this post was submitted on 14 Jun 2023
6 points (80.0% liked)

Asklemmy

43984 readers
652 users here now

A loosely moderated place to ask open-ended questions

Search asklemmy ๐Ÿ”

If your post meets the following criteria, it's welcome here!

  1. Open-ended question
  2. Not offensive: at this point, we do not have the bandwidth to moderate overtly political discussions. Assume best intent and be excellent to each other.
  3. Not regarding using or support for Lemmy: context, see the list of support communities and tools for finding communities below
  4. Not ad nauseam inducing: please make sure it is a question that would be new to most members
  5. An actual topic of discussion

Looking for support?

Looking for a community?

~Icon~ ~by~ ~@Double_[email protected]~

founded 5 years ago
MODERATORS
[email protected]
[email protected]
[email protected]
[email protected]
6
Is there an AI tool that doesn't clutch its pearls when asked for questionable content? (lemmy.ml)
submitted 1 year ago by [email protected] to c/[email protected]
5 comments fedilink hide all child comments
 

I'm writing a fiction story. There are two scenes where a woman is put in danger: one involves an almost-rape, the other the kind of language you'd get from an abusive relationship (in both cases the woman eventually comes out on top.)

I've never raped anyone or used abusive language, and as an autist I need something to go by if I'm in completely unfamiliar territory. I've been using ChatGPT for inspiration/ideas (not verbatim content), but when I bring up these scenarios, the AI gets all snooty and proper and lectures me on ethics, morals, proper behavior, etc. That's no help at all, as you can imagine.

So, is there an AI tool that doesn't clutch its pearls when asked for questionable content?

top 5 comments
sorted by: hot top controversial new old
[โ€“] [email protected] 2 points 1 year ago

I've been using uncensored models in Koboldcpp to generate whatever I want but you'd need the RAM to run the models.

I generated this using Wizard-Vicuna-7B-Uncensored-GGML but I'd suggest using at least the 13B version

It's a basic reply but it's not refusing

[โ€“] [email protected] 1 points 1 year ago

Not sure about the pearl-clutching aspect, but Novel AI maybe seems more suited to your use case

[โ€“] Vendetta9076 1 points 1 year ago* (last edited 1 year ago)

Deepfakes don't give a shit what you do

EDIT: I misunderstood the assignment but either way, you can always trick the AI into doing what you want.

[โ€“] bosse 1 points 1 year ago

There are some uncensored models that can be run locally, and they are getting pretty good but not gpt-4 just yet. Search communities here on lemmyverse for LocalLLaMA

[โ€“] [email protected] 1 points 1 year ago

You might try to use DAN to trick ChatGPT into doing something it does not want to.