If you're into movies or filmmaking, it's a fantastic AI tool for consistent, fully-intentional scenes with deliberate set and actor blocking.
It's also the cheapest model aggregator service out there. You can log into every AI image and video provider directly and don't have to pay me anything to use the tool. You can use your Sora account, Midjourney account, Grok account, etc. It'll soon let you log into other aggregators like OpenArt, plug in your FAL API key, etc. so you can use your credits/funds wherever they happen to live.
Unlike the other "model aggregator" websites like Higgsfield, this is a desktop app written in Rust that you can keep. It also has highly intentional 2D and 3D design surfaces especially built for design.
Text prompting sucks for artists and designers, so I'm trying to put image and video design onto canvases that you can intuitively mold like clay.
I spend way too much time on short videos on Instagram and TikTok. I've got Sora and I've tried a bunch of times and I'm just baffled by its success so far.
First of all I almost never use these video services with the sound on. I know I'm not the only one, because many services have captions on by default. Sora doesn't seem to have a solution for this yet.
Second, I have almost no desire to see videos with @Sama cameos. I get served a bunch of them every time. Along with MLK, Lincoln, Kennedy etc. @Sama isn't funny to me, and raping the likenesses of some of those figures doesn't really work for me.
Third, there's not enough creativity and range in the videos. I see way too many of the same videos over and over and over. The riffs on the 1980s/90s TV commercial with the kid opening the sucky Christmas present. Ok, maybe there's a small iota of humor once or twice, but not enough to sustain endless remixes of the same thing.
I also hate just about all Jim Carrey films (except maybe the Truman Show) but many other people seem to love them. Perhaps Sora just isn't for me.
The fundamental issue is extremely pro-AI people do not respect any creative endeavor at all, and therefore all of this is unknowable to them. They don't know why people are interested in art or why people want to make art; movies, images, music, any form of creative expression; they simply know that people are interested in those things, and that various AI models will let them make those things without investing into the skills to execute, or indeed, even the vision to know what they might want to make. To them, a movie is not a thing you make because you want to express something: a movie is a product, brought to market, and people then pay tickets to see, and they see things like Sora as a way to make movies that people will want to see. This is the only way that the way they talk makes any sense: because they don't know why people want to see movies, them wanting to see it is treated as a given. Of course they'll want to see it, it's a movie.
And like, this isn't that weird. This is the mindset of every money-minded media executive that exists, and is why studio meddling in projects is almost universally bad, and the more meddling there is, the more frantic focus-group testing that's done, the more diluted and bad the art comes out in the end. Classic example is the original theatrical release of Blade Runner which was subject to tons of studio interference before it's release, and the studio's version is so legendarily bad that the studio itself, decades later, has all but fucking buried it at sea.
It's this particular kind of brain worms that infests people who want to get into that space, but have no desire to cultivate skill. They want to make things people love and/or get the money for doing so, but they have no idea why people make things, or why people want the things people make. Hence why all of them are tripping over themselves for their imagined AI future.
And if you think I’m being elitist or whatever, I’m just reading the marketing here. All of this stuff is sold with the primary selling point being you can make stuff without learning the skills required to make stuff so it’s naturally going to be most appealing to people who lack skills and see the acquisition of them, or more likely the difficulty involved in that, as something they don’t want to deal with.
I was so stupid when GPT3 came out. I knew so little about token prediction, I argued with folks on here that it was capable of so many things that I now understand just aren't compatible with the tech.
Over the past couple of years of educating myself a bit, whilst I am no expert I have been anticipating a dead end. You can throw as much training at these things as you like, but all you'll get is more of the same with diminishing returns. Indeed in some research the quality of responses gets worse as you train it with more data.
I am yet to see anything transformative out of LLMs other than demos which have prompt engineers working night and day to do something impressive with. Those Sora videos took forever to put together, and cost huge amounts of compute. No one is going to make a whole production quality movie with an LLM and disrupt Hollywood.
I agree, an LLM is like an idiot savant, and whilst it's fantastic for everyone to have access to a savant, it doesn't change the world like the internet, or internal combustion engine did.
OpenAI is heading toward some difficult decisions, they either admit their consumer business model is dead and go into competing with Amazon for API business (good luck), become a research lab (give up on being a billion dollar company), or get acquired and move on.
Relatedly, I start to think more and more the AI is great for mediocre stuff. If you just need to do the 1000th website, it can do that. Do you want to build a new framework? Then there will probably be less many useful suggestions. (Still not useless though. I do like it a lot for refactoring while building xrcf.)
EDIT: One reason that lead me to think it's better for mediocre stuff was seeing the Sora model generate videos. Yes it can create semi-novel stuff through combinations of existing stuff, but it can't stick to a coherent "vision" throughout the video. It's not like a movie by a great director like Tarantino where every detail is right and all details point to the same vision. Instead, Sora is just flailing around. I see the same in software. Sometimes the suggestions go towards one style and the next moment into another. I guess AI currently is just way lower in their context length. Tarantino has been refining his style for 30 years now. And always he has been tuning his model towards his vision. AI in comparison seems to always just take everything and turn it into one mediocre blob. It's not useless but currently good to keep in mind I think. That you can only use it to generate mediocre stuff.
Ha, my previous comment was before your new edit mentioning Sora. There is a good reason why the accompanying research report to the Sora demo isn't titled "Awesome Generative Video," but references world models. The interesting feature is how many apparently (approximations to) physical properties emerge (object permanence, linear motion, partially elastic collisions, as well as many of the elements of grammar of film), and which do not (notably material properties of solid and fluids, creation of objects from nothing, etc.)
Hacker Cinema
7 comments on Hacker News
From the linked article, "Sora is not going to generate movies." How's Sora doing these days?
ArtCraft [1], a source available Adobe + AI.
Here's a demo / trailer that shows it off:
https://www.youtube.com/watch?v=NNHSTfWbkaA
If you're into movies or filmmaking, it's a fantastic AI tool for consistent, fully-intentional scenes with deliberate set and actor blocking.
It's also the cheapest model aggregator service out there. You can log into every AI image and video provider directly and don't have to pay me anything to use the tool. You can use your Sora account, Midjourney account, Grok account, etc. It'll soon let you log into other aggregators like OpenArt, plug in your FAL API key, etc. so you can use your credits/funds wherever they happen to live.
Unlike the other "model aggregator" websites like Higgsfield, this is a desktop app written in Rust that you can keep. It also has highly intentional 2D and 3D design surfaces especially built for design.
Text prompting sucks for artists and designers, so I'm trying to put image and video design onto canvases that you can intuitively mold like clay.
Here are some short films made with ArtCraft:
https://www.youtube.com/watch?v=H4NFXGMuwpY
https://www.youtube.com/watch?v=tAAiiKteM-U
https://www.youtube.com/watch?v=SuVW8l-_O3I
Would love to hear feedback if anyone tries it out.
[1] https://getartcraft.com
I spend way too much time on short videos on Instagram and TikTok. I've got Sora and I've tried a bunch of times and I'm just baffled by its success so far.
First of all I almost never use these video services with the sound on. I know I'm not the only one, because many services have captions on by default. Sora doesn't seem to have a solution for this yet.
Second, I have almost no desire to see videos with @Sama cameos. I get served a bunch of them every time. Along with MLK, Lincoln, Kennedy etc. @Sama isn't funny to me, and raping the likenesses of some of those figures doesn't really work for me.
Third, there's not enough creativity and range in the videos. I see way too many of the same videos over and over and over. The riffs on the 1980s/90s TV commercial with the kid opening the sucky Christmas present. Ok, maybe there's a small iota of humor once or twice, but not enough to sustain endless remixes of the same thing.
I also hate just about all Jim Carrey films (except maybe the Truman Show) but many other people seem to love them. Perhaps Sora just isn't for me.
The fundamental issue is extremely pro-AI people do not respect any creative endeavor at all, and therefore all of this is unknowable to them. They don't know why people are interested in art or why people want to make art; movies, images, music, any form of creative expression; they simply know that people are interested in those things, and that various AI models will let them make those things without investing into the skills to execute, or indeed, even the vision to know what they might want to make. To them, a movie is not a thing you make because you want to express something: a movie is a product, brought to market, and people then pay tickets to see, and they see things like Sora as a way to make movies that people will want to see. This is the only way that the way they talk makes any sense: because they don't know why people want to see movies, them wanting to see it is treated as a given. Of course they'll want to see it, it's a movie.
And like, this isn't that weird. This is the mindset of every money-minded media executive that exists, and is why studio meddling in projects is almost universally bad, and the more meddling there is, the more frantic focus-group testing that's done, the more diluted and bad the art comes out in the end. Classic example is the original theatrical release of Blade Runner which was subject to tons of studio interference before it's release, and the studio's version is so legendarily bad that the studio itself, decades later, has all but fucking buried it at sea.
It's this particular kind of brain worms that infests people who want to get into that space, but have no desire to cultivate skill. They want to make things people love and/or get the money for doing so, but they have no idea why people make things, or why people want the things people make. Hence why all of them are tripping over themselves for their imagined AI future.
And if you think I’m being elitist or whatever, I’m just reading the marketing here. All of this stuff is sold with the primary selling point being you can make stuff without learning the skills required to make stuff so it’s naturally going to be most appealing to people who lack skills and see the acquisition of them, or more likely the difficulty involved in that, as something they don’t want to deal with.
I was so stupid when GPT3 came out. I knew so little about token prediction, I argued with folks on here that it was capable of so many things that I now understand just aren't compatible with the tech.
Over the past couple of years of educating myself a bit, whilst I am no expert I have been anticipating a dead end. You can throw as much training at these things as you like, but all you'll get is more of the same with diminishing returns. Indeed in some research the quality of responses gets worse as you train it with more data.
I am yet to see anything transformative out of LLMs other than demos which have prompt engineers working night and day to do something impressive with. Those Sora videos took forever to put together, and cost huge amounts of compute. No one is going to make a whole production quality movie with an LLM and disrupt Hollywood.
I agree, an LLM is like an idiot savant, and whilst it's fantastic for everyone to have access to a savant, it doesn't change the world like the internet, or internal combustion engine did.
OpenAI is heading toward some difficult decisions, they either admit their consumer business model is dead and go into competing with Amazon for API business (good luck), become a research lab (give up on being a billion dollar company), or get acquired and move on.
Relatedly, I start to think more and more the AI is great for mediocre stuff. If you just need to do the 1000th website, it can do that. Do you want to build a new framework? Then there will probably be less many useful suggestions. (Still not useless though. I do like it a lot for refactoring while building xrcf.)
EDIT: One reason that lead me to think it's better for mediocre stuff was seeing the Sora model generate videos. Yes it can create semi-novel stuff through combinations of existing stuff, but it can't stick to a coherent "vision" throughout the video. It's not like a movie by a great director like Tarantino where every detail is right and all details point to the same vision. Instead, Sora is just flailing around. I see the same in software. Sometimes the suggestions go towards one style and the next moment into another. I guess AI currently is just way lower in their context length. Tarantino has been refining his style for 30 years now. And always he has been tuning his model towards his vision. AI in comparison seems to always just take everything and turn it into one mediocre blob. It's not useless but currently good to keep in mind I think. That you can only use it to generate mediocre stuff.
Ha, my previous comment was before your new edit mentioning Sora. There is a good reason why the accompanying research report to the Sora demo isn't titled "Awesome Generative Video," but references world models. The interesting feature is how many apparently (approximations to) physical properties emerge (object permanence, linear motion, partially elastic collisions, as well as many of the elements of grammar of film), and which do not (notably material properties of solid and fluids, creation of objects from nothing, etc.)