"We’ll be taking several important safety steps ahead of making Sora available in OpenAI’s products. We are working with red teamers — domain experts in areas like misinformation, hateful content, and bias — who will be adversarially testing the model."
- To make sure that the perfectly unbiased algorithms are biased against bias. So in essence, red teamers as in commies I suppose.
Sora: Creating video from text
761–770 of 1001 posts
Re: Sora: Creating video from text
#762All current form of entertainment will be impacted, all of them. Except for live sporting events. This is why I think megacorps all going to bid for sport league streaming right. That's the only one that AI can't touch.
Anyway to benefit economically from this trend?
Re: Sora: Creating video from text
#763https://openai.com/sora?video=big-sur In this video, there's extremely consistent geometry as the camera moves, but the texture of the trees/shrubs on the top of the cliff on the left seems to remain very flat, reminiscent of low-poly geometry in games. I wonder if this is an artifact of the way videos are generated. Is the model separating scene geometry from camera? Maybe some sort of video-NeRF or Gaussian Splatti…
It's possible it was pre-trained on 3D renderings first, because it's easy to get almost infinite synthetic data that way, and after that they continued the training on real videos.
In the car driving on the mountain road video you can see level-of-detail popping artifacts being reproduced, so I think that's a fair guess.
Re: Sora: Creating video from text
#764AI will eventually be capable of performing most of the tasks humans can do. My neighbor's child is only 6 years old now. What advice do you think I should give to his parents to develop their child in a way that avoids him growing up to find that AI can do everything better than he can?
If you want an honest answer you should tell the parents to vote for politicians prepared to launch missile strikes on data centers to secure their child's future. People who are worried purely about employment here are completely missing the larger risks. Realistically his child is going to be unemployable and will therefore either starve or be dependant on some kind of government UBI policy. However UBI is complete…
I'm sure people felt similarly when the first sewing machines were invented. And of course, sewing machines did completely irreversibly change the course of humanity and altered (and even destroyed) many lives. But ultimately, most humans managed, and -- in the end (though that end may be farther away than our own lifetimes) -- benefited.
I'm not sure you're actually under-estimating the impact of this AI meteor that's currently hitting humanity, because it is a huge impact. But I think you're grossly under-estimating the vastness of human endeavors, ingenuity, and resilience. Ultimately we're still talking about the bottom falling out of the creative arts: storytelling, images, movies, even porn -- all of that is about to be incredibly easy to create mediocre versions of. Anyone who thrived on making mediocre art, and anyone who thrived second-hand on that industry, is going to have a very bad time. And that's a lot of people, and it's awful. But we're talking about a complete shift in the creative industries in a world where most people drive trucks and work in restaurants or retail. Yes, many of those industries may also get replaced by AI one day, and rapidly at that, but not by ChatGPT or Sora.
Of course you're right that our near future may suddenly be an AI company hegemony, replacing the current tech hegemony, which replaced the physical retail hegemony, which replaced the manufacturing hegemony, which replaced the railway hegemony, which replaced the slave-owning plantation hegemony, which replaced the guilds hegemony, which replaced the ...
You're also under-estimating how much business can actually be relocated outside the U.S., and also how much revolution can be wrought by a completely disenfranchised generation.
Re: Sora: Creating video from text
#765I think the implications go much further than just the image/video considerations. This model shows a very good (albeit not perfect) understanding of the physics of objects and relationships between them. The announcement mentions this several times. The OpenAI blog post lists "Archeologists discover a generic plastic chair in the desert, excavating and dusting it with great care." as one of the "failed" cases. But t…
It doesn't understand physics. It just computes next frame based on current one and what it learned before, it's a plausible continuation. In the same way, ChatGPT struggles with math without code interpreter, Sora won't have accurate physics without a physics engine and rendering 3d objects. Now it's just a "what is the next frame of this 2D image" model plus some textual context.
> It just computes next frame based on current one and what it learned before, it's a plausible continuation.
...
> Now it's just a "what is the next frame of this 2D image" model plus some textual context.
This is incorrect. Sora is not an autoregressive model like GPT, but a diffusion transformer. From the technical report[1], it is clear that it predicts the entire sequence of spatiotemporal patches at once.
[1]: https://openai.com/research/video-generation-models-as-world...
Re: Sora: Creating video from text
#766Ad generation usecases are getting interesting with Video generation + Controlnet + Finetuning
Re: Sora: Creating video from text
#767This is insane. But I'm impressed most of all by the quality of motion . I've quite simply never seen convincing computer-generated motion before . Just look at the way the wooly mammoths connect with the ground, and their lumbering mass feels real. Motion-capture works fine because that's real motion, but every time people try to animate humans and animals, even in big-budget CGI movies, it's always ultimately obvio…
Serious: Can one just pipe an SRT (subtitle file) and then tell it to compare its version to the mp4 and then be able to command it to zoom, enhance, edit, and basically use it to remould content. I think this sounds great!
Re: Sora: Creating video from text
#768Welp, goodbye internet, it was fun to know you.
Re: Sora: Creating video from text
#769AI will eventually be capable of performing most of the tasks humans can do. My neighbor's child is only 6 years old now. What advice do you think I should give to his parents to develop their child in a way that avoids him growing up to find that AI can do everything better than he can?
He should become a massage therapist or a Circus performer would be solid advice.
Re: Sora: Creating video from text
#770AI will eventually be capable of performing most of the tasks humans can do. My neighbor's child is only 6 years old now. What advice do you think I should give to his parents to develop their child in a way that avoids him growing up to find that AI can do everything better than he can?
If you want an honest answer you should tell the parents to vote for politicians prepared to launch missile strikes on data centers to secure their child's future. People who are worried purely about employment here are completely missing the larger risks. Realistically his child is going to be unemployable and will therefore either starve or be dependant on some kind of government UBI policy. However UBI is complete…
So vote for Putin?