Live data from Hacker News

Sora is here

openai.com

321–330 of 1001 posts

Re: Sora is here

#321

A little worried how young children watching these videos may develop inaccurate impressions of physics in nature. For instance, that ladybug looks pretty natural, but there's a little glitch in there that an unwitting observer, who's never seen a ladybug move before, may mistake as being normal. And maybe it is! And maybe it isn't? The sailing ship - are those water movements correct? The sinking of the elephant int…

Me too. While I'm generally optimistic about generative art, at this point the models still have this dreamlike quality; things look OK at first glance, but you often get the feeling something is off. Because it is. Texture, geometry, lights, shadows, effects of gravity, etc. are more or less inconsistent.

I do worry that, as we get exposed more and more to such art, we'll become less sensitive to this feeling, which effectively means we'll become less calibrated to actual reality. I worry this will screw with people's "system 1" intuitions long-term (but then I can't say exactly how; I guess we'll find out soon enough).

Re: Sora is here

#322
sorry for the tangent: can't remember a launch they've had where you could just use it. it's always "rollout", "later this quarter", "select users", what's the deal here?

it's given openai this tinge to me that i probably won't ever manage to forget.

Re: Sora is here

#323

A little worried how young children watching these videos may develop inaccurate impressions of physics in nature. For instance, that ladybug looks pretty natural, but there's a little glitch in there that an unwitting observer, who's never seen a ladybug move before, may mistake as being normal. And maybe it is! And maybe it isn't? The sailing ship - are those water movements correct? The sinking of the elephant int…

Fair! I watched a lot of Superman as a kid and I killed myself jumping off a building

Don't be an asshole. When learning to fly, learn by starting on the ground first, not from a tall building. --Bill Hicks

Re: Sora is here

#324
post #199

Earlier quoted context omitted.

The past few years' innovation in AI has roughly been split into two camps for me. LLMs -- Awesome and useful. Disruptive, and somewhat dangerous, but probably more good than harm if we do it right. 'Generative art' (i.e. music generation, image generation, video generation) -- Why? Just why? The 'art' is always good enough to trick most humans at a glance but clearly fake, plastic, and soulless when you look a bit c…

I understand your take but it's only going to get better and incredibly fast. I'm a huge film nerd and I can only dream of a future where I could use these type of tools (but more advanced) to create short films about ideas I've had. It's very exciting to me

I somehow doubt it's (lack of) technology that's stopping you from creating your ideas.

Re: Sora is here

#325
post #299
post #271

Earlier quoted context omitted.

There's big difference between cartoonishly incorrect and uncanny valley plausibly correct.

There's a huge amount of such stuff in movies. Special effects, weapons physics, unrealistic vehicles and planes, or the classic 'hacking'.

Not a bad point, those representations have, in some cases, caused widespread misunderstandings among people who learn about those concepts from movies... and this is all while simultaneously knowing "it's just a movie".

Re: Sora is here

#326

Earlier quoted context omitted.

What are the leading alternatives? (Open source or otherwise)

You have to be specific . What's more important to you? - uncensored output (SD + LoRa) - Overall speed of generation (midjourney) - Image quality (probably midjourney, or an SDXL checkpoint + upscaler) - Prompt adherence (flux, DALL-E 3) EDIT: This is strictly around image generation. The main video competitors are Kling, Hailuo, and Runway.

SD does not generate video, does it?

Re: Sora is here

#328
post #9

I've found using these and similar tools that the amount of prompts and iteration required to create my vision (image or video in my mind) is very large and often is not able to create what I had originally wanted. A way to test this is to take a piece of footage or an image which is the ground truth, and test how much prompting and editing it takes to get the same or similar ground truth starting from scratch. It is…

The adage "a picture is worth a thousand words" has the nice corollary "A thousand words isn't enough to be precise about an image". Now expand that to movies and games and you can get why this whole generative-AI bubble is going to pop.

You are half right. Its funny because I use the same same. Mine is "A picture is worth a thousand words. thats why it takes 1000 words to describe the exact image that you want! Much better to just use Image to Image instead".

Thats my full quote on this topic. And I think it stands. Sure, people won't describe a picture. instead, they will take an existing picture or video, and do modifications of it, using AI. That is much much simpler and more useful, if you can file a scene, and then animate it later with AI.

Re: Sora is here

#329

Earlier quoted context omitted.

What were you working on? It took a month to render 2 seconds of video?

VFX heavy feature for a Disney subsidiary. Each frame is rendered independently of each other - it’s not like video encoding where each frame depends on the previous one, they all have their own scene assembly that can be sent to a server to parallelize rendering. With enough compute, the entire film can be rendered in a few days. (It’s a little more complicated than that but works to a first order approximation) I d…

Are they still using CPUs and not GPUs for rendering?

Weren't the rendering algos ported to CUDA yet?

Post reply on HN