News for hackers. Stuff that matters.
Re:Stable Diffusion Textual Inversion by frebord on Monday August 29, 2022 @05:27PM (#32643791)
How many years until we can generate a feature length film from a script?
Re:Stable Diffusion Textual Inversion by bottlepalm on Monday August 29, 2022 @05:47PM (#32643984)
It looks like this is trending towards making our dreams/thoughts reality in a way in that what we imagine can easily be turned into media - music, books, movies, etc.. Pair this up with VR 'the metaverse' and you literally do get the ability to turn thoughts into personalized explorable realities.. what happens after that?

* Do we get lost in it?

* Does today's 'professional' fiction become a lot less lucrative when we can create our own?

* Is there a to leverage this technology the improve the human condition somehow?

Re:Stable Diffusion Textual Inversion by xdfgh1112 on Monday August 29, 2022 @05:53PM (#32644030)
I can create and explore realities using my imagination alone, though. I personally don't think having it become actual 2d or 3d art will have a lasting impact. It might be fun for a while, but it will get old.
Re:Stable Diffusion Textual Inversion by bottlepalm on Monday August 29, 2022 @06:03PM (#32644144)
It's kind of like your imagination on steroids as the system creates worlds using you imagination as the seed and augments it with summation of all the human creations used to train the network. Give Stable Diffusion a sentence for example, it will create something way beyond what you could of imagined and/or created on your own.
Re:Stable Diffusion Textual Inversion by mejutoco on Tuesday August 30, 2022 @03:21AM (#32647562)
Stable diffusion is impressive, but still is a subset of what one _can_ imagine.
Re:Stable Diffusion Textual Inversion by afro88 on Monday August 29, 2022 @06:12PM (#32644221)
I think it will encourage novel ideas in all forms of art. In other words, genuinely new styles and expression will be scarce, because there wasn't thousands of forms of it to train a model on yet.

We will also adjust to AI generated art like have other creative technologies and the novelty will wear off. We will become good at identifying AI generated art and think of it as cheap.

Still, extremely exciting.

Re:Stable Diffusion Textual Inversion by namrog84 on Tuesday August 30, 2022 @12:48AM (#32646730)
> Do we get lost in it?

That is one hypothesis for the fermi paradox, Kardashev scale, and the great filter. At some point all civilizations essentially create infinite dreams/thoughts/Matrix style tech in where we all will retreat inward and have an infinite world to play with and essentially become gods in a virtual reality.

Re:Stable Diffusion Textual Inversion by wnkrshm on Tuesday August 30, 2022 @09:01AM (#32649850)
Nobody would be interested in it really, since everyone (would have) (edit) their own thing they want others to check out. It's like fan fiction collections online, or the 80% of deviantart that you really don't want to spend time with - only now everything looks hollywood polished.
Re:Stable Diffusion Textual Inversion by Loveaway on Monday August 29, 2022 @06:11PM (#32644204)
this probably already all happened before mate
Re:Stable Diffusion Textual Inversion by anigbrowl on Monday August 29, 2022 @05:43PM (#32643941)
5

You could do storyboards from a shooting script* now, but generalizing to synthesizing character and camera movement as well as object physics is a ways off.

* A version of the script used mainly by director and cinematographer with details of each different angle to be used covering the scene.

Re:Stable Diffusion Textual Inversion by bitwize on Monday August 29, 2022 @05:34PM (#32643865)
I want to see the Batman film where the Joker gives Batman a coupon for new parents but it is expired. That should really be a real film in theatres.
Re:Stable Diffusion Textual Inversion by djmips on Monday August 29, 2022 @06:15PM (#32644254)
you 'might' enjoy. Teen titans fixing the timeline.
Re:Stable Diffusion Textual Inversion by goldenkey on Monday August 29, 2022 @05:57PM (#32644082)
I loled.
Re:Stable Diffusion Textual Inversion by Eji1700 on Monday August 29, 2022 @06:17PM (#32644272)
I suspect at least 10+ depending on your definition.

Tools like this will absolutely be used by professionals to cut out portions of the workload, but there's still a large gap between something like this and actually making a coherent, cohesive, consistent, paced, well framed and lit story from text alone.