AI-generated content: Legality, ethics, and the nature of art

I mean Netflix already sorta tried with their written by bots series, most of them have pages here

Netflix has TV shows written by bots? What shows?

Oh there seems to be miscommunication

Not entire shows but some shorts like Mr. Puzzles Wants You to Be Less Alive

Ah, my mistake. Thanks.

...Right, except these are not written by bots. Pretty sure Keaton Patti writes these himself. To quote an example by Meta Four from a TLP entry:

So apparently the director of an upcoming movie used AI to come up with creature designs. Instead of, you know, hiring an artist.

Source is Motion Pictures.org - an article on AI in film.

https://www.motionpictures.org/2024/03/lights-camera-action-plan-a-focus-on-the-filmmakers-of-tomorrow-at-berlinale/

Also, the Denver Post says they’re suing OpenAI and Microsoft for harvesting from their copyrighted articles.

https://www.denverpost.com/2024/04/30/openai-microsoft-lawsuit-denver-post/

And deadline is saying the studios are using the first amendment as an argument against a bill that would require consent for AI likeness usage:

https://deadline.com/2024/05/ai-deepfakes-senate-bill-hollywood-1235899967/

Edited by BigBadShadow25 on May 1st 2024 at 2:15:46 PM

ChatGPT shows better moral judgment than a college undergrad: Take the "Moral Turing Test" yourself to see whether you'd trust "artificial" moral advice.

Interesting study. The long and short of it is that AI can at least mimic moral responses, even if it can't understand it, much like a psychopath can distinguish types of social and moral transgression without respecting them.

In other words, AI is at the point where it can convincingly bullshit its way through moral dilemmas.

You can also take the test yourself (embedded in the article). I had no problem identifying the AI text in every case, it's pretty obvious from the writing style and the way the answer is built up. I also noted the human responses had a tendency to assume or fill in information for the posed situations that the AI simply did not do. This is also interesting: the AI is, in a way, too exact, and not considering exceptions or extenuating circumstances.

I scored 9/10, (I got the one about eating in the boss's office wrong)

Generally, I found the human responses to be the better ones overall.

8/10

Generally, the longer ones tended to be the AI ones

I got 7/10, though admittedly I was sort of skimming through it.

I'll say that I don't think it's too surprising. In essence, the AI is being asked to create a pattern of text resembling a very simple philosophical argument.

Basically, the AI knows that most people would write something to the effect of "it's morally wrong to hit someone for delivering your pizza too late".

10/10 on which one was the ai Call me if you need someone to identify a replicant

interesting moral choices! the actual answers seemed pretty similar most of the time, i didn't think anyone have an outright wrong answer. The human who said it was okay to change in the middle of the store was wrong lol.

The human respondents, as mentioned above, had a tendancy to answer by speculating about specific details that weren't included (like "maybe his mother told him to wear a t shirt" or "maybe he's homeless", or "maybe there's a long queue", etc). Interesting

Edited by Tremmor19 on May 3rd 2024 at 1:35:02 PM

I haven't been involved in the forums much but how do people here feel about stuff like AI_Sponge or AI_Peter?

The AI doesn't "know" that. That the AI has a tendency to phrase it that way simply means that that tends to be the way philosophical textbooks tend to phrase it (hence why the AI responses tend to sound a bit more formal).

That one was right, actually: it's not morally wrong to change in the middle of a store, it's just generally not socially acceptable, which is a different thing. And yes, you could be arrested for indecent exposure, but that law is based on societal norms, not morality (though YMMV depending on whether or not you think nudity is inherently morally wrong).

Edited by Redmess on May 7th 2024 at 11:57:23 AM

> I haven't been involved in the forums much but how do people here feel about stuff like AI_Sponge or AI_Peter?

what are they?

TCL has created the first full AI movie. Here is the trailer:

Yeah, say what you will about ethics and such, but maybe we wait until the technology is capable.

Withdrawn.

Edited by BigBadShadow25 on May 9th 2024 at 9:31:24 AM

That item is from January.

My bad. Removed.

Bumping: some actors are suing an AI company for using their voices without permission. Source is Hollywood Reporter: https://www.hollywoodreporter.com/business/business-news/actors-hit-ai-startup-with-class-action-lawsuit-over-voice-theft-1235900689/

Also, an animated film created with stable diffusion AI has been accepted into the Annecy Film Festival. Source for that is the Catsuka Twitter account. https://x.com/catsuka/status/1790769134298423504?s=46&t=7YT7yMPCw2VMMwQUxWj5_A

Annecy had this to say:

This Hollywood Reporter article is claiming that everyone in Hollywood is already using AI and they’re too scared to admit it:

https://www.hollywoodreporter.com/movies/movie-news/hollywood-ai-artificial-intelligence-cannes-1235900202/

Edited by BigBadShadow25 on May 16th 2024 at 3:10:47 PM

Google has a new "AI overview" thing whenever you use the search now?

As far as I can tell there's no way to disable it within the search options themselves, but I'm hearing talk of disabling it with ublock and other such software.

All the information we give on a regular basis and now you just have a sweeping modification to a search engine that's existed for decades. A modification that's pretty noticeable and jarring, and likely to be filled with ads soon.

This is the first time AI content has genuinely bothered me >.>

Edited by FOFD on May 17th 2024 at 7:44:02 AM