Hey y’all! 🤩

The past several months I’ve been fucking around with Stable Diffusion and, as you might’ve guessed, saw the value in NSFW image generation.

This idea is totally off the cuff, steal it too if you’d like:

What if we marketed and created an entirely ethical porn website (guaranteed) that started with AI generated porn images and gifs and eventually moved to fully AI videos?

A lot of people are at least somewhat concerned with the morals of traditional porn websites (including OF and their creator ‘age issues’) so why not guarantee ethical and legit porn via AI?

In theory you could generate whatever preferences you have (perhaps even allowing the users to specify their own preferences to reference a cache/db or generate on the fly)… right?

What are your thoughts?

  • lialdd@lemmynsfw.com
    link
    fedilink
    English
    arrow-up
    10
    ·
    1 year ago

    I think ethics will also have to grapple with the fact that the thousands of humans who created the training data, often without their knowledge or consent, aren’t compensated and not only didn’t consent to their likeness or work being included but also can’t effectively request to be excluded. I see plenty of Emma Watson or other celebrity lookalikes in the same models I’ve been using, and I don’t think you can call yourself ethical while using tools that contain such stuff. No more ethical than the ethics of a person choosing to take a nude photo of themselves and charge money in order to see it, certainly.

    • pornthrowaway2@lemmynsfw.comOP
      link
      fedilink
      English
      arrow-up
      1
      ·
      1 year ago

      This is unfortunately a sad fact of reality but one that I had indeed recognized as an issue - this website would need to use a custom trained model using consenting models/artists/performers/workers/etc.

      There is forever a very important contingency to this plan - if you sign your consent to this sort of thing, you are technically giving the company (my hypothetical company) rights to your image and to make transformations to it.

      See the first episode of the new Black Mirror season for reference.

  • Rreeddeerr@latte.isnot.coffee
    link
    fedilink
    English
    arrow-up
    8
    arrow-down
    2
    ·
    1 year ago

    A concern I have is that ai generated women/men are too perfect. You speak of ethical porn. Are unrealistic people something we should be aiming towards to? The bar is already set too high by actors and models, I don’t want to know what this perfect porn could cause to the youth in terms of self esteem, or interpersonal relationships

    • PutItOnMyBill@lemmynsfw.com
      link
      fedilink
      English
      arrow-up
      10
      ·
      1 year ago

      Are you telling me that being invited in to bang an absurdly hot mother and daughter duo while I deliver a pizza is unrealistic

    • Ookami38
      link
      fedilink
      English
      arrow-up
      7
      ·
      edit-2
      1 year ago

      The unreal expectations are always going to be there. Right now, anything less than perfect gets airbrushed, photoshopped, make-upped, etc, and you never know. At least with a website like OP mentioned, it’s laid out from the start - these are fake people created to an impossible standard.

    • pseudoanonymouse@lemmynsfw.com
      link
      fedilink
      English
      arrow-up
      2
      ·
      1 year ago

      I think that it would be an interesting exercise to try and force AI generation of more realistic body types. Since so many of the AI companies are talking about re-ingesting AI-generated content for future training cycles it could even be seen as a way to build in that type of content for future versions of these systems.

    • pornthrowaway2@lemmynsfw.comOP
      link
      fedilink
      English
      arrow-up
      1
      ·
      edit-2
      1 year ago

      Woah, never heard of it. I’ll have to check out the competition

      Edit: haha! Check out the man section!

      I ultimately think this idea boils down to the model in use and the training data.