This is seriously insane, in particular as someone mentioned the quality of it. I can't wait to play around with this. SICK!
Sora: Creating video from text
421–430 of 1001 posts
Re: Sora: Creating video from text
#422This is leaps and bounds beyond anything out there, including both public models like SVD 1.1 and Pika Labs' / Runway's models. Incredible.
It's very good. Unclear how far ahead of Lumiere it is ( https://lumiere-video.github.io/ ) or if its more of a difference in prompting/setttings.
The big stand out to me beyond almost any other text video solution is that the video duration is tremendously longer (minute+). Everything else that I've seen can't get beyond 15 to 20 seconds at the absolute maximum.
Re: Sora: Creating video from text
#423Finally new TNG episodes!
Re: Sora: Creating video from text
#424Does anyone know how to handle the depression/doom one feels with these updates? Yes, it's a great technical achievement, but I just worry for the future. We don't have good social safety nets, and we aren't close to UBI. It's difficult for me to see that happen unless something drastic changes. I'm also afraid of one company just having so much power. How does anyone compete?
I agree. I see how misinformation is effecting my family members now. I can't imagine how this is all going to effect what's coming.
Re: Sora: Creating video from text
#425Does anyone know how to handle the depression/doom one feels with these updates? Yes, it's a great technical achievement, but I just worry for the future. We don't have good social safety nets, and we aren't close to UBI. It's difficult for me to see that happen unless something drastic changes. I'm also afraid of one company just having so much power. How does anyone compete?
this is so exciting! why do you feel depressed.
Btw, a year or two from now you'll be able to run a more powerful open model locally. So, not they aren't having some outsized amount of power
Re: Sora: Creating video from text
#426This is all very impressive. I can't help to wonder though. How is text-to-video going to benefit humanity? That's what OpenAI is supposedly about, right? We'll get some groundbreaking film content out of this in the hands of a few talented creatives, and a vast ocean of mediocre content from the hands of talentless people who know how to type. What's the benefit to humanity, concretely?
For models to interact with real-world objects, they first need to understand those objects. These videos demonstrate just how advanced that awareness is. The goal is not to generate videos. Of course, they could and likely will build products on this capability, but the long-term goal is bigger.
Sure, if that's not just marketing. I haven't seen enough evidence to conclude this will go towards that kind of thing yet, but I'm open to the possibility.
Re: Sora: Creating video from text
#427This is all very impressive. I can't help to wonder though. How is text-to-video going to benefit humanity? That's what OpenAI is supposedly about, right? We'll get some groundbreaking film content out of this in the hands of a few talented creatives, and a vast ocean of mediocre content from the hands of talentless people who know how to type. What's the benefit to humanity, concretely?
This vast amount of human talent and computational power could be channeled into fighting disease and death.
Re: Sora: Creating video from text
#428Countdown to when studios licensing this for "unlimited" episodes of your favorite series. There was Seinfeld "Nothing, Forever" AI parody, but once the models improve enough and are cheap enough to deploy, studios will license their content for real and just have endless seasons. Or even custom episodes. Imagine if every episode of a TV show was unique to the viewer.
I am thinking of Stargate SG-1 Season 11. And remaking Game of Thrones after Season 5
Re: Sora: Creating video from text
#429This inside VR goggles would make it amazing. probably it wouldnt even need to render 360, it would generate it on demand. I better go get some feeding tube
That's the difference between Donkey Kong Country and the N64 (or perhaps between Pixar and Quake).
The amount of power needed to generate this can't be feasible for real time VR today. There's a reason even the company that invented (massive and free) Gmail is charging for its top tier generative AI.
Re: Sora: Creating video from text
#430Many might miss the key paragraph at the end: "Sora serves as a foundation for models that can understand and simulate the real world, a capability we believe will be an important milestone for achieving AGI." This also helps explain why the model is so good since it is trained to simulate the real world, as opposed to imitate the pixels. More importantly, its capabilities suggest AGI and general robotics could be cl…
What I want is an AI trained to simulate the human body, allowing scientists to perform artificial human trials on all kind of medicines. Cutting trial times from years to months.