Did they claim the video was unedited Sora output? It doesn’t sound like they had to do all that much to the output to get what they wanted. There aren’t any AI tools right now that always output exactly what you want without any alterations, so of course they had to regenerateate clips many times and fix them up manually. They still ended up with a video that required no actual filming, and that’s impressive.
Has generative video’s problem with faces and hands been solved? Not quite. We still get glimpses of warped body parts. And text is still a problem (in another video, by the creative agency Native Foreign, we see a bike repair shop with the sign “Biycle Repaich”). But everything in “Air Head” is raw output from Sora. After editing together many different clips produced with the tool, Shy Kids did a bunch of post-processing to make the film look even better. They used visual effects tools to fix certain shots of the main character’s balloon face, for example.
weird OpenAI neglected to mention that what the real artists were doing with the technology was spend a lot of time heavily editing and fucking rotoscoping its output to look barely passable
but the result was still uninteresting garbage that’s only barely notable if you think generative AI did it, and we’ve established that all the coherent parts of this were done (as usual) with the hard work of a team of uncredited humans
I am confused, was the expectation really a magic automate entire movie clip button? Because thats not how any kind of creative generative ai works in my experience.
llms are not sentient, they cannot perform “intentional reasoning” of course the showcased art is a human work. Of course the raw output has hallucinations, gpt4 is not except of that either but its still a great drafter.
The results stands to appear technologically very impressive. This kind of thing was perceived as never to be possible and improves quickly.
No cameras, no physical shooting, no actors. Just a few creatives and something to compute.
tonight’s promptfans are fucking boring and I’m cranky from openai’s shitty sora page crashing my browser so I guess all you folks doing free marketing for Sam Altman can fuck off now
The results stands to appear technologically very impressive. This kind of thing was perceived as never to be possible and improves quickly.
No cameras, no physical shooting, no actors. Just a few creatives and something to compute.
like @gnomicutterance@hachyderm.io I am begging generative AI idiots to realize how out of touch “no cameras, no physical shooting, no actors” is as a supposed milestone when it applies equally well to Xavier: Renegade Angel… except Xavier looked fucked up on purpose
No it doesn’t, ive read the available documents its a work in progress tech demo being made available to industry experts for the goal of gaining feedback.
Your interpretation on the “as a simulation paper” feels to me (and yes that is subjective) as a lack of reading comprehension if you analyze what is actually all being stated.
“Our results suggest that scaling video generation models is a promising path towards building general purpose simulators of the physical world.“
“Promised path” does not imply the current state is capable of creating or based on a simulation,
it also does not imply it can do this without additional components playing along, neither is it proof that it will be 100% feasible.
If this was a journalist piece id agree that the title was clickbait but this a tech company doing a tech paper for something that there not even turning into a product yet. (and when they do it be a dumb down chatgpt feature because shits expensive to run in full).
I don’t even care about openai its just another silicon valley money cult but i have a fascination for everything computer science and i can read the small print of a tech paper.
Whats being performed is revolutionary (across the field) but people need to chill out thinking they will see ai doing anything significant on its own for a few more years.
Again, reading comprehension. I never said actors aren’t creatives. If your really think i am arguing in such a bad faith does it even matter to you what my argument is really about.
Maybe we got off on the wrong foot, i dont like fighting people.
I see many people hear holding the sentiment that its openai who lied about this specific balloon video… What i dont see is people talking about or with the people who actually made that video.
These people are established artists with a professional career with real life experience working with traditional tools and this one project with Sora.
Is be interested to know if they themselves think about
If the technology made their work easier
How they rate the quality versus there usual non ai work
if they felt openai was transparent about the details of the assignment they where given when publicizing their work.
yes and people 4 months ago were like “haha stupid ai art can’t draw hands” and now that’s just like not a valid argument because the tech has matured to a degree that its pretty reasonable to create something with little to no imperfections, and obviously that will happen again
Did they claim the video was unedited Sora output? It doesn’t sound like they had to do all that much to the output to get what they wanted. There aren’t any AI tools right now that always output exactly what you want without any alterations, so of course they had to regenerateate clips many times and fix them up manually. They still ended up with a video that required no actual filming, and that’s impressive.
deleted by creator
fucking called it
putting the pro in prognostication
This is the article that I read at launch:
I think what they claimed “this is what real artists can do with this technology”
Which appears exactly what this video is.
weird OpenAI neglected to mention that what the real artists were doing with the technology was spend a lot of time heavily editing and fucking rotoscoping its output to look barely passable
but the result was still uninteresting garbage that’s only barely notable if you think generative AI did it, and we’ve established that all the coherent parts of this were done (as usual) with the hard work of a team of uncredited humans
I am confused, was the expectation really a magic automate entire movie clip button? Because thats not how any kind of creative generative ai works in my experience.
llms are not sentient, they cannot perform “intentional reasoning” of course the showcased art is a human work. Of course the raw output has hallucinations, gpt4 is not except of that either but its still a great drafter.
The results stands to appear technologically very impressive. This kind of thing was perceived as never to be possible and improves quickly.
No cameras, no physical shooting, no actors. Just a few creatives and something to compute.
oh come the fuck off it, OpenAI’s marketing presents sora as exactly a magic automate entire movie clip button. here’s OpenAI marketing the stupid thing as a world simulator which is fucking laughable if it can’t maintain even basic consistency. here’s an analysis of how disappointing sora actually is
tonight’s promptfans are fucking boring and I’m cranky from openai’s shitty sora page crashing my browser so I guess all you folks doing free marketing for Sam Altman can fuck off now
also:
like @gnomicutterance@hachyderm.io I am begging generative AI idiots to realize how out of touch “no cameras, no physical shooting, no actors” is as a supposed milestone when it applies equally well to Xavier: Renegade Angel… except Xavier looked fucked up on purpose
Still proud of promptfans
* promptfondlers
Tbh I find both acceptable, and not solely because I thought of the one. Current working mental taxonomy:
Fans: the internet weird-nerds choosing to be bodyshields for this shit absent any other reason whatsoever
Fondlers: those that write the thonkpieces as demonstrated elsethread (the infosec panic one)
No it doesn’t, ive read the available documents its a work in progress tech demo being made available to industry experts for the goal of gaining feedback.
Your interpretation on the “as a simulation paper” feels to me (and yes that is subjective) as a lack of reading comprehension if you analyze what is actually all being stated.
“Our results suggest that scaling video generation models is a promising path towards building general purpose simulators of the physical world.“
“Promised path” does not imply the current state is capable of creating or based on a simulation, it also does not imply it can do this without additional components playing along, neither is it proof that it will be 100% feasible.
If this was a journalist piece id agree that the title was clickbait but this a tech company doing a tech paper for something that there not even turning into a product yet. (and when they do it be a dumb down chatgpt feature because shits expensive to run in full).
I don’t even care about openai its just another silicon valley money cult but i have a fascination for everything computer science and i can read the small print of a tech paper.
Whats being performed is revolutionary (across the field) but people need to chill out thinking they will see ai doing anything significant on its own for a few more years.
it’s quite telling that you don’t think that actors are “creatives” but think that “gpt-4 is a great drafter”.
Again, reading comprehension. I never said actors aren’t creatives. If your really think i am arguing in such a bad faith does it even matter to you what my argument is really about.
Maybe we got off on the wrong foot, i dont like fighting people.
I see many people hear holding the sentiment that its openai who lied about this specific balloon video… What i dont see is people talking about or with the people who actually made that video.
These people are established artists with a professional career with real life experience working with traditional tools and this one project with Sora.
Is be interested to know if they themselves think about
If the technology made their work easier
How they rate the quality versus there usual non ai work
if they felt openai was transparent about the details of the assignment they where given when publicizing their work.
well, that’s true from a certain point of view
marketing department said it so it must be right
yes and people 4 months ago were like “haha stupid ai art can’t draw hands” and now that’s just like not a valid argument because the tech has matured to a degree that its pretty reasonable to create something with little to no imperfections, and obviously that will happen again
Bold to post this in a thread about how it had many many imperfections and what it outputs has to be manually reworked by humans, still.
🎶 Iiiiiiiiii-eye-eye … have become … confidently wrong 🎶
(with only the tiniest apology for the massacre of waters’ lyrics this happens to be)
lying is immoral
they still can’t draw hands