Sora is here • Hacker News Board

Re: Sora is here

#281

Post by cj » Mon, Dec 09, 2024, 7:47 PM UTC

sccomps wrote: post #253

lacoolj wrote: post #235

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…

> 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? Well, none of the existing animation movies follow exact laws of physics.

Take the example to the extreme: In 10 years, I prompt my photo album app "Generate photorealistic video of my mother playing with a ladybug".

The juxtaposition of something that looks extremely real (your mother) and something that never happened (ladybug) is something that's hard for the mind to reconcile.

The presence of a real thing inadvertently and subconsciously gives confidence to the fake thing also being real.

Re: Sora is here

#282

Post by spullara » Mon, Dec 09, 2024, 7:47 PM UTC

sccomps wrote: post #253

lacoolj wrote: post #235

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…

> 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? Well, none of the existing animation movies follow exact laws of physics.

gravity acts immediately, you don't hover in the air for few seconds before falling

Re: Sora is here

#283

Post by TeMPOraL » Mon, Dec 09, 2024, 7:47 PM UTC

randomcatuser wrote: post #195

jsheard wrote: post #139

Earlier quoted context omitted.

> What will save it is that, no matter how picky you are as a creator, your audience will never know what exactly was that you dreamed up, so any half-decent approximation will work. Part of the problem is the "half decent approximations" tend towards a clichéd average, the audience won't know that the cool cyberpunk cityscape you generated isn't exactly what you had in mind, but they will know that it looks like eve…

a somewhat counterintuitive argument is this: AI models will make the overall creative landscape more diverse and interesting, ie, less "average"! Imagine the space of ideas as a circle, with stuff in the middle being more easy to reach (the "cliched average"). Previously, traversing the circle was incredibly hard - we had to use tools like DeviantArt, Instragram, etc to agglomerate the diverse tastes of artists, hop…

To extend the analogy, imagine the circle as a probability distribution; for simplicity, imagine it's a bivariate normal joint distribution (aka. Gaussian in 3D) + some noise, and you're above it and looking down.

When you're commissioning an artist to make you some art, you're basically sampling from the entire distribution. Stuff in the middle is, as you say, easiest to reach, so that's what you'll most likely get. Generative models let more people do art, meaning there's more sampling happening, so the stuff further from the centre will be visited more often, too.

However, AI tools also make another thing easier: moving and narrowing the sampling area. Much like with a very good human artist, you can find some work that's "out there", and ask for variations of it. However, there are only so many good artists to go around. AI making this process much easier and more accessible means more exploration of the circle's edges will happen. Not just "more like this weird thing", but also combinations of 2, 3, 4, N distinct weird things. So in a way, I feel that AI tools will surface creative art disproportionally more than it'll boost the common case.

Well, except for the fly in the ointment that's the advertising industry (aka. the cancer on modern society). Unfortunately, by far most of the creative output of humanity today is done for advertising purposes, and that goal favors the common, as it maximizes the audience (and is least off-putting). Deluge of AI slop is unavoidable, because slop is how the digital world makes money, and generative AI models make it cheaper than generative protein models that did it so far. Don't blame AI research for that, blame advertising.

Re: Sora is here

#284

Post by byteknight » Mon, Dec 09, 2024, 7:48 PM UTC

sccomps wrote: post #253

lacoolj wrote: post #235

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…

> 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? Well, none of the existing animation movies follow exact laws of physics.

Feels like you're looking for a strawman argument, and may have found one.

I would retort that animation and real-life-looking video do different things to our psyche. As an uneducated wanna-be intellectual, I would lean toward thinking real-looking objects more directly influence our perception of life than animations.

Re: Sora is here

#285

Post by lelandfe » Mon, Dec 09, 2024, 7:49 PM UTC

pen2l wrote: post #80

Every day that passes I grow fonder of Google's decision to delay or otherwise keep a lot of this under the wraps. The other day I was scrolling down on YouTube shorts and a couple videos invoked an uncanny valley response from me (I think it was a clip of an unrealistically large snake covering some hut) which was somehow fascinating and strange and captivating, and then scrolling down a few more, again I saw someth…

I saw my first AI video that completely fooled commenters: https://imgur.com/a/cbjVKMU

This was not marked as AI-generated and commenters were in awe at this fuzzy train, missing the "AIGC" signs.

I'm quite nervous for the future.

Re: Sora is here

#286

Post by natvert » Mon, Dec 09, 2024, 7:49 PM UTC

tetris11 wrote: post #274

natvert wrote: post #145

anyone done a comparison with the open-source hunyuanvideoai.com?

0 stars, and no comments the last time this was posted. Maybe too good to be true?

I've ran hunyuanvideoai from their GitHub and it seems to generate realistic videos. It is a bit slow (30-60min per video clip) and requires ~50GB VRAM. I wonder how the quality compares though?

Re: Sora is here

#287

Post by FridgeSeal » Mon, Dec 09, 2024, 7:49 PM UTC

sccomps wrote: post #253

lacoolj wrote: post #235

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…

> 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? Well, none of the existing animation movies follow exact laws of physics.

Well none of the existing animation movies…a to be anything other than animation?

You just know there’ll be people making content within the week for social media that will be trying to pass itself off as real imagery.

Re: Sora is here

#288

Post by imiric » Mon, Dec 09, 2024, 7:49 PM UTC

huehehue wrote: post #225

quenix wrote: 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…

There's this FinTech ad on the NYC subway right now. I can't remember the company, but the entire ad is just a picture of a guitar and some text. Anyway, the guitar is AI generated, and it's really bad. There are 5 strings, which morph into 6 at the headstock. There's a trem bar jammed under the pickguard, somehow. There's a randomly placed blob on the guitar that is supposed to be a knob/button, but clearly is not.…

> You're trying to sell me on something, why would you put so little effort into your advertising? Why would you not just...take a picture of a real guitar?

Is this not evident? Because using AI is much cheaper and faster. Instead of finding the right guitar, paying for a good photographer, location, decoration, and all the associated logistics, a graphics designer can write a prompt that gets you 90% of the vision, for orders of magnitude less cost and time. AI is even cheaper and faster than using stock images and talented graphic designers, which is what we've been doing for the past few decades.

All our media channels, in both physical and digital spaces, will be flooded with this low-effort AI garbage from here on out. This is only the beginning. We'll need to use aggressive filtering and curation in order to find quality media, whether that's done manually by humans or automatically by other AI. Welcome to the future.

Re: Sora is here

#289

Post by skybrian » Mon, Dec 09, 2024, 7:50 PM UTC

lacoolj wrote: post #235

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…

Yes, entertainment spreads lots of myths. But bad physics from AI movies is only a tiny part of the problem. This is similar to worries about the misconceptions people might get from playing too many video games, reading too many novels, watching too much TV, or participating too much in social media.

It helps somewhat that people are fairly aware that entertainment is fake and usually don’t take it too seriously.

Re: Sora is here

#290

Post by lanthissa » Mon, Dec 09, 2024, 7:50 PM UTC

pen2l wrote: post #80

Every day that passes I grow fonder of Google's decision to delay or otherwise keep a lot of this under the wraps. The other day I was scrolling down on YouTube shorts and a couple videos invoked an uncanny valley response from me (I think it was a clip of an unrealistically large snake covering some hut) which was somehow fascinating and strange and captivating, and then scrolling down a few more, again I saw someth…

exactly one lab has passed the test of morals vs profit at this point, and thats deepmind, and they were thoroughly punished for it.

Every value oAI has claimed to have hasn't lasted a milisecond longer than there was profit motive to break it, and even anthropic is doing military tech now.