want good things to those you spend time cherishing. It has nothing to do with whether you think it's real, fake, obtainable or not. BTS aren't popular because people want to be like them, they're popular because people want to be with them.
The males want to aspire to be a member and females to be close. You were both right.
Perhaps, but on the other hand something like Hatsune Miku has been around for over a decade and generated a little industry around itself. I don't know about this particular attempt but I wouldn't dismiss the idea as a whole.

https://en.wikipedia.org/wiki/Hatsune_Miku

Agreed, though the title of this article is deeply misleading, there's no AI involved apparently (maybe they swapped CGI/AI for clicks?).

That changes the likelihood of this succeeding though, not only because it has been done before, but because influencers are already deeply artificial (no pun intended), and in the process of becoming successful having a near perfect control of your image through computers might help.

Fails to mention that 80% of the profits are coming from the AI's OnlyFans account. Just kidding. It's insane how many niches like this there are to make some quick money with available open source libraries, if you know what you're doing and you have some time on your side.
I understand where you're coming from in that the technical abilities to create something (someone?) like this may seem like a low barrier to entry.

However, I disagree that this is "quick" money or just a fad. The value of this AI for sponsorships/ads comes more from a team of actual people that scan social media to search for trends, photo styles and carefully helps curate its presence. For example, look at how the first "photo" in the article is created. Notice how the AI isn't looking at the camera? Or how the perspective is somewhat tilted? Overall, the image has many subtle details also very in line with how many millienial/gen-Z Korean girls upload birthday photos from cafes on Instagram.

It feels like Rozy represents a broader movement towards abstracting out the elements that make us resonate with a brand or image more. I think brand strategists and marketers have figured out that the model is more of a vessel to store and convey those elements/emotions and for particular types of ads there is no real need for the model to actually exist in reality. Think about how many ads where you don't know the model in any other context.

Rozy probably adds more value than the nameless models in some sense because there's a persona/brand the team has built up in social media/other ads. Definitely not easy money.

>> Notice how the AI isn't looking at the camera?

It's not an AI!

:pulls pigtails:

> if you know what you're doing and you have some time on your side.

Key point. The AI generated character is the medium; the vast majority of effort was put into the marketing and the "story" of this character.

> Just kidding

Holy crap, I 100% believed you.

Following trends like virtual humans is interesting, and honestly I don't know what things might look like with virtual humans in 15 or 20 years - but I do wish reporting took care not to veer into P.R. territory.

EDIT: deleted a bit here as i misread the word sponsorship.

Not scholarship. Sponsorship. Product representation.
thank you, i misread that somehow.
The article mentions scandals several times. Are scandals that common for celebrities in SK? They seem very confident in their ability to maintain their AI girl’s image. I wonder if they considered the Rule 34 folks.
I know that, in some instances, things like idols publicly dating constitute as a scandal.
> Are scandals that common for celebrities in SK

Oh yes. Everything is a scandal in such a competitive society, where cover ups are rampant just to avoid scalndal. Take this article about Seo Yea-ji for example:

https://www.scmp.com/magazines/style/celebrity/article/31298...

I think they are working to change the cover up culture, as it's almost always represented negatively in dramas.

Sexual awareness, like holding a boy's hand, is considered a major scandal. SK celebrities are often expected to be highly sexualised in clothing, their songs, etc. whilst maintaining an entirely "pure" persona. i.e. Never considered kissing, etc.

This "impure purity" is expected for both male & female celebs, though there are some differences and boys tend to get a little more leeway.

The celeb also tends to take the blame for their creepy fans. It's their fault for 'encouraging' stalkers, etc.

It's an impossible culture that chews people up and spits them out.

As I understand it, western streamers also may lose a big chunk of their followers engagement simply by mentioning their relationship. Lonely twitch watchers are sometimes substantial part of auditory, who have feelings to their pure idol. Also, look at the percentage of regular popular instagram (formerly fb) girls who NEVER post their boyfriends. It’s not only a culture, it sells much better.

That said, SK undoubtedly pushes it beyond any sense.

It's the same in the west though, or, used to be with the boy / girl bands from the 90's / 2000's. Pretty sure K-pop ensembles are loosely based on that.
It's really not even near to the same degree. For example, an unstable fan may bring a knife to a handshake event, see someone else shake their idol's hand, lose it, and start knifing people.

The idol is blamed for the incident. For the state of mind of their fan. For doing anything that might incite jealousy.

this is depressing.
I relate. It might also be Darwinian.
On the other side of the coin there's an army of young people dedicating their life to be instagramable.
in other news, skilled performancer without a pretty face get more jobs.
Are there any companies listed on public exchanges doing this kind of work? This will surely change advertising forever.
I don't think so. You still need a body model, so it kind of defeats the economic aspect.
It might make economic sense. Instead someone who knows their value and asks for it you can get the cheapest aspiring gig worker and replace them as needed.
basically the Stig but with a fake face.
The face can be somehow detected as AI generated. At least I can still detect a difference.

What surprised me was the hands. The hands are perfect. I can't see anything at all that makes me think they are not real.

It is amazing.

Eh, that's because hands are real. They are face swapping rendered face to real body. Source: easy to find if you can read Korean.
oh. makes much more sense. I don’t see how this helps avoid scandals unless they’re using multiple interchangeable models plus the face? https://www.instagram.com/rozy.gram/ I can’t see the face swap at all tbh
They are using multiple interchangeable models. They stated they already have four models.
(comment deleted)
Ahhh, then it is disappointing. Anyone can do photoshop.
The teeth are still wrong, though.

I'm not sure what's so hard about teeth, but (so far) weird teeth has always been a reliable tell for me.

Virtual influencers have been around for a while. Couldn't tell from the article what the "AI generated" part adds to it though.

https://www.bloomberg.com/news/features/2020-10-29/lil-mique...

Yes, Rozy is exactly like Lil Miquela and creators pretty much say so. I am not sure why this coverage is different.
Hatsune Miku from 2007
> "These days, celebrities are sometimes withdrawn from dramas that they have been filming because of school violence scandals or bullying controversies. However, virtual humans have zero scandals to worry about."

Wait till they make the virtual human “talk” (or put up messages and replies online) and you’ll have scandals very soon. Others have learned this the hard way before.

If we don't already have it now, it's only a matter of time before we have a combination of deep-fake AIs with conversational AIs which can break a voice call mid-stream and begin influencing the conversation in a particular direction. Think of it like a man-in-the-middle attack.

Our intelligence agencies would be absolutely stupid not to work on this. Any voice communication channel you can intercept, you can influence, and you can use the trust established with vocal familiarity to enable you assume the influence level of one or both of the speakers. The key would be training the AI to know when to alter the conversation and when to let it pass through.

Not only deep fakes. What happens if/when the generated/virtual human turns out to look identical to a real (living) human being.

There are many kind of possibilities for things getting weird at that junction. eg if the real person is a porn star, political figure, <etc>

Why do things like this generate so many responses along the lines of "this is depressing," "technology is terrifying," "humanity is doomed" and etc? We've always loved entertainment ever since the dawn of our species. "Generated people" just seem like another form of entertainment medium, enabled by advancements in technology, to me.

Sure, I see the points about how we're more and more isolated and anti social and stuff.. but is that really a problem? Our evolutionary programming has the primary objective of genetic propagation, but it's ultimately up to us to decide our purpose in life. Someone said that real "AI-takeover" scenarios would involve such enthralling fantasies and entertainment that humans cannot even begin to imagine, and humans won't even put up a fight against AIs. As for me, I'm not alarmed; I absolutely look forward for that scenario to happen.

Surely, you can't be this credulous
Maybe I am this credulous
There are other serious concerns I do have about AI, but on this topic I think it's a reasonable argument.

What you have to remember is that the average person sits in front of a TV for 3-4 hours a day, watching characters who aren't real act out stories that never happened. Some people enjoy the shows where real people stand in front of a camera, while others (guilty!) prefer the ones where cartoonists draw fake people. In both genres, we've long since reached the saturation point where you could watch every second of your free time, but most people don't do that because they don't want to. The same dynamics generalize to Instagram, especially with filters allowing "real people" to look perfect in a way no actual person can - I don't know of reliable statistics on exactly how much people use Instagram, but I'm pretty confident it's not bounded by the number of pictures they could find if they wanted.

Does adding in an intermediate category of "cartoon but looks like a real person" fundamentally change these dynamics? I'm not 100% sure, but I lean towards no.

“Generated people” are no different than animation videos and movies in this regard. Another art form. No?
It becomes another thing when people aren't able to tell they're looking at something entirely artificial and are led to believe the person they're seeing is an actual human being. That's already happening with good deepfakes and it can be used for highly effective manipulation.
Everyone aware knows that “influencing” is full of the unreal stuff, including personalities. Those who do not realize it are manipulated either way. Only social education can fix that.
I can foresee a future where this (in practice) eliminates choice/individuality because people will allow their responses to be dictated by an AI for "best results."

Then whoever controls the AI controls society.

Tag onto that how AI/ML often reinforces systemic problems, and one can foresee how it would further entrench the status quo.

You make a really great point. This AI stuff is certainly powerful for whoever controls them, though I personally believe that's a problem with the operator or societal system rather than the AI itself.
Exactly. Human models are easy to manipulate in the same exact way (maybe even easier except “scandals” part). Pour enough money on them and they will relay anything you want in the format you demanded.
> Why do things like this generate so many responses along the lines of "this is depressing," "technology is terrifying," "humanity is doomed" and etc?

Don't worry about those comments, they were made by bots.

We are a global startup in Seoul, the company next to our office creates this type of character. I am not sure what they use their models for though. Very talented work they create.
I think in addition to personal finance, they should teach two more topics:

1. How to win friends and influence people

2. How to not be influenced easily

They can't mouth and teeth right...
Advertisements are already photo-shopped and completely fake. Knowing that the models are AI generated doesn't change much about that. It could actually have a beneficial effect. If young people (especially girls) know that the images they see on-screen are 100% computer generated, maybe it will reduce their effectiveness at convincing them that the ad is an object worthy of emulation.

I'm sure the advertisers will scientifically test for this effect and use whatever strategy keeps them buying though...

> If young people (especially girls) know that the images they see on-screen are 100% computer generated, maybe it will reduce their effectiveness at convincing them that the ad is an object worthy of emulation.

You think so? Nobody ever wanted to be like Sailor Moon?

Did Sailor Moon push facial creams?

Wasn't the protagonist supposed to be somewhat of a slob?

Not creams, exactly. But Sailor Moon has always been heavily merchandised. And the Japanese name is "pretty guardian sailor moon" and she says "moon crystal power make-up" and her transformation tool has the shape of a cosmetic kit and ok I'll stop now.

To be clear, this is just the most superficial stuff. That television series was definitely pushing a certain specific and not at all enlightened standard of female behavior and beauty. Not that I think it should be cancelled on Twitter or anything, but I think it's helpful to criticize the media we consume.

Plenty of western cartoons made to push toys though, e.g. Masters of the Universe.
Or the current menace of parents, Paw Patrol. So far we've managed to avoid exposing our daughter to it, but we know it's only a matter of time.
Or, to take the canonical Western example, there's been a lot of discussion about the absurd proportions of barbie dolls.
Good point. I don't know why I ever try to be optimistic about industrial practices.
That's way too optimistic. K-pop idols are already entirely superficial and 90% fake. This doesn't stop people putting them on pedestals for things that are totally constructed.
So are western pop stars though, mostly manufactured by record labels and producers.
You're right. I had a moment of weakness and overoptimism. We really need regulations on advertisers, it's the only way to reduce the harm.
I'd love that a visit to a studio to see how these images are made be part of the mandatory school curriculum.
Interesting bit at the end:

> As CEO Baek said, the reason for the popularity of virtual humans is that there is no fear that advertisements will be suspended due to unsavory privacy scandals after the AI model is selected as the advertising model. In addition, the location and scene can be created through computer graphics, so the virtual model is not limited in time and space, and unlike real people. The other advantage is that period in which the model can be active is very long or eternal because the virtual human doesn't get sick or grow old.

Offensive (by whatever definition of offensive in the context) images can be generated and poison the perception of the virtual persona.
Interesting to read about virtual human on a website with the design that is the next level of evil. Another one which doesn't account that the content can be zoomed in on a desktop too, not just on mobile platforms.
(comment deleted)
Anyone with more knowledge know how much of this is BS?

She has 60k followers on Instagram which (even if all real) isn't that much. Certainly not anywhere big enough to make 800k a year that the creator claims he plans on making.

They say they have 100 sponsorships but this seems like a marketing statement...influencers all the time talk about how popular they are and I don't see why we should trust the creator of AI influencer more.

I don't see any major brands on her Instagram though I'm not South Korean so maybe I'm missing something - but more likely explanation seems to be that he has partnered with a few companies for a pittance and is overhyping his creation to get more money in.

South Korean here. Shinhan Life ran a TV commercial of Rozy. It's not BS.
If you were South Korean, you'd realise how much the marketing sphere of Korea is influenced by passing fads and memes. There's never a single thing that hasn't been picked up by at least one larger person to try run their luck at a new trend.

TV commercial does not make it not BS. There are lots of things in Korea that are utter BS but get a lot of attention, and people are generally fearful of missing out.

WTF if you were South Korean? I was born in South Korea and lived here for 35 years.

I am saying creators are probably not lying when they project $800k yearly revenue. Yes, you can earn $800k in South Korea entirely by fads, but that was not the question.

Yes. You're the one bringing up the fact that you're South Korean for some reason. I'm rebutting your comment with more information that actually suits Korea.

I know what you're saying, I'm not sure what being Korean has to do with that. And no, no one said you could not earn money in Korea through fads. What does that have to do with what I wrote?

I'm saying they are most definitely capable of lying, especially in cases where lying grants you a huge boost to future results.

Are you replying to the right comment here?

handmodel mentioned "I'm not South Korean", that's why I mentioned it. That's all.
Yes. You mentioned that you were South Korean, so I mentioned it. That's all. So I don't quite understand what warrants a WTF.

I backed up why I disagree that a single spark determines whether it's not BS. I assumed, being in Korea for so long, it'd be a given first thing to bring up when someone asks if hype is actually genuine. So I brought it up. Every fad with even the slightest traction will be overvalued more in Korea, and it will also fall out of favour much faster too. The cycle is very much filled with BS that hasn't started to stink yet.

You should see the investment funded startup scene in Korea. It's very close to throwing anything at the wall and waiting for the one that sticks. Which is a shame given how technologically empowered Koreans are.

I'm still unsure why it made anyone upset. But then again, this is the internet.

p.s: Post is rate limited now, so there was a delay in this reply.

FYI perhaps your misunderstanding of the WTF is because you didn't realise what you said in English?

"If you were South Korean" is exactly saying "You are not South Korean", which sanxiyn rightly took offense to.

As a native English speaker from the colonies, I have zero idea whether "If you were" in this usage is the subjunctive, or if it is a conditional clause, or something else. Maybe I need to take an ESOL course to learn English grammar.

No. It is exactly what it is written as. "If you were at the beach, you would have seen the shark." It's not saying you were there or not, it's saying that you probably should maintain knowledge of the following clause. I suppose it could also mean you were not there. But not always.

I can see why they'd get upset if I was assuming they're not Korean. But that's not what I said or meant.

It would be considered directly saying that if I were to say "You can't be Korean because ____" or "You would be Korean if you ______".

It may not be what you meant, but it is what you inadvertently said.

I think you might be thinking of a similar construct which doesn't imply it's false: "If you're Korean, you realise ...". The "you" here is non-specific, and the phrase doesn't say you're Korean or not Korean. The present tense and lack of speculative "would" are what make the difference.

"If you were at the beach, you would have seen the shark." is saying you were not at the beach, and implying you didn't see the shark.

"If you were tall, you would have been able to see it" is saying you are not tall, and implying you couldn't see it.

The "if", "were", and "would" make the phrase a hypothetical counterfactual, imagining that the statement is true and supplying a consequence, and this means that the statement is actually false (i.e. you were not at the beach). This meaning holds when speaking about a specific "you", like the person you're replying to. I'm sure there's a name for this grammatical construct, but I don't know it.

The sentence may be incorrect, in which case the person you're speaking to could correct it with "I was at the beach, but I didn't see it", "I saw it by climbing on a rock", or "WTF I am Korean". It's possible the speaker may suspect their sentence is incorrect and be using it to ask a question indirectly, inviting correction.

A more general meaning is possible when you're not talking about a specific person, but using the conditional to imagine a group of people or something, e.g. "If you were employed, you could afford it, and if you were on a benefit you could afford it, so most people can afford it." I think this usage is uncommon though, as there are much clearer ways to phrase that sort of thing. It's definitely not the usage you wrote.

Source: Native English speaker, but not trained grammarian. Also wrote way too much because it's late at night and I can't be succinct.

> bringing up the fact that you're South Korean for some reason

It's completely normal to qualify one's perspective, and often beneficial. in fact for people to know whether your comment is an opinion, or based on some actual experience, knowledge, skill, etc, it adds value to the discussion.

In the context of South Korean culture, it's absolutely useful to know the commenter is South Korean.

If we're talking about tea, crumpets, or the rain, I might point out that I'm British.

If you were British, you would know where York is.

It's not saying you don't know. It's saying that, this should be part of your knowledge based on you being British.

I don't see how that's offensive.

Because as others have already explained, that's not what that phrasing suggests. "If you were/you would" suggests that it's a purely theoretical perspective and thus strongly implies "you aren't".

"If you were British..." and "If you are British..." have very different subtext.

Nuance of the English language, but, "If you were British, you would know where York is", suggests I'm not British, and implies I don't know where York is.

That specific form of the sentence has negative connotations and might be considered divisive in (British) English.

It's essentially a challenge to my honesty, and had you said "if you are" rather than "if you were" it'd be a more amiable challenge, implying I should respond with proof that I know where York is, and not suggest I'm being dishonest.

As I said, it's divisive, but to some may be considered offensive too.

Just one word makes all of the difference.

If you want to suggest I would know where York is by virtue of being British, you might say;

"You will know where York is, because you're British" or "You must know where York is, you're British".

No. It does not suggest that. I use British English and that's not at all what it means. Maybe to those that don't know how to use it correctly and are seeking emotional connotations.

The statement is at most doubting whether you are British, but not dismissing or claiming the opposite.

>but to some may be considered offensive too.

It's offensive that someone might claim to be from somewhere and not understand a very common piece of knowledge from that place. Or at least I might want to claim it to be.

She does make money, but that's not the only part of the equation.

Rather, people are becoming more and more annoyed by the repeating news articles trying to hype this thing. Also, young people around me gives absolute-zero-f** about this thing.

I should put it this way: boomers don't realize that influencer marketing is the most boomer thing to do these days.

I don't agree with you. I have seen it first hand. Girls find makeup artists they like on youtube that do looks they like and then they follow their recommendations. There are so many options out there. How do you choose? Follow the advice of an influencer. It repeats across the entire market, across all demographics. Building a relationship and then leveraging it. Doing (buying, in this case) something based on the recommendation of another person is basic human behavior. I don't think you should write it off as a boomer thing so easily.
> There are so many options out there. How do you choose? Follow the advice of an influencer.

That's the boomer perspective that I wanted to point out. Where's the internet in this flow? Where's the community? Where's google? I know the above statement is simplified, but what matter is the linking logic, which can be radically different b/w generations.

One thing boomers don't recognize is that, unlike the celebrities from the last generation, influencers are contributors, not rulers or leaders. They make practical contributions to their communities, and, in return, they earn trust, which then can be utilized for monetization. It's not like someone great bam comes in and become a star (which is the exact strategy of this artificial influencer. So boomer, eh?)

Also it's worth noting that the internet community is that it offers more information and raw data than the sum of influencers can ever offer. Follow the recommendation of an influencer? Pf, it's just mere a drop in the vast ocean. A proper digital-literate - and MZs are the first digital native generation - will search other sources, cross check the information, and, naturally, expand it even further. As more effort goes in to these steps, the significance of influencers drops rapidly.

So, I want to say that the influencer marketing can't be a key strategy these days. Rather, it's just an element in the usual advertisement campaign. The most critical part is the community engagement.

(comment deleted)
It's all BS.

"Influencing" is a ponzy scheme. If you can convince people you are trending, you'll trend. If you can convince people you are influencing, you'll influence. Sadly, like all trends, the biggest players are all big fish trying to get in before it dies. This doesn't legitimise it though.

This is just an art project that is trending. They're real photos, with real girls and then the creator will place a flavour (ala thispersondoesnotexist.com) through the use of AI. It's akin to automating your photoshop meme pipeline, but way less creative.

It's marketing all the way down. Instagram, Facebook, YouTube, etc. are all just marketing spirals that occasionally spit out unique content that was made with real hard work. The rest is just cycling through other people's work with faux congrats and advertising for your own "channel".

I think you use a different definition of bullshit than the discussion here.
I'll never know. My definition of bullshit follows every definition of the people I've ever discussed with though. That is: Something that does not actually have substance, but may be propped up to be perceived as so.
OP wasn't making any value judgement like this. They were simply questioning if there is actually a story here.

The media tends to notice little phenomena like this and paint them as if it's a cultural megatrend.

Even then, the misunderstanding is in reading the question "is this story BS?" as "is influencing BS?", not in the definition of bullshit.
Using different criteria is more or less a different definition of being BS.
I don't see a way to confirm the story if you do not investigate the likelihood of the value it is "presenting" though.

I'm still confused by these two comments. I don't think we can verifiably claim any personal story on any media outlet is BS or not.

You're talking about whether this kind of account is a scam.

They're wondering whether this specific account is making money.

Those are very different questions. An account can be both, neither, or one or the other.

No. I'm talking about whether it's bullshit. Scam is not the same as bullshit.

I guess Americans use the word bullshit differently to the rest of the English speaking world? Or something else?

If one were to speak bullshit, they'd be doing exactly what I just wrote. No money is involved.

What do you mean by bullshit? People who like it and follow the fake personae on social media are wasting their time? Or what exactly? Are all invented things bullshit, so also books, music, movies, paintings...? Maybe posting this artificial person on social media is just another way to tell a story?
If I photoshop the moon into a picture and post it to a photography forum, most people will pick it up as a bullshit photo.

If I post the same thing to Instagram, most people will applaud me for being such a great photographer.

Everyone is wasting their time. Lots of people like bullshit.

> No. I'm talking about whether it's bullshit. Scam is not the same as bullshit.

You called it a ponzi scheme!

Either way, please listen to the distinction I'm making.

You're talking about ways the industry is bad.

They're asking if the article is flat-out wrong about this single account being successful inside that industry.

Those are two separate ideas to call bullshit on. A bullshit industry can have bullshit success stories, but it can also have true success stories.

>You called it a ponzi scheme!

Yes. And then people had a problem with the word bullshit, for some reason.

It can be both bullshit and a ponzi scheme. That doesn't make bullshit necessarily a scam.

According to Harry Frankfurt, author of On Bullshit, it's when someone makes statements without any regard to their truth.
But how do you define substance? It seems to me if they make money, they have substance, whether what they do is useful in the greater scheme of things or not.

I must admit I don't quite understand your attitude. I personally for example have no interest in "the Kardashians" and would consider their story "bullshit" in a way. On the other hand, they clearly made a shitload of money. So maybe it would be better to analyse what is going on and why it works.

Another example that I personally consider mostly BS is professional sports. But clearly billions of people are into it and a lot of money is being made. Rather than calling it BS, perhaps it is more interesting to think about what might be going on, and what human need seems to be fulfilled by these things.

I was also reminded of "Hello Kitty" which seems to have been going strong for decades. Maybe, among other things, people just buy into this phantasy of a happy world were everything is fine. I personally bought some Hello Kitty mugs for fun a while ago because using them gave me a little jolt of happiness. Even though it is "commercial BS".

Substance is much more than money (to me). Why use the word if we all use it differently indeed...
Eh it's not a Ponzi scheme as far as influencers go. They are the new TV/movie celebrities and share fame with those groups. Plenty have been around for years now, sure there are rises and falls and fans are more fickle than in the past. Unless instagram and twitter and tiktok die soon they're going to be around a while, even though boomers and gen-x often greatly dislike them; those aren't the audiences they are trying to appeal to.
When you try to start your own channel on YouTube, or get viral in any form, you realise just how true of a ponzy scheme it is. Even original, well created work is hidden until someone grabs it and runs with it.

Easiest way for that to happen? Spam. Lots of spam. Instagram audiences aren't fickle, or at least they're not deterred by playing the game. They are extremely easy to offend though, and follow very serious moral trains. i.e. the head of the dragon changes directions fast but the body keeps in toe regardless of how they feel.

Just click on any of the top trending videos on Reddit or YouTube, you'll see a bunch of familiar faces saying the following. The more niche the topic, the more niche the channel spammers.

"Hey this video was great! Good job! -- seoulmetro walking tours".

I'm young so all these social outlets are what are left for me. But they're just way too tiring so I usually ignore the feeds.

You might not have noticed, but most people are into celebrities, musicians, actors, YouTubers, TikTokers, Instagrammers, etc. Famous people.

We evolved to compare ourselves to others in order to elevate the fitness of ourselves and our in-group. We look to peers. We take an interest in their gossip. We look at beautiful people, smart people, successful people. We're curious. We're also intrigued by scandalous and dangerous people. It's how we update our mental model of the world and contrast ourselves against it.

I'd be utterly shocked if you didn't, at some level, take an interest in the personal details of some person you don't know.

Virtual celebrities and influencers provide the same sort of dopamine hits that their real life counterparts do, except they can be 100% controlled by a corporation.

You don't even have to look at virtual influencers to know that people will eat them up: lots of people are already addicted to fictional characters. Movies, super heroes, even cartoons and anime. They find pieces of themselves in it. They start to care and take deep interest.

Virtual celebs are the future, and a cottage industry is already cropping up in support of it.

And just wait until people use AR/VR to optimize their own appearances and behaviors.

I'm not trying to argue that being a celebrity is a Ponzi scheme. The virtual ladder being incentivized on current social media platforms is. It's a game, being a celebrity is the end result but mostly people are just in it to make money.

I definitely don't connect to any celebrity other than those I admire for their intelligence and/or skill. So I'm not normal.

> It's a game, being a celebrity is the end result but mostly people are just in it to make money.

Being a celebrity is mostly a way to make money. And with influencers, the "career progression" that almost every upstart vlogger/instagrammer aspires to is pretty well-defined:

1) Provide some kind of entertainment content as a vector to push other people's crap

2) Over time, build up your user base and revenue base, until you become a well-known brand in a given market segment

3) Transition to pushing your own crap, e.g. a line of products with your name on it

Of course, selection is very strong there. Most would-be influencers spend years and fail at step 1. Few achieve step 3. And it can't really be any different, because attention economy is the ultimate zero-sum game: people have only so much hours in the day.

> "Influencing" is a ponzy scheme.

It's not. Influencers ask prices that are relative to how much money they'll make to a particular brand/store. That's why local influencers is a thing. And I was, too, surprised at how much sales they can drive for a simple store.

And what's wrong with marketing? Why is it bullshit? Marketing/influencing costs money, drives sales and drives adoption, all of which are very real and have a real impact in the society they're a part of.
Marketing is selling a product.

Bullshit is selling a lie.

Both are very inefficient and lazy ways to generate interest.

Shinhan (mentioned in the article) is a pretty big financial conglomerate. Twosome Place (in the second picture) is a well-known luxury coffee chain.

Financial chains in South Korea do not usually do influencer marketing. I would guess that the companies are buying not because of the influencer's reach, but because of the novelty of AI and the quality of the art. Presumably their goal is using the character and derived artwork in more traditional advertisements.

Edit: sanxyin confims that Shinhan used the character in a TV ad. I don't watch TV, so I guess I missed it.

I think I don't even believe she's AI generated.
60k followers isn't low either especially if it's organic. I have 976 posts and 343 followers :)
I suppose my point, knowing a few people with online followings, is that 60k followers does not translate to 800,000 marketing dollars as the article claims - or even 80k. Even if your had a really profitable niche (which she doesn't).

That being said I see the AI was in TV ads (is this influencer based? or just actor?) so 800k is possible.

"I looked around western social media and don't understand why I don't see what is popular in asia"
article is annoyingly uninformative about what “AI generated girl” actually means or represents here
just like 90 % of the kpop community
Hatsune Miku was there before.