EroScripts is a community for scripts that sync interactive sex toys with porn. Learn more.
To find and download scripts, create an account.
Just a fun question. Assume the AI can script anything and everything you throw at it just as good as a hand scripted one in seconds. You can keep the AI forever. All your favorite videos will be scripted.
Very interesting question indeed.
The answer is not very intuitive, since it would realistically save me thousands of USD I might end up spending on scripts in my lifetime. But it doesn’t actually feel worth that much?
When I look at it in practice, a service like SLR offers lifetime newly created (handmade and AI) scripts for 300 USD with the current discount and that’s not enough to be worth it for me.
This hypothetical AI would be worth much more since it’d actually make all the scripts I’m interested in while SLR might make 2 scripts I care about per year (in my personal case).
Free scripts on EroScripts provide so much value that it completely throws off the calculation. The best scripters on here release the majority of their library of work for free only to contribute to the community.
I thought it was an interesting question to gauge how much people care. For people who have Terabytes of porn, I would imagine they wouldn’t mind paying $1k+ to have them all scripted in a day for their enjoyment.
Theoretical price is high as it takes tons of time scripting manually. If this is made automatically then scripting manually gets worthless. This will come eventually. If the SLR script maker would be public no one would buy 5$ scripts.
how good is the SLR script maker though?
I’d say its way better than an amateur scripter, but not as good as an elite scripter
This is a dependent kind of question.
What comprises the software, is it opensource?
Is it an online service? Can I run the AI locally?
what kind of horsepower does the model need?
If it requires a lot of horsepower and it’s an online service due to needing datacenter hardware, probably low cost recurring revenue would be fair.
If I can run it on my machine, maybe like $50 onetime purchase with the software + model.
$100 for the same plus additional features probably.
For ex, if there were 1 software and I could buy a couple models that are optimised for different things (action-based / music-based / Hybrid PMV-HMVs / cock heros / etc.)
probably $20 per model etc.
If mine ever works, I will give it away. The technology is there, it’s more a question of the right model and the right amount of training. Script generation ought to be something a single GPU can do.
But… An app based on one of this models to clean up the AI’s mistakes would be cool, but isn’t that pretty much any existing scripting tool?
I was actually just thinking about this as a personal budgeting exercise in anticipation of the release of SLR Drive and its supposedly related capabilities.
I think the biggest qualifier for me is that I would never actually be able to “use” all those possible scripts. I don’t have the time nor the sexual fortitude.
On the other hand, my libido craves novelty so very rarely do I revisit scripts more than a couple times. Having a virtually unlimited and unending supply would give me the variety I desire.
This is a personal budgeting consideration, but I think I’d be cool with $10 a month.
The idea of “lifetime” adds some complications to the equation. How long before technology bypasses the concept of funscripts as we know it now? How much longer would I even find it useful?
So maybe need to use a shorter payback period. I think I’d settle on 1 year/$120. That’s about the same as I paid for lifetime SLR scripts.
I have a couple TB of porn and I would be willing to spend a couple hundred on it as one time payment. If it is a subscription, I would simply ignore it completely.
However it would have to be something that I can run on my own hardware, without needing any cloud service. Open Source would be highly prefered, but then again, I can’t program for shit, so I wouldn’t even be able to tell if the AI itself is any good.
The question I would have is, what is an accurate script? Outside of PMV/HMV content, probably nothing is really accurate. From my very limited experience of scripting, the actual stroke length doesn’t 100% match movement on screen. And that is often times intentional.
There are different styles of scripts that people prefer and different styles of scripts that scripters make. For example intense climax, or Vibrations in the script to make it more intense and so on.
So instead of setting the expectation or goal to “100%” accurate, I would prefer to good scripts with high accuracy. If it could be customized using prompts for example it would be even better.
90% of it is getting cowgirl right with the girl(s) in the scene locking eyes with me. What’s messing up my work so far is that older scenes are framed differently and I’m predominantly interested in actresses and scenes from 6-10 years ago that never got scripted in the first place.
It’s meant as a fun question without getting too technical. Let’s say the AI will be able to provide a script that is perfect to your liking or perfect to what’s happening in the sex scene. It’ll always be a good script to your preferences. How much would you pay for that?
If someone could write an AI to script videos accurately, they could probably create configurations to stylize scripts, incorporating things like acceleration into exaggerating the strokes, handling tongue parts, bottoming out on deepthroats, etc.
Doing these things wouldn’t even be too hard to implement code-wise as a way to apply these adjustments to a base “accurate” script.
Sounds nice, how far are you with it?
Since this is a imaginary exercise, I won’t over complicate thinking about it…
For me, being an old fart and a hoarder of my personal “golden age of porn”, the thought of my entire archive being scripted perfectly… that would be heaven on earth. I’d happily spend thousands of dollars for that to happen.
RIGHT WITH YOU BRO!
Just FYI SLR offers lifetime with scripts for $400, $300 is the price without scripts.
For me, 100% accurate isn’t always what’s best in a script. For instance, hybrid PMV scripts are some of my favorite because the scripter gets pretty close to the action but also takes some liberties to make it more interesting.
Clearly, there’s still a ton of value there because the AI could make the base script and scripter could add personal touches from there.
It is hard and hard work to match a script to the action. It is easy and fast to unmatch it even if it means interesting stuff as you say. So the goal and the five star script is a matching one. That’s why AI is mostly four stars as it’s mostly incomplete where there are angles the AI can’t understand or action it does not see.
this:
I would favor this rating:
one star: Finished (all action scripted)? yes/no
one star: Endpoints matching? yes/no
one star: Well exaggerated? yes/no
one star: Advanced scripting (scripting speed change)? yes/no
one star: Speed sub 600U/s ? yes/no (devices can follow)