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

Veo 3 and Imagen 4, and a new tool for filmmaking called Flow

blog.google

Post reply on HN
Print view View original

411–420 of 570 posts

Re: Veo 3 and Imagen 4, and a new tool for filmmaking called Flow

#411

Post by sim7c00 »

derektank wrote: post #400
sebstefan wrote: post #371

Earlier quoted context omitted.

They're never going to manage to do that, just on a technical level Plus some users might want to legitimately upload things with AI-generated content in it

I'm pretty sure YouTube saves the metadata from all the video files uploaded to it. It seems pretty trivial to exclude videos uploaded without camera model or device setting information. I seriously doubt even a tiny fraction of people uploading AI content to YouTube are taking the time to futz about with the XMP data before they upload it. Sure, they'll miss out on a lot of edited videos doing that, but that's proba…

"Since launching in 2023, SynthID has watermarked over 10 billion images, videos, audio files and texts, helping identify them as AI-generated and reduce the chances of misinformation and misattribution. Outputs generated by Veo 3, Imagen 4 and Lyria 2 will continue to have SynthID watermarks.

Today, we’re launching SynthID Detector, a verification portal to help people identify AI-generated content. Upload a piece of content and the SynthID Detector will identify if either the entire file or just a part of it has SynthID in it.

With all our generative AI models, we aim to unleash human creativity and enable artists and creators to bring their ideas to life faster and more easily than ever before."

From the page linked in the post....

So there's different ways to detect AI generated content (videos/images atleast). (https://www.nature.com/articles/s41586-024-08025-4 <-- paper on synthID / watermarking and detecting it with LLMs)

Re: Veo 3 and Imagen 4, and a new tool for filmmaking called Flow

#413

Post by sim7c00 »

mrklol wrote: post #376
_ncuy wrote: post #311

Google hit the jackpot with their acquisition of YouTube and it's now paying dividend. YouTube is the largest single source of data and traffic on the Internet, and it's still growing fast. I think this data will prove incredibly important to robotics as well. It's a shame they sold Boston Dynamics in one of their dumbest ever moves because of bad PR.

Why should YouTube be here at the advantage? Every competitor also has access to these videos(?)

presumed datasets: 1. its petabytes of data in the public/listed/free tier videos. 2. there's paywalled videos. 3. there's private/unlisted videos.

google will have access to all of these. competitors will have to do tons of network interactions with google to pull in only the first set. (which google could detect and block depending on how these competitors go about it)

Re: Veo 3 and Imagen 4, and a new tool for filmmaking called Flow

#414

Post by ben_w »

DrScientist wrote: post #406
thomashop wrote: post #398

Earlier quoted context omitted.

The AI is not paying for watching videos yet

Indeed, it's the advertisers who are paying for AI to watch videos....

And paying for my sofa to watch a unskippable 50s ad while I make a coffee.

Re: Veo 3 and Imagen 4, and a new tool for filmmaking called Flow

#415

Post by TeMPOraL »

ipnon wrote: post #310
bigmadshoe wrote: post #243

Earlier quoted context omitted.

I'm curious why you decide to declare victory after one successful attempt, but try many times for unsuccessful models. Are you trying to measure whether a model _can_ get it right, or whether it frequently _does_ get it right? I feel like success rate is a better metric here, or at least a fixed number of trials with some success rate threshold to determine model success.

It is indicative of marginal improvements instead of new breakthroughs. iPhone 1 was a paradigm shift. iPhone 10 was essentially iPhone 9 with tweaks. As an AI optimist I would be disappointed to find we are already seeing diminishing returns on R&D.

Just moments ago, I managed to turn a photo of a person into a short clip of them dancing, in half-decent quality, fully locally, on a mid-range gaming GPU (RTX 4070 Ti, 12GB VRAM). I almost run out of RAM (32GB), but it worked, worked well, and took only couple of minutes.

Half a year ago, that was sort of possible for some genius really bent on making it happen. A year ago, that was unthinkable. Today, it's a matter of drag&dropping a workflow to a fresh ComfyUI install and downloading a couple dozen GB of img2vid models.

The returns on R&D are not diminishing, the progress is just not happening everywhere evenly and at the same time.

Re: Veo 3 and Imagen 4, and a new tool for filmmaking called Flow

#416

Post by matsemann »

baxtr wrote: post #353

A whale coming out of the street in Manhattan, a women with a Jellyfish belly walking in the woods. Why is it that all these AI concept videos are completely crazy?

There is a point in that since you don't know how these really should look you can't really judge them on small idiosyncrasies, and hence you get a better impression compared to uncanny valley if it's something common.

However, I also think this is to show that it can create anything, not just copies of stuff it has seen. If you ask for a painting of a woman and it shows you mona lisa, that's not very impressive.

Re: Veo 3 and Imagen 4, and a new tool for filmmaking called Flow

#417

Post by vbezhenar »

informal007 wrote: post #387
_ncuy wrote: post #311

Google hit the jackpot with their acquisition of YouTube and it's now paying dividend. YouTube is the largest single source of data and traffic on the Internet, and it's still growing fast. I think this data will prove incredibly important to robotics as well. It's a shame they sold Boston Dynamics in one of their dumbest ever moves because of bad PR.

Why videos are important for robotics?

If you can generate realistic video stream, responding to player movements and interactions, you can train your robot using that video stream. It's much more scalable, compared to building physical environments and performing real-world training.

Of course the alternative is to use game engines, but it's possible that AI would generate more realistic video stream for the same money spent. Those recent AI-generated videos certainly look much more realistic than any game footage I ever saw.

Re: Veo 3 and Imagen 4, and a new tool for filmmaking called Flow

#418

Post by _xhhf »

DoesntMatter22 wrote: post #393
anilgulecha wrote: post #383

I'd made a prediction/bet a month ago, predicting 6 months to a full 90 minute movie by someone sitting on their computer. [0] The pace is so crazy that was an over estimation! I'll probably get done in 2. Wild times. 0: https://www.linkedin.com/feed/update/urn:li:activity:7317975...

It's doable now. Someone just needs to do it. With voice now it's completely doable. Just throw it all together add some effects and you've got a great movie... In theory

It's not a theory, at Cannes a feature movie has premiered that is generated entirely by AI. Made in Spain.

Re: Veo 3 and Imagen 4, and a new tool for filmmaking called Flow

#420

Post by _xhhf »

seydor wrote: post #399
nrjames wrote: post #70

This is technically impressive and I commend the team that brought it to life. It makes me sad, though. I wish we were pushing AI more to automate non-creative work and not burying the creatives among us in a pile of AI generated content.

What is non-creative work? I think the term reeks of elitism. Every job is creative, even picking up garbage can become an art when one puts effort in it. There is a more sensical distinction between work that is informational in nature, and work that is physical and requires heavy tools in hard-to-reach places. That's hard to do for big tech, because making tests with heavy machinery is hard and time consuming

> even picking up garbage can become an art when one puts effort in it.

good lord. talk about pedantic.

Post reply on HN