|
|
|||
| forums: groups: | |||
|
"Is it A.I or is it Memorex?"
Love you, too
If you pie someone in the woods and they don't make a sound...do you still have to notify the park ranger?
Just wanted your attention for a question to ponder...assuming it's still legal to ponder in your area.
So, now that I have kinda sorta almost figured out how to create some semi decent A.I. WAM I was thinking about a question someone asked me a while back.To wit: (or in my case "half wit")...Did I remember every WAM session I had that there is no video for.
I said "Every WAM session?" I don't even remember a company Zoom call from yesterday. Maybe I have early onset piesheimers. (Apologies to those who have dealt with the real thing. That is not fun).
Anyway, his question gave me an idea. So about a week ago I simply tried to remember every "encounter" i had over the years that are NOT on video and began creating mini vids from those memories.
I cannot BELIEVE how close A.I. came in almost every instance so far in getting everything right. I did use the original faces, but geez I mean in some cases virtually perfect. (Yes, I know something can't be virtually perfect. It's either perfect or it's not).
So I showed one of my friends the vid and she freaked. "That's me but like 30 years ago" which is what I thought as well.
Keep that thought....
Now, what if I didn't have video of the WAM play, but DID have pictures? Like when vid cams were too expensive so for fun pics to look at we used a timer release which snapped pics every 30 secs or so.
So now I have these pics and of course I could upload them with my friend's permission.
But What if I don't upload them? What if I just describe them in a prompt?
And here's the question(s):
If I upload the original picture...then the original content isn't synthetic. But if I take a program and make it move? Is it synthetic? My personal thought was yes...but was curious as to others thoughts.
Then....what if I create a picture from my text prompts FROM the original picture and it looks virtually identical?
Then it's synthetic and the original content is real?
As you might have surmised (assuming you were enough of a dick to say "surmised"), I have created a number of vids now that were real content, but I made them move in video form with A.I.
To be honest, it's actually kind of freaky when it's someone I know. It's almost like a WAM time machine.
All of the above overlaps in my mind especially when I'm thinking of creating a "hybridvid" that will have A.I. creations and honest to goodness real people interacting.
Hey, look, I actually kept this missive short. (cough)
Robby
Just wanted your attention for a question to ponder...assuming it's still legal to ponder in your area.
So, now that I have kinda sorta almost figured out how to create some semi decent A.I. WAM I was thinking about a question someone asked me a while back.To wit: (or in my case "half wit")...Did I remember every WAM session I had that there is no video for.
I said "Every WAM session?" I don't even remember a company Zoom call from yesterday. Maybe I have early onset piesheimers. (Apologies to those who have dealt with the real thing. That is not fun).
Anyway, his question gave me an idea. So about a week ago I simply tried to remember every "encounter" i had over the years that are NOT on video and began creating mini vids from those memories.
I cannot BELIEVE how close A.I. came in almost every instance so far in getting everything right. I did use the original faces, but geez I mean in some cases virtually perfect. (Yes, I know something can't be virtually perfect. It's either perfect or it's not).
So I showed one of my friends the vid and she freaked. "That's me but like 30 years ago" which is what I thought as well.
Keep that thought....
Now, what if I didn't have video of the WAM play, but DID have pictures? Like when vid cams were too expensive so for fun pics to look at we used a timer release which snapped pics every 30 secs or so.
So now I have these pics and of course I could upload them with my friend's permission.
But What if I don't upload them? What if I just describe them in a prompt?
And here's the question(s):
If I upload the original picture...then the original content isn't synthetic. But if I take a program and make it move? Is it synthetic? My personal thought was yes...but was curious as to others thoughts.
Then....what if I create a picture from my text prompts FROM the original picture and it looks virtually identical?
Then it's synthetic and the original content is real?
As you might have surmised (assuming you were enough of a dick to say "surmised"), I have created a number of vids now that were real content, but I made them move in video form with A.I.
To be honest, it's actually kind of freaky when it's someone I know. It's almost like a WAM time machine.
All of the above overlaps in my mind especially when I'm thinking of creating a "hybridvid" that will have A.I. creations and honest to goodness real people interacting.
Hey, look, I actually kept this missive short. (cough)
Robby
Saturation Hall, Langstonedale
Lord of the Pies
From a UMD POV I believe that would count as synthetic, as while created from non-synthetic stills, the video would count as AI and not Memorex (whatever happened to them - I had lots of their audio cassettes once upon a time).
MM anticipated things like this hence the rules that AI images or video based on real people (whether created from a photo or from a prompt) can only be posted if full consent and model releases are in place. But as it looks like you have or can get those, you'd be able to post the videos here, as long as it's made clear that they are synthetic. I think it'd also be fine to sell them but check with MM to be absolutely sure on that aspect.
Personally, a rule I've imposed on myself is that even though I have signed releases for all of the people I've ever shot with, all of whom were fully aware it was fetish modelling and signed for "any legal form of trade" use, I won't generate any AI images or videos of any of them without obtaining a fresh and specific permission for that purpose. On the one hand it'd be an interesting way to "videofy" some of our very early photo-only scenes. But on the other, the models twenty years ago could never have imagined how things have developed since, and there are ethical considerations given people thought they were only consenting to still photography.
At this point I've no idea if AI WAM will eventually become part of the WAM marketplace or not, if it does I expect models will charge massively more to have an "AI Use" clause included in the release, as it'll allow the producer to go on churning out scenes of them, almost cost-free, with them gaining no additional income from it, so they'll need to front-load the charges to compensate.
How saleable it is will also depend on exactly what any specific buyer wants. If they are just after the visuals, then quality AI (which will cost rather more than today's experimental ones do) may well be viable. But there will always be those who don't just want to see a woman in a formal dress having her cleavage filled with custard, but want to know it actially happened, and that a real breathing human woman experienced all the sensations as the gloop was poured in. Of course your scenes created from actual photos during the taking of which a real woman genuinely did get messy, adds a kind of third level - it's both AI and real at the same time.
Video works through persistence of vision - you see a series of still frames taken very fast one after the other and your mind turns it into a moving image. So from one POV, AI doing the same thing but filling in the gaps between still images taken much more slowly is kind of the same but different? Absolutely fascinating area.
It'll certainly be interesting to see how things develop over the next ten years or so.
MM anticipated things like this hence the rules that AI images or video based on real people (whether created from a photo or from a prompt) can only be posted if full consent and model releases are in place. But as it looks like you have or can get those, you'd be able to post the videos here, as long as it's made clear that they are synthetic. I think it'd also be fine to sell them but check with MM to be absolutely sure on that aspect.
Personally, a rule I've imposed on myself is that even though I have signed releases for all of the people I've ever shot with, all of whom were fully aware it was fetish modelling and signed for "any legal form of trade" use, I won't generate any AI images or videos of any of them without obtaining a fresh and specific permission for that purpose. On the one hand it'd be an interesting way to "videofy" some of our very early photo-only scenes. But on the other, the models twenty years ago could never have imagined how things have developed since, and there are ethical considerations given people thought they were only consenting to still photography.
At this point I've no idea if AI WAM will eventually become part of the WAM marketplace or not, if it does I expect models will charge massively more to have an "AI Use" clause included in the release, as it'll allow the producer to go on churning out scenes of them, almost cost-free, with them gaining no additional income from it, so they'll need to front-load the charges to compensate.
How saleable it is will also depend on exactly what any specific buyer wants. If they are just after the visuals, then quality AI (which will cost rather more than today's experimental ones do) may well be viable. But there will always be those who don't just want to see a woman in a formal dress having her cleavage filled with custard, but want to know it actially happened, and that a real breathing human woman experienced all the sensations as the gloop was poured in. Of course your scenes created from actual photos during the taking of which a real woman genuinely did get messy, adds a kind of third level - it's both AI and real at the same time.
Video works through persistence of vision - you see a series of still frames taken very fast one after the other and your mind turns it into a moving image. So from one POV, AI doing the same thing but filling in the gaps between still images taken much more slowly is kind of the same but different? Absolutely fascinating area.
It'll certainly be interesting to see how things develop over the next ten years or so.
Saturation Hall - Forth! The Gungemaidens!
Ask me about WAM stories
DungeonMasterOne said: From a UMD POV I believe that would count as synthetic, as while created from non-synthetic stills, the video would count as AI and not Memorex (whatever happened to them - I had lots of their audio cassettes once upon a time).
MM anticipated things like this hence the rules that AI images or video based on real people (whether created from a photo or from a prompt) can only be posted if full consent and model releases are in place. But as it looks like you have or can get those, you'd be able to post the videos here, as long as it's made clear that they are synthetic. I think it'd also be fine to sell them but check with MM to be absolutely sure on that aspect.
Personally, a rule I've imposed on myself is that even though I have signed releases for all of the people I've ever shot with, all of whom were fully aware it was fetish modelling and signed for "any legal form of trade" use, I won't generate any AI images or videos of any of them without obtaining a fresh and specific permission for that purpose. On the one hand it'd be an interesting way to "videofy" some of our very early photo-only scenes. But on the other, the models twenty years ago could never have imagined how things have developed since, and there are ethical considerations given people thought they were only consenting to still photography.
At this point I've no idea if AI WAM will eventually become part of the WAM marketplace or not, if it does I expect models will charge massively more to have an "AI Use" clause included in the release, as it'll allow the producer to go on churning out scenes of them, almost cost-free, with them gaining no additional income from it, so they'll need to front-load the charges to compensate.
How saleable it is will also depend on exactly what any specific buyer wants. If they are just after the visuals, then quality AI (which will cost rather more than today's experimental ones do) may well be viable. But there will always be those who don't just want to see a woman in a formal dress having her cleavage filled with custard, but want to know it actially happened, and that a real breathing human woman experienced all the sensations as the gloop was poured in. Of course your scenes created from actual photos during the taking of which a real woman genuinely did get messy, adds a kind of third level - it's both AI and real at the same time.
Video works through persistence of vision - you see a series of still frames taken very fast one after the other and your mind turns it into a moving image. So from one POV, AI doing the same thing but filling in the gaps between still images taken much more slowly is kind of the same but different? Absolutely fascinating area.
It'll certainly be interesting to see how things develop over the next ten years or so.
MM anticipated things like this hence the rules that AI images or video based on real people (whether created from a photo or from a prompt) can only be posted if full consent and model releases are in place. But as it looks like you have or can get those, you'd be able to post the videos here, as long as it's made clear that they are synthetic. I think it'd also be fine to sell them but check with MM to be absolutely sure on that aspect.
Personally, a rule I've imposed on myself is that even though I have signed releases for all of the people I've ever shot with, all of whom were fully aware it was fetish modelling and signed for "any legal form of trade" use, I won't generate any AI images or videos of any of them without obtaining a fresh and specific permission for that purpose. On the one hand it'd be an interesting way to "videofy" some of our very early photo-only scenes. But on the other, the models twenty years ago could never have imagined how things have developed since, and there are ethical considerations given people thought they were only consenting to still photography.
At this point I've no idea if AI WAM will eventually become part of the WAM marketplace or not, if it does I expect models will charge massively more to have an "AI Use" clause included in the release, as it'll allow the producer to go on churning out scenes of them, almost cost-free, with them gaining no additional income from it, so they'll need to front-load the charges to compensate.
How saleable it is will also depend on exactly what any specific buyer wants. If they are just after the visuals, then quality AI (which will cost rather more than today's experimental ones do) may well be viable. But there will always be those who don't just want to see a woman in a formal dress having her cleavage filled with custard, but want to know it actially happened, and that a real breathing human woman experienced all the sensations as the gloop was poured in. Of course your scenes created from actual photos during the taking of which a real woman genuinely did get messy, adds a kind of third level - it's both AI and real at the same time.
Video works through persistence of vision - you see a series of still frames taken very fast one after the other and your mind turns it into a moving image. So from one POV, AI doing the same thing but filling in the gaps between still images taken much more slowly is kind of the same but different? Absolutely fascinating area.
It'll certainly be interesting to see how things develop over the next ten years or so.
Well said...and there is a third level....and my guess is that eventually there will be fourth level as A.I. evolves.
Robby
RobbyWLP said: But if I take a program and make it move? Is it synthetic?
Yes - the video has still been generated by AI.
A while back, I saw the movie "Alien Autopsy".
https://en.wikipedia.org/wiki/Alien_Autopsy_(2006_film)
The basic premise (spoiler alert!) is that there genuinely was an alien spacecraft which crashed in Roswell in 1947, and there was a real video made of the autopsy. However, that video degraded over time (i.e. the physical tape), so a couple of people re-enacted it and sold their version as the real thing.
Even if you believe in the spaceship, government conspiracy, etc., the video being sold was still a fake.
flank said:
Yes - the video has still been generated by AI.
A while back, I saw the movie "Alien Autopsy".
https://en.wikipedia.org/wiki/Alien_Autopsy_(2006_film)
The basic premise (spoiler alert!) is that there genuinely was an alien spacecraft which crashed in Roswell in 1947, and there was a real video made of the autopsy. However, that video degraded over time (i.e. the physical tape), so a couple of people re-enacted it and sold their version as the real thing.
Even if you believe in the spaceship, government conspiracy, etc., the video being sold was still a fake.
RobbyWLP said: But if I take a program and make it move? Is it synthetic?
Yes - the video has still been generated by AI.
A while back, I saw the movie "Alien Autopsy".
https://en.wikipedia.org/wiki/Alien_Autopsy_(2006_film)
The basic premise (spoiler alert!) is that there genuinely was an alien spacecraft which crashed in Roswell in 1947, and there was a real video made of the autopsy. However, that video degraded over time (i.e. the physical tape), so a couple of people re-enacted it and sold their version as the real thing.
Even if you believe in the spaceship, government conspiracy, etc., the video being sold was still a fake.
Crap, there goes my idea of pieing alien women.
Robby
Kind of...in my various short artworks (of late), I use my own material (drawings, photos) as 'seeds' which I then (via prompts and noise/prompt fidelity adjustments) 'animate' ('bring to life'). This seems similar to what you do with your wam pics. I call these artworks 'hybrids', being derived from (or evolved from) original human-made art...not fully human-crafted, but not completely 'synthetic' either (since the 'identity' or recognizable starter image came from my mind, my work).
Now, here on the umd, it seems that only the process used to make the final output is what defines 'synthetic'...So, that is how it will label your work. But rest assured, in the art world, such works are increasingly accepted and considered 'liminal', or 'in-between' the real and the artificial. Thus, they are more accurately called hybrids. It will take awhile for some to recognize and accept this distinction.
Now, here on the umd, it seems that only the process used to make the final output is what defines 'synthetic'...So, that is how it will label your work. But rest assured, in the art world, such works are increasingly accepted and considered 'liminal', or 'in-between' the real and the artificial. Thus, they are more accurately called hybrids. It will take awhile for some to recognize and accept this distinction.
'Our Favorite (Pie) Hits' now available!
I don't think it really matters what the prompt is; if you're using an AI programme to create (or modify) it, it's AI content.
That comports with my statement ('process')...but my point was not about the process, but the SEED (or starter image, 'pre-evolved', which is original to the producer). Of course, the model did not actually do what is being depicted, so, that is a concern (without permission).
So, I think the umd is correct in its labeling rule.
But, as a nuanced but significant distinction: most of the gen-ai models have been trained on actual human made artworks (in the millions), and sample a limited set of these (based upon a sampling algorithm), thus nearly every 'output' from these gen-AI engines is 'derivative' of human art (whether paintings or videos). One thus has to look at the word "synthetic" and recognize that what is being 'synthesized' is many human works via prompt-targeting algorithms (diffusion models or other)...and so, the word gets complicated in what it actually refers to or means.
I'm rambling...you do you.
So, I think the umd is correct in its labeling rule.
But, as a nuanced but significant distinction: most of the gen-ai models have been trained on actual human made artworks (in the millions), and sample a limited set of these (based upon a sampling algorithm), thus nearly every 'output' from these gen-AI engines is 'derivative' of human art (whether paintings or videos). One thus has to look at the word "synthetic" and recognize that what is being 'synthesized' is many human works via prompt-targeting algorithms (diffusion models or other)...and so, the word gets complicated in what it actually refers to or means.
I'm rambling...you do you.
'Our Favorite (Pie) Hits' now available!
I'm rambling...you do you.
Yeah, I hate when people ramble on. (cough) Apologies to Led Zeppelin.
Robby
Memorex...been using them to tape on VHS...then CD/DVD...from the pre internet days of trying to catch WAM on mainstream media...good times!
DungeonMasterOne said:
But as it looks like you have or can get those, you'd be able to post the videos here, as long as it's made clear that they are synthetic. I think it'd also be fine to sell them but check with MM to be absolutely sure on that aspect.
But as it looks like you have or can get those, you'd be able to post the videos here, as long as it's made clear that they are synthetic. I think it'd also be fine to sell them but check with MM to be absolutely sure on that aspect.
Be VERY careful with this.
First, not all "model releases" are created equal. This has been a subject of debate amongst the modeling community who have been now scrambling to update all their compliance paperwork on C4S. Many of them literally lost hundreds (upwards to 1000 or more) of videos off their existing online library that fell out of compliance.
There is also case precedent challenging poorly written or old releases (hence some platforms desire to change the terms of their compliance documentation). For instance, he might be able to share the photos fine. The minute he decides to toss it through AI, it's no longer a photo but rather a new product in which he would need to reach out to get an updated release. Is it likely a model will chase him down to have it removed? Don't know. Is it technically still a breach? Yes and no. It depends largely on the language of the original model release (and it needs to be an actual document and not something scribbled on a piece of paper) and it would be something I would recommend the host of the content to review with their attorney.
A large sum of the releases being rejected are both digital signature and just poorly written or missing in perpetuity so I would suggest that any documentation be taken to a lawyer to make sure it would stand any judicial test before going any further.
Making smut is easy. Making art takes more!
Saturation Hall, Langstonedale
Lord of the Pies
Nostalgic Erotica Prod said:
Be VERY careful with this.
First, not all "model releases" are created equal. This has been a subject of debate amongst the modeling community who have been now scrambling to update all their compliance paperwork on C4S. Many of them literally lost hundreds (upwards to 1000 or more) of videos off their existing online library that fell out of compliance.
There is also case precedent challenging poorly written or old releases (hence some platforms desire to change the terms of their compliance documentation). For instance, he might be able to share the photos fine. The minute he decides to toss it through AI, it's no longer a photo but rather a new product in which he would need to reach out to get an updated release. Is it likely a model will chase him down to have it removed? Don't know. Is it technically still a breach? Yes and no. It depends largely on the language of the original model release (and it needs to be an actual document and not something scribbled on a piece of paper) and it would be something I would recommend the host of the content to review with their attorney.
A large sum of the releases being rejected are both digital signature and just poorly written or missing in perpetuity so I would suggest that any documentation be taken to a lawyer to make sure it would stand any judicial test before going any further.
DungeonMasterOne said:
But as it looks like you have or can get those, you'd be able to post the videos here, as long as it's made clear that they are synthetic. I think it'd also be fine to sell them but check with MM to be absolutely sure on that aspect.
But as it looks like you have or can get those, you'd be able to post the videos here, as long as it's made clear that they are synthetic. I think it'd also be fine to sell them but check with MM to be absolutely sure on that aspect.
Be VERY careful with this.
First, not all "model releases" are created equal. This has been a subject of debate amongst the modeling community who have been now scrambling to update all their compliance paperwork on C4S. Many of them literally lost hundreds (upwards to 1000 or more) of videos off their existing online library that fell out of compliance.
There is also case precedent challenging poorly written or old releases (hence some platforms desire to change the terms of their compliance documentation). For instance, he might be able to share the photos fine. The minute he decides to toss it through AI, it's no longer a photo but rather a new product in which he would need to reach out to get an updated release. Is it likely a model will chase him down to have it removed? Don't know. Is it technically still a breach? Yes and no. It depends largely on the language of the original model release (and it needs to be an actual document and not something scribbled on a piece of paper) and it would be something I would recommend the host of the content to review with their attorney.
A large sum of the releases being rejected are both digital signature and just poorly written or missing in perpetuity so I would suggest that any documentation be taken to a lawyer to make sure it would stand any judicial test before going any further.
Very good points. In this particular case I gather Robby is still in close touch with one of the models featured and so would presumably be able to obtain a brand new "I consent to having my still images used to create AI content" release, hence "But as it looks like you have or can get those". Agreed that relying on old releases to cover usages that could not have been imagined back in the day would be dangerous. As I said, personally I'd only do it if I obtained, in advance, a new, specific, "I consent to AI usage" release, with the model's full informed consent of what was going to be done. Though at the moment I've no plans to use AI in my own WAM shoots, customers still like "real-people-real-events-that-really-happened" WAM so as long as it goes on selling that's what I'll be producing.
I just edited what will be next week's Langstonedale mud release over the weekend, featuring Isabeau absolutely throwing herself into the mudbanks on what was the very last day of Mud Season this year, Saturday 6th September. The sheer joy and happiness she shows as she gets utterly drenched in mud is way beyond anything even the best AI can produce. In terms of how I compare AI development with the steam locomotive, we started at 1804's "Puffing Billy" just a couple of years ago, and we're already passed "Locomotion No1" (1825) and "Rocket" (1832), and probably in the realms of the Stirling "single wheelers" (2-2-2s) of the 1850s. So that's 40+ years development in what, 2, 3 years? But there is a very long way to go before it's actually possible to produce full WAM scenes that completely and perfectly mimic human model behaviour.
Saturation Hall - Forth! The Gungemaidens!
Sponsors
VIP Coupons
Design & Code ©1998-2025 Loverbuns, LLC 18 U.S.C. 2257 Record-Keeping Requirements Compliance Statement
Epoch Billing Support Log In