Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Sora: Creating video from text

openai.com

Post reply on HN
Print view View original

301–310 of 1001 posts

Re: Sora: Creating video from text

#301

Post by lucisferre »

nopinsight wrote: post #257

Many 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…

> since it is trained to simulate the real world

Is it though? Or is this just marketing?

Re: Sora: Creating video from text

#302

Post by psb217 »

jenny91 wrote: post #253

Absolutely insane. It's very odd where the glitches happen. Did anyone else notice in the "stylish woman ... Tokyo" clip how her legs skip-hop and then cross at 0:30 in a physically impossible way. Everything else about the clip seems so realistic, yet this is where it trips up ?

She's also wearing a different jacket at the end of the video. Continuity is not maintained when the video zooms back out to a wider shot after the close-up on her face. See, e.g., no zipper on end jacket and obvious zipper on jacket earlier in the video, or placement of the silver "buttons" and general structure of the lapels.

The background details are particularly "slippery" in these videos. E.g., in the initial video of walking along a snowy street in Japan, characters on the left just sort of merge into/out of existence. It's impressive locally, but the global structure and ability to paint in finer-grained details in a physically plausible way fails similarly to current image gen models, but more noticeably with the added temporal dimension.

Re: Sora: Creating video from text

#303

Post by hooande »

This really seems like "DALL-E", but for videos. I can make cool/funny videos for my friends, but after a while the novelty wears off.

All of the AI generated media has this quality where I can immediately tell that it's ai, and that becomes my dominant thought. I see these things on social media and think "oh, another ai pic" and keep scrolling. I've yet to be confused about whether something is ai generated or real for more than several seconds.

Consistency and continuity still seem to be a major issues. It would be very difficult to tell a story using Sora because details and the overall style would change from scene to scene. This is also true of the newest image models.

Many people think that Sora is the second coming, and I hope it turns out to have a major impact on all of our lives. But right now it's looking to have about the same impact that DALL-E has had so far.

Re: Sora: Creating video from text

#304

Post by seydor »

chasing wrote: post #24

Yeah, you just can't let all media, all the cost and hard work of millions of photographers, animators, filmmakers, etc be completely consumed and devalued by one company just because it's a very cool technical trick. The more powerful these services become the more obvious that will be. What OpenAI does is amazing, but they obviously cannot be allowed to capture the value of every piece of media ever created — it'll…

They could pay people to capture it. They could buy out one of the stock video companies. this is not important

Re: Sora: Creating video from text

#305

Post by resolutebat »

macrolime wrote: post #241
thepasswordis wrote: post #100

OpenAI demonstrating the size of their moat. How many multi-million-dollar funded startups did this just absolutely obsolete? This is so, so, so much better than every other generative video AI we've seen. Most of those were basically a still image with a very slowly moving background. This is not that. Sam is probably going to get his $7T if he keeps this up, and when he does everybody else will be locked out foreve…

>100% they would pay a lot of money to be able to hang out with Joe Rogan, or some only fans person, and those pornstars or podcasts hosts will never disagree with them, never get mad at them, never get bored of them, never thing they're a loser, etc. All of these things are against the terms of service and attempting them may result in a ban.

There are no terms of service for the open-source clone of this that we'll have in 6 months.

Re: Sora: Creating video from text

#306

Post by kevmo314 »

nopinsight wrote: post #257

Many 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 is latent space if not a representation of the real world?

Re: Sora: Creating video from text

#307

Post by ldjkfkdsjnv »

thepasswordis wrote: post #265
break_the_bank wrote: post #160

In less than a few hours Gemini 1.5 is old news. Sam is doing live demos on Twitter while Google just released a blog. Didn't think Google would be the first of the Facebook, Apple, Google and Microsoft to get disrupted.

The fact that SamA just seems to go off the cuff on twitter pretty frequently is such a breath of fresh air.

Hes a real CEO, Sundar is just a political appointment

Re: Sora: Creating video from text

#309

Post by burkaman »

Imnimo wrote: post #49

https://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…

Maybe it was trained on a bunch of 3d Google Earth videos.
Post reply on HN