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

Make-A-Video: AI system that generates videos from text

makeavideo.studio

Post reply on HN
Print view View original

111–120 of 399 posts

Re: Make-A-Video: AI system that generates videos from text

#111

Post by bottlepalm »

EmilyHughes wrote: post #49
RupertEisenhart wrote: post #35

Earlier quoted context omitted.

Check out the original simulation argument paper[0]. The issue is, if we think we are heading to a world in which we can do simulations, it becomes increasingly likely that we are in one of those (presumably very many) worlds, rather than in the one world that existed before the advent of such simulations. [0] https://www.simulation-argument.com/simulation.pdf

Sounds good because it seems like it explains what our reality is, but it really doesn't. It just pushes the fundamental problem up X simulations into the supposed "real reality". It also assumes that simply simulating a universe would generate consciouss beings which nobody knows the answer to, but my guess is that it would not.

It does push the problem up, but it does let us factor out a lot of complex particle physics which are not needed to run the simulation.

Remember in this case the simulation is not running a model of physical reality - it’s only running a neural net that is fed into your senses.

In that case there are no ‘atoms’ it’s just a concept fed into your mind. Just like in dream it’s hard to question reality, you just go along with it.

Re: Make-A-Video: AI system that generates videos from text

#112

Post by echelon »

indigodaddy wrote: post #109

So how long before we can “Hey music-stable-diffusion-service,” play me “relaxing baroque violin and flute music in the style of game of thrones” ?

My start-up is working on this! We'll be launching a web and downloadable version soon.

Eg. https://www.youtube.com/watch?v=r_0JjYUe5jo --> https://vocaroo.com/1hgjjnVNqWjk

We're also working on film generation.

Re: Make-A-Video: AI system that generates videos from text

#113

Post by anonymouse008 »

indigodaddy wrote: post #109

So how long before we can “Hey music-stable-diffusion-service,” play me “relaxing baroque violin and flute music in the style of game of thrones” ?

I think we all can view a video of 'nails on a chalkboard' before we hear audio of nails on a chalkboard.

For some reason, unacceptable uncanny sounds is a much wider valley than unacceptable videos/pictures. The hand holding is uncanny in the family video, but I'm fine watching it for a second - it doesn't cause pain the way that same error would in music.

Re: Make-A-Video: AI system that generates videos from text

#114

Post by harrisonjackson »

More examples here: https://make-a-video.github.io/

Some of these look like existing videos that have just be garbled - check the clownfish one and the litter of puppies - maybe that is because the prompts aren't that detailed or they just need to up the "creative/randomness" factor.

The sloth one though has a more specific prompt and came out looking more better? and more original.

With the huge amount of data being used to back these, how do we know the "uniqueness" of the generated content? is it original or is it just mangling of existing content?

Even if it is just garbling existing content, it is pretty amazing. Frame by frame the mutant unicorns float across the beach hah.

Re: Make-A-Video: AI system that generates videos from text

#115

Post by Bakary »

mensetmanusman wrote: post #7
Datenstrom wrote: post #2

The > A golden retriever eating ice cream on a beautiful tropical beach at sunset, high resolution example is terrifying.

They forgot ‘trending on onlyfans’ But actually, this technology is super exciting. Imagine a future where movies and games are choose your own adventure.

We'll have procedural generation that will be hard to distinguish from human-made content. Goodbye repetitive Skyrim filler caves!

Re: Make-A-Video: AI system that generates videos from text

#116

Post by rany_ »

mach1ne wrote: post #96
tpmx wrote: post #8

Earlier quoted context omitted.

In case anyone else had problems finding it - it's at the bottom of the page: https://makeavideo.studio/assets/a_golden_retriever_eating_i... (webp) That grasp though. These things still feel a bit like e.g. Google/GCP services to me: Super appealing at first glance, quite close to what you want, but somehow never quite there. Maybe they'll asymptotically get there, eventually? Perhaps that statistical model can't re…

It may be that it's the deep learning tech which will never quite get there. GPT-3 has similar shortcomings in its mimicry. We're 95% there, I guess, but may never quite reach 100%.

> the deep learning tech which will never quite get there

Never say never, we've come a long way since GPT-2! All this was unthinkable back then

Re: Make-A-Video: AI system that generates videos from text

#117

Post by castlecrasher2 »

sphars wrote: post #107
castlecrasher2 wrote: post #98

Earlier quoted context omitted.

Have you by chance tried out NeROIC? I'm a 3D printing enthusiast, mostly video game stuff, and it seems like like it would be excellent for that purpose.

I actually have been trying it out this week, and in fact it's currently trying to process the video generation, like their example shows. While I was able to follow their steps for training using their dataset, and generate the lighting/depth maps for the milkcarton example, the video generation is taking a long time (over 24 hours so far, using a 3070Ti with 8GB VRAM). From what I understand with NeROIC, it's not p…

Thanks! I imagine one could use something like ZBrush's dynamesh to create a usable mesh from the output. Shame the library doesn't provide it by default, though.

Re: Make-A-Video: AI system that generates videos from text

#119

Post by Verdex »

generalizations wrote: post #105
asutekku wrote: post #88

Earlier quoted context omitted.

With the future version of this you could theoretically prototype movies way faster and try ideas with test audiences without requiring to actually film them

In the future version of this, the end user asks for a movie and gets a custom movie generated just for them. You want a thought-provoking Bourne-style action thriller with hints of Jane Austen and a bollywood dance sequence? How about a Matrix sequel that lives up to the first one, but ends just the way you like it? Just ask.

Xenomorph cop. It's a Dirty Harry style movie except the lead is a xenomorph. The other characters are only vaguely aware that he's a xenomorph.

Or how about the movie Clue with the three endings except an infinite sequence of "or maybe it happened this way" sequences. I mean how else are we going to get a sequence where Darth Vader and Tim Curry reenact the "No I am your father" scene while Martin Mull dies in the background due to a heart attack.

Re: Make-A-Video: AI system that generates videos from text

#120

Post by mach1ne »

rany_ wrote: post #116
mach1ne wrote: post #96

Earlier quoted context omitted.

It may be that it's the deep learning tech which will never quite get there. GPT-3 has similar shortcomings in its mimicry. We're 95% there, I guess, but may never quite reach 100%.

> the deep learning tech which will never quite get there Never say never, we've come a long way since GPT-2! All this was unthinkable back then

It's certainly possible. I find it somewhat unlikely though, despite the fast-paced progress.
Post reply on HN