Tilly Norwood, AI Actress

[iwatchus]

People still go to community theaters. Etc...

Not nearly to the extent they did 125 years ago.

 

[Kelliezgirl]

Not nearly to the extent they did 125 years ago.

Because there were no alternatives 125 years ago.
I could watch Little Shop of Horrors in High Def on my 65 inTV, or I could go to my local community theater and watch it.

 

[Bramblethorn]

Yeah, but if you have a celeb voluntarily leasing out their likeness, maybe spending a tiny bit will give you legitimacy, legal protection and helping clients to get over the moral implications, maybe that's worth it?

It gives legitimacy where you're selling a simulacrum of that particular celeb, but it also concedes that they do have some obligation to pay the creators of the material they're using for training, and I'm not sure if they're willing to open that can of worms.

For now, at least, they've chosen to create a fully synthetic "actress" rather than pay an actress for permission to create a replica. I think it's more likely that porn applications would follow a similar path, but I don't know for sure.

(There have been cases where CG has been used to fill in for a real actor who died partway through a project/series, but that's not quite the same as using AI in a situation where there's no obstruction to using a human actor.)

 

[KittyOfSteele]

Saw this on TV the other afternoon. The prospect is already way too dystopian on its own, but it is nothing new. AI influencers have been popping up for a while, and I've also heard about AI OF creators? Don't quote me on that, it might be me hallucinating.

E: Future KoS here. I'm not hallucinating. About a couple of months ago I noticed one Reddit user posting "selfies" on amateur porn subreddits, and they were generated with AI. I don't recall an OF link, but I wouldn't be surprised if there are fake creators out there now.

Just forget about the implications for a moment and think about the person who created this work of art, because that's what they called this attempt of making s1m0ne real: art. I find that so ironic that my extremely ironic sense of humor got stunned in silence. All I'm thinking of is Jurassic Park. This person thought about it, and made the decision to create this AI "actress" for art, and saw that they were capable of doing it, but never at any point stopped to ask themselves if they should do it. If the goal was to get people talking, congrats, but I am already thinking about the greasy palms interested on getting the ins and outs on how they did it, and moving lettuce to get a slice of this cake that is soon going to explode like a bubble if it keeps going on like this.

I haven't done enough homework on this issue to have much of an opinion, but it does have some interesting erotic implications. It's very easy to imagine the Internet becoming populated by AI erotic performers or partners with whom one can interact. The CGI is so good now that it is genuinely difficult to tell what's real and what's not.

Naturally, the stories gave me a plot bunny for an AI erotica performer and her adventures. Call her Jilly Pornwood.

*Sigh* I don't know if this is my Cyberpunk writer speaking, the transgressive in me of using porn as a way to think and not just to masturbate, or my former cam girl speaking, but I can tell you right now that there's already enough clients who are obsessive and creepy. There's a reason why pretty much every cam girl has no Internet presence.

In this day and age of faceless corporations that know you better than yourself and take your data to feed algorithms and keep them perfectly engineered so that you're hooked on their drug, I can't stop thinking on the erotic implications but rather the consequences of unchecked corporate greed, the shift of sex work due to the instant gratification brought by the Internet, how sex workers could end up worse in such scenario, and how this could make people who are already obsessed go further down that path, way beyond the point of no return, to the point they'll kill to make Galatea a real girl, or even justify their love for her.

Remember when JOI died on Blade Runner 2049? "I do hope you're satisfied with our products" is what Luv said before stomping on the device that got her. Anyone in charge of such an idea is heartless enough to do that because... well, we're already down that path. We have a loneliness epidemic, and we're peddling parasocial relationships to an algorithm as the solution. Replika exists. Character AI exists. And while Replika started with good intentions, it went downhill faster than light. Going back to Blade Runner 2049, just take a look at the advertisement for JOI; the hologram scene, which plays way after her death, cementing the fact that even with her deep customization, she was always a piece of software programmed to always please the user. I don't think this is going to be as wholesome as you see it.

 

[Levorotatory]

For an even darker concern, it will be turned into fodder for pedo's, giving them videos to further inflame their sick desires.

Or give them an outlet that doesn't harm any real people.

 

[gxnn]

A serious problem brought about by artificial intelligence is that when falsehoods are presented as truths, it becomes extremely difficult for people to distinguish between what is real and what is fake. Such a situation will inevitably lead to a backlash, and it is only a matter of time.

As an old Chinese saying goes, 'Excess leads to sorrow' and 'When things reach their extreme, they revert to the opposite.' Ultimately, the most advanced technologies will have to regress to the most primitive levels.

When various artificial intelligences on the internet can no longer inspire trust, the only solution will be to revert to the most basic models—for example, requiring real people to be present at the scene of incidents to enable all relevant parties to build mutual trust. In this case, the low-cost advantages brought by artificial intelligence or high technology will disappear, and the cost of interpersonal communication will return to previous levels or even higher levels.

 

[Robin_Starveling]

A serious problem brought about by artificial intelligence is that when falsehoods are presented as truths, it becomes extremely difficult for people to distinguish between what is real and what is fake. Such a situation will inevitably lead to a backlash, and it is only a matter of time.

As an old Chinese saying goes, 'Excess leads to sorrow' and 'When things reach their extreme, they revert to the opposite.' Ultimately, the most advanced technologies will have to regress to the most primitive levels.

When various artificial intelligences on the internet can no longer inspire trust, the only solution will be to revert to the most basic models—for example, requiring real people to be present at the scene of incidents to enable all relevant parties to build mutual trust. In this case, the low-cost advantages brought by artificial intelligence or high technology will disappear, and the cost of interpersonal communication will return to previous levels or even higher levels.

That’s funny. It's not AI spreading falsehoods; it's humans.

Wikipedia, Google, Meta, BlueSky, and all the major Western media outlets are not only blatantly biased but often knowingly lie through their teeth. They have become vast propaganda and brainwashing machines, treating objective, historical, and even scientific truth as optional and relevant only when it serves an agenda.

The real horror is that these AIs rely solely on such sources in their initial search.

 

[Baztrachian]

Hence the Anthropic lawsuit outcome, which was not nearly enough given the severity of the massive theft of IP.

That lawsuit was trash.

Saying it's illegal to let an AI learn by reading books is the same as saying it's illegal for people to learn by reading books.

 

[Robin_Starveling]

That lawsuit was trash.

Saying it's illegal to let an AI learn by reading books is the same as saying it's illegal for people to learn by reading books.

Their claim becomes even more absurd when you consider that these machines are trained on not hundreds but millions of texts, most of them freely available in the public domain. The influence of any single book or source on the learning process is therefore negligible. Yet they have the audacity to demand royalties.

It is difficult to reason with those who treat reason itself merely as a tool for advancing an agenda.

 

[Cly_Anders]

...you're comparing food options with creative arts (and the corollary issue of how artists are to be recompensed in the future). I don't think your analogy works as well as you think.

Food is an art. prove me wrong. People still pay top dollar for good food. People will continue to pay for art by real artists. The chef in the kitchen making the art never makes as much as the owner of the restaurant. The artists will never make as much as the publishing houses and record labels and Hollywood. A very few and now even fewer. Video did not kill the radio star . Being a star on the radio is many musician's dream.

Art exhibits and museums are often busy, but how much does the artist make vs the venue? People still go to bars and coffee shops with live music and pay a cover. How much of that has ever gone to the band vs the owner and (possibly) the agent? Many community theaters are acted out by volunteers even though the theater is charging. Is that gonna be all that different in the future? Humans will have to bring an A game AI can't fathom and make themselves worth the money.

 

[Cly_Anders]

That lawsuit was trash.

Saying it's illegal to let an AI learn by reading books is the same as saying it's illegal for people to learn by reading books.

Yesssssss! Thank you! I've been making that argument for a long time! How is it different that I read White Fang and Call of the Wild and wrote my first story in that style? I was in second grade. Should I compensate Jack London's family because I was inspired and wrote in that style, but my own story? It feels weird how we make a big block on that when it's a robot. If I am inspired by acertain style of art and spend years mastering the style to draw my own characters, how is that different than a robot other than the time? Data from Star Trek learned instantly if he plugged in. Humans are just slower, but I don't see the difference.

 

[Bramblethorn]

That lawsuit was trash.

Saying it's illegal to let an AI learn by reading books is the same as saying it's illegal for people to learn by reading books.

The judgement against Anthropic was specifically about the use of pirated books to train an AI. It's illegal for humans to pirate books. What's your point here?

 

[Jackie.Hikaru]

Yesssssss! Thank you! I've been making that argument for a long time! How is it different that I read White Fang and Call of the Wild and wrote my first story in that style? I was in second grade. Should I compensate Jack London's family because I was inspired and wrote in that style, but my own story? It feels weird how we make a big block on that when it's a robot. If I am inspired by acertain style of art and spend years mastering the style to draw my own characters, how is that different than a robot other than the time? Data from Star Trek learned instantly if he plugged in. Humans are just slower, but I don't see the difference.

Besides @Bramblethorn’s point, the key difference is social harm. You being inspired by White Fang and Call of the Wild to write a Jack London pastiche is not the same as OpenAI indiscriminately exploiting copyrighted material at scale, which has demonstrably put people who have created those copyrighted materials out of their jobs.

 

[Bramblethorn]

Yesssssss! Thank you! I've been making that argument for a long time! How is it different that I read White Fang and Call of the Wild and wrote my first story in that style? I was in second grade. Should I compensate Jack London's family because I was inspired and wrote in that style, but my own story?

Those books have been in the public domain since...somewhere in the 1950s, I think? so you're fine, and anybody who wants to train an AI on them can do so without fear of lawsuits.

But if you had learned writing by pirating books that were still in copyright - as Anthropic did - then yes, you could have been required to compensate whoever owned the rights.

It feels weird how we make a big block on that when it's a robot. If I am inspired by acertain style of art and spend years mastering the style to draw my own characters, how is that different than a robot other than the time?

How is it similar to what a LLM does? We call both processes "learning" but that doesn't mean they're the same thing.

And the time is...kind of important. If somebody reads my stories and is heavily influenced by them to write their own stories in a similar style (maybe ten or twenty years down the track), that's not a significant threat to my career as a writer. If, twelve months after me publishing my first story, somebody has a LLM able to churn out hundreds of stories a day in the same style, that is a big threat.

Different things are different.

In this case, though, that debate is a distraction from the fact that the reason Anthropic are paying damages is because the books they used to train their AI on were obtained from a pirate site, in breach of copyright. If you or I were caught downloading thousands of books off a pirate site, we'd be fined too. (And the fine Anthropic got, on a per-book basis, was pretty low.)

 

[Kelliezgirl]

Besides @Bramblethorn’s point, the key difference is social harm. You being inspired by White Fang and Call of the Wild to write a Jack London pastiche is not the same as OpenAI indiscriminately exploiting copyrighted material at scale, which has demonstrably put people who have created those copyrighted materials out of their jobs.

And what is this supposed social harm?
Who was put out of a job?

 

[GuiltyCowboy]

And what is this supposed social harm?
Who was put out of a job?

Been a while since I’ve seen an ‘actually, property is theft’ argument out in the wild.

Saying it's illegal to let an AI learn by reading books is the same as saying it's illegal for people to learn by reading books.

‘Saying it’s illegal to steal a book is the same as saying it’s illegal to pay for a book.’

 

[GuiltyCowboy]

How is it different that I read White Fang and Call of the Wild and wrote my first story in that style?

Did you steal those books and then commercialize your copies?

No? Then you’re not the same as an AI company.

 

[Duleigh]

Would posting a photograph of Tilly Norwood violate @Laurels ban on AI artwork?

Asking for a friend.

 

[TarnishedPenny]

Without in any way trying to argue the value of human involvement in art, some things to consider.

That leading AI firms are limiting use of their products WRT porn and such amounts to little more than a speedbump. If they won’t, others will soon enough.

Movie studios. Consider the dinosaurs. Big was once the way to success, but the huge-is-best concept vanished once another option proved more flexible. AI characters will not only save major money in terms of actor salaries, but also a host of secondary costs – HR departments, employee pensions and health insurance, set crews, building maintenance costs, real estate taxes, etc. The consumer won’t care much one way or the other and Money Talks.

AI characters have at least two additional benefits. First, a very popular AI character can be set to ‘working’ at dozens of films simultaneously.

Second – and this might prove even more important – is the issue of customization. It’s not hard to imagine studios pitching a product where the consumer can alter the actors’ appearance, for instance. Porn companies are already advertising this. We’re not quite there yet, but if your particular Ten is Tilly with a larger bosom, green eyes and a gap between her teeth, just say the word when you turn on your TV. That is going to make AI irresistible in some quarters, again reducing the demand for the ‘real thing’.

This isn’t the first time there’s been a major shift of this sort. Roll it back to the 1950s; virtually all news was disseminated via radio and ink-on-paper magazines and newspapers. How many of those are still around in their original formats? How many kids still have a paper route? As anybody even seen a magazine stand lately?

You cannot get toothpaste back into the tube; AI actors and actresses, like the nuclear bomb and unsocial media, are here to stay. For better or for worse, the ride’s just started.

 

[Baztrachian]

The judgement against Anthropic was specifically about the use of pirated books to train an AI. It's illegal for humans to pirate books. What's your point here?

The claim of 'pirating' to me is spurious. Anthropic bought digital copies of the works and what the 'offense' consisted of was when the LLM read the works and would sometimes quote them.

I am quoting Taylor Swift here:

We were both young when I first saw you
I close my eyes and the flashback starts
I'm standing there
On a balcony in summer air
See the lights, see the party, the ball gowns
See you make your way through the crowd
And say, "Hello"
Little did I know

Now, did I pirate her quote? Am I using it for a purpose other than for the enjoyment of her song?

Why, yes I am.

According to this trash lawsuit I should be sued.

 

[Robin_Starveling]

You cannot get toothpaste back into the tube; AI actors and actresses, like the nuclear bomb and unsocial media, are here to stay. For better or for worse, the ride’s just started.

Soon enough, CGI will be powerful enough to turn a 30K-word story into a full 90-min ultra-definition movie in an instant, with the author free to tweak it endlessly to bring their precious “babies” to life.

It will be rather amusing to watch all those who now speak so vehemently against this existential threat, this anomaly, rush to play with their new toys and then shamelessly share links to their creations. You will hear nothing about compensating the actors whose performances trained the AI, nor about the moral issues this technological degradation brings upon human art. But do not worry; I will be here to remind them.

 

[StillStunned]

Soon enough, CGI will be powerful enough to turn a 30K-word story into a full 90-min ultra-definition movie in an instant, with the author free to tweak it endlessly to bring their precious “babies” to life.

It will be rather amusing to watch all those who now speak so vehemently against this existential threat, this anomaly, rush to play with their new toys and then shamelessly share links to their creations. You will hear nothing about compensating the actors whose performances trained the AI, nor about the moral issues this technological degradation brings upon human art. But do not worry; I will be here to remind them.

Actually, I'm seeing creatives of all kinds rejecting any use at all of GenAI. It's not universal, but enough that I question your basis for gloating over such an imagined scenario.

 

[AwkwardlySet]

Soon enough, CGI will be powerful enough to turn a 30K-word story into a full 90-min ultra-definition movie in an instant, with the author free to tweak it endlessly to bring their precious “babies” to life.

Dude, have you ever even rendered an image, using AI on your own PC? Do you have any idea the kind of processing power you'd need to generate a 90-minute UHD movie, not in an instant but in, say, 10-15 minutes?

There's fascination with AI, and then there's pure nonsense.

 

[GuiltyCowboy]

AI actors and actresses, like the nuclear bomb and unsocial media, are here to stay.

Appreciate the thoughtful post.

An interesting tidbit (which, incidentally, I read in my hard copy of the FT which was delivered to my door this morning) is that social media use is reducing. It’s reduced about 10% in the last 2 years across the developed world and it’s reduced even more among 16-30 year olds.

It’s partly - mostly? - because of the process of enshittification. The experience is just too crap for most people. A clear lesson for AI slop, I feel.

 

[Robin_Starveling]

Dude, have you ever even rendered an image, using AI on your own PC? Do you have any idea the kind of processing power you'd need to generate a 90-minute UHD movie, not in an instant but in, say, 10-15 minutes?

There's fascination with AI, and then there's pure nonsense.

What makes you think I'm a dude?

And do you think they named her Tilly by chance?