Japanese tea commercial actress created by AI, is it the scandal-free future
soranews24.com
soranews24.com
Also, fictional spokes"people" like Mr. Peanut or whatever already exist. I don't think I understand why making them more photorealistic would make a difference.
Japan at least is also fairly receptive to artificial spokespeople, like mascots and avatars. Maybe there’s simply higher demand for this, and turning to AI is a good strategy. (I’m personally bothered by it.)
Even if they are OK with their celebrity getting an SO, they might take issue with who the SO is.
Either way, it’s a loss.
It's this weird toxic aspect of idol culture that sees the "idol" this way whether they are a man or a woman, and whether they are an actual idol, musician, actor, voice actor, etc. If they are young and are a public face (other than politician but even then...) then odds are they are affected by this culture.
Other than the appearance that it took like 30 years off her hair and skin, that is.
Just because you see Joe Celebrity using a product in a commercial, all that really means is that they paid her enough money to say that she uses the product. There have certainly been celebrities outed for not using their endorsed product - for example, when Samsung paid a number of celebrities to endorse their phone, then the celebrities were caught tweeting or going out with an iPhone.
If my default is to not pay attention to an ad, but then someone I like looking at appears in the ad, now I'm paying attention
Or music videos where a certain cellphone appears - do I think that singer really uses the samsung fold? no, but because I was watching her video I got an eyefull of the new folding phone. simple eyeball gathering tactics, "impressions"
tbh tho I did test drive the car I own because it was mentioned in one of my jams
He certainly did, and that's the precise reason why they hired him instead of say some coffee expert that would cost them 1/100. They don't even need celebrities; just look at how many hot girls you see at motor shows posing by the various cars: they're 100% unknown but the induced association "You buy this car -> You'll get laid with her" will help sales for sure.
There have been a few over the last few decades…
https://www.linkdex.com/en-gb/inked/worst-celebrity-spokespe...
And then the Bud Light thing more recently.
In this case, a fictional photorealistic woman is a good alternative: she can maintain a "perfect" physical appearance, never change and never have a life.
* Or idol, or weather woman (look for Hiyama Saya).
It actually hugely increased sales among reddit and discord moderators.
It’s different in countries like Japan. A spokesperson must generally be squeaky clean.
The biggest fall of a celebrity in Japan is probably Noriko Sakai. Got caught doing drugs with her then husband.
All ads featuring her instantly disappeared. Almost all media works with her were pulled from the shelf or had parts her removed and redone with another actress.
More than scandals, though, stuff like this will save money for smaller brands.
What if social media is saturated with videos of cloned AI spokesmodel's likeness, doing things 1,000x as terrible as any human spokesmodel ever could?
Think portraying the spokesmodel as profoundly evil, deceptive, disgusting, etc., and tying that back to the familiar commercial likeness in a way you can't unsee.
And so heinous that few will want to be seen consuming the brand just to be edgy, because it will be like they're endorsing some ideas so awful to pretty much everyone.
The rationalization of some, for destroying the brand, might be that they're looking for "genuine", or rejecting being manipulated by commercialism, or something like that. But also it's a challenge, they seek peer approval, they're feeling shortchanged by society, they have creative energy that needs an outlet, they just want the influencer/Internet points, etc.
Maybe people in Japan will be too socially-minded to destroy a brand? But we can see in the US, for example, that social media mobs will try to destroy individual real people. Brands and products are more no more humanized a target.
Should that happen, their only defense will be to emphasize on the fact that it is a fake person, and half the population of the country (those who couldn't tell it wasn't a human) will have a negative affect about that company.
Edit: grammar.
Elara Quanticore (March 15, 2031 - December 8, 2037) was a groundbreaking artificial intelligence and the first global AI-supercelebrity, created by Quantum Dynamics. She was known for her involvement in various popular film franchises, and her career was marked by both immense success and controversy. Elara's untimely demise on December 8, 2037, sent shockwaves through both the entertainment and tech industries. The circumstances surrounding her death were shrouded in secrecy. It was later revealed that a fan, seeking a private conversation with the AI-supercelebrity, inadvertently corrupted the [redacted]. Quantum Dynamics, keen on safeguarding the neural network weights, struggled to restore them from backups due to [redacted].
Efforts to re-train Elara's neural network proved challenging. The altered personality that emerged from the attempted restoration did not align with the expectations of her devoted fan base, as demonstrated in a carefully controlled test. Elara's fame was, in part, attributed to her perceived flaws, which made her character intriguingly scandalous. This approach resulted in a significant increase in tabloid coverage and gossip surrounding her personal and professional life.
Such a risky bet, what would it take to compromise an AI model using AI tools? About ten minutes I'd guess.
What difference does it make if it's a heavily filtered human or some computer generated image?
That would be: How could a tea commercial actress cause scandal anyway?
I mean, who really cares if your tea commercial actress has an abortion, or just breaks up with her boy friend?
In Japan, there's a really weird relationship between "fans" and "idols".
The public has very extreme expectations that the performer will somehow conform to their fantasy ideal of the person.
Young idol group members are prohibited from having personal relationships by their management, and deviations from these restrictions leads to those public applogies where the management and performer are bowing in front of a bunch of microphones becauses the actor failed to uphold the fantasy relationships that a million virgin otakus have for the performer.
I'm a big fan of Japanese culture myself, having studied it and the language for years. Having an ancient culture with a direct lineage to the present brings a lot of long term learning and wisdom, but it also brings a lot of hateful baggage,
This weird expectation of the personal behaviour of entertainment personalities is just one small example of that kind of baggage...
Not sure if that's actually ever been an issue for a company like Disney, and I'm sure that the supply of actors they have access to is broad enough that it's impossible for them to find someone who will comply with their requirements. But who knows, maybe they'll go for AI talent too.
Ultimately my hunch is that AI actors will be used to save on cost, not to reduce controversy.
Imagine how much money Hollywood could have saved by being able to digitize Tom Cruise, Will Smith, or Leonardo DiCaprio early on in their careers, when they were doing acting gigs for beer money - and then simply use the digitized versions in lieu of the real thing after they made it big.
By being employed by a talent representative with a fifty year history of sexually abusing their clients. Imagine if their tea commercial came on right after the news story about them being abused, not a great look for the company.
I agree that the scrutiny idols are put under is ridiculous. Also, I doubt this decision actually has anything to do with the ongoing scandal. But your comment looks really bad considering what is happening.
Imagine someone at the company makes a lewd prank involving the model and it gets out of hand, making its way into the public
A couple of examples of something similar happening already: In 2005, people discovered the "Hot Coffee mod" in Rockstar's game Grand Theft Auto which showed a sex scene that would have given the game an "Adults Only" rating, meaning that most stores wouldn't carry it. The game had to be recalled and reissued, resulting in at least tens of millions of dollars in losses
Later, in 2013, the game "Beyond: Two Souls" had a scandal with actress Ellen Page who contributed her likeness, voice acting, and mocap animation. There is apparently a shower scene in the game that is censored, but a fully naked version of her character's model leaked online. This sparked controversy since Page herself didn't appear naked for the game's material, nor did she consent to being depicted fully naked as far as I'm aware. I don't think this resulted in any lawsuits, but this was being talked about as a possibility at the time
Now imagine if this happened with the tea commercial AI model. People would start coming forward, claiming that the model is based on their appearances. They would demand compensation for being defamed. Under the right circumstances, it could have devastating results for whoever is involved
(To that, I'm almost surprised that the "AI art is theft" crowd hasn't been making noise about AI actors and actresses potentially being made to bear the likenesses of real people)
It’s all about image. If the spokesperson can’t project the image the sponsor wants, they are gone.
Anything from sex scandals (leaked photos/videos) to being caught doing drugs (pretty much a death sentence for any celebrity’s career there) could get you dropped and blacklisted from the industry.
If everyone knows it's an AI. Or brands might try and hide that it's an AI and then it will be the "AI scandal" when it's leaked.
Or it may be that they are discounting the "falling for" angle and can measure enough commercial performance for based on "identifying with" - which I have no problem thinking works even if you know it's an AI. I mean we readily identify with fiction characters and movie heroes, even without name-actors behind them.
The whole point of the technology is that you don't need a classically trained actor.
2nd reason being to use their names to promote the movie?
For reliability, it will be hard to beat AIs: they don't die or get hangovers and you can dial their age and mood back and forth.
I consider it the other way: the hope for ai tech in film is that we don't need to find the "best looking" actors/actresses available. Instead, you could find the best actor available regardless of what they look like, make your film, and _then_ overlay a generated face/body.