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

Veo 2: Our video generation model

deepmind.google

Post reply on HN
Print view View original

281–290 of 342 posts

Re: Veo 2: Our video generation model

#281

Post by ttul »

ks2048 wrote: post #266
0xcb0 wrote: post #206

Imho is stunning, yet what is happening there is super dangerous. These videos will and may be too realistic. Our society is not prepared for this kind of reality "bending" media. These hyperrealistic videos will be the reason for hate and murder. Evil actors will use it to influence elections on a global scale. Create cults around virtual characters. Deny the rules of physics and human reason. And yet, there is no w…

Are Apple and other phone/camera makers working on ways to "sign" a video to say it's an unedited video from a camera? Does this exist now? Is it possible? I'm thinking of simple cryptographic signing of a file, rather than embedding watermarks into the content, but that's another option. I don't think it will solve the fake video onslaught, but it could help.

I think this will be a thing one day, where photos are digitally watermarked by the camera sensor in a non-repudiable manner.

Re: Veo 2: Our video generation model

#282

Post by ferguu_ »

bsenftner wrote: post #273
WickyNilliams wrote: post #254

Earlier quoted context omitted.

The naturalistic fallacy is weak at best, but this is one of the weirdest deployments of it I've encountered. It's not evolution, it's nothing like it. If it's kill or be killed, we should do away with medicine right? Only the strong survive. Why are we saving the weak? Sorry but this argument is beyond silly

Deception is a key part of life, and the inability to discriminate fact from fiction is absolutely a key metric of success. Who said "kill or be killed"? Not I. It is survival or not, flourish or not, succeed or not.

But why must the deception take place? Evolution is natural, The development of AI generated videos takes teams of people, years of effort and millions of pounds. Why should those that are more easily deceived be culled? Do you believe that the future of technology is weeding out the weak? Do you believe the future of humanity is the existence of only those that can use the technologies we develop? You might very well find yourself in a position, a long time from now, where you are easily deceived by newer technologies that you are not familiar with.

Re: Veo 2: Our video generation model

#283

Post by arjie »

yurylifshits wrote: post #163
simonw wrote: post #131

I got access to the preview, here's what it gave me for "A pelican riding a bicycle along a coastal path overlooking a harbor" - this video has all four versions shown: https://static.simonwillison.net/static/2024/pelicans-on-bic... Of the four two were a pelican riding a bicycle. One was a pelican just running along the road, one was a pelican perched on a stationary bicycle, and one had the pelican wearing a weird…

There's another important contender in the space: Hunyuan model from Tencent My company (Nim) is hosting Hunyuan model, so here's a quick test (first attempt) at "pelican riding a bycicle" via Hunyuan on Nim: https://nim.video/explore/OGs4EM3MIpW8 I think it's as good, if not better than Sora / Veo

I was curious how it would perform with prompt enhancement turned off. Here's a single attempt (no regenerations etc.): https://www.youtube.com/watch?v=730cb2qozcM

If you'd like to replicate, the sign-up process was very easy and I was easily able to run a single generation attempt. Maybe later when I want to generate video I'll use prompt enhancement. Without it, the video appears to have lost a notion of direction. Most image-generation models I'm aware of do prompt-enhancement. I've seen it on Grok+Flow/Aurora and ChatGPT+DallE.

    Prompt
    A pelican riding a bicycle along a coastal path overlooking a harbor
    Seed
    15185546
    Resolution
    720×480

Re: Veo 2: Our video generation model

#284

Post by tomp »

0xcb0 wrote: post #206

Imho is stunning, yet what is happening there is super dangerous. These videos will and may be too realistic. Our society is not prepared for this kind of reality "bending" media. These hyperrealistic videos will be the reason for hate and murder. Evil actors will use it to influence elections on a global scale. Create cults around virtual characters. Deny the rules of physics and human reason. And yet, there is no w…

We've had realistic sci-fi and alternate history movies for a very long time.

Re: Veo 2: Our video generation model

#285

Post by DonHopkins »

tomp wrote: post #284
0xcb0 wrote: post #206

Imho is stunning, yet what is happening there is super dangerous. These videos will and may be too realistic. Our society is not prepared for this kind of reality "bending" media. These hyperrealistic videos will be the reason for hate and murder. Evil actors will use it to influence elections on a global scale. Create cults around virtual characters. Deny the rules of physics and human reason. And yet, there is no w…

We've had realistic sci-fi and alternate history movies for a very long time.

[flagged]

Re: Veo 2: Our video generation model

#286

Post by oldmanhorton »

tomp wrote: post #284
0xcb0 wrote: post #206

Imho is stunning, yet what is happening there is super dangerous. These videos will and may be too realistic. Our society is not prepared for this kind of reality "bending" media. These hyperrealistic videos will be the reason for hate and murder. Evil actors will use it to influence elections on a global scale. Create cults around virtual characters. Deny the rules of physics and human reason. And yet, there is no w…

We've had realistic sci-fi and alternate history movies for a very long time.

Which take millions of dollars and huge teams to make. These take one bored person, a sentence, and a few minutes to go from idea to posting on social media. That difference is the entire concern.

Re: Veo 2: Our video generation model

#287

Post by Eisenstein »

code_for_monkey wrote: post #100
terhechte wrote: post #59

Earlier quoted context omitted.

Back when computers took up a whole room, you'd also have asked: "but what exactly is this useful for? B-Roll some simple calculations that anybody can do with a piece of paper and a pen."? Think 5-10 years into the future, this is a stepping stone

this is kind of an unfair comparison. Whats the endpoint of generating AI videos? What can this do that is useful, contributes something to society, has artistic value, etc etc. We can make educational videos with a script but its also pretty easy for motivated parties to do that already, and its getting easier as cameras get better and smaller. I think asking "whats the point of this" is at least fair.

The end point is enabling people to put into video what is in their mind. Like a word processor for video. When you remove the need to have a room full of VFX artists to make a movie, then anyone can make a movie. Whether this is beneficial is dubious, but that's an end goal if you are looking for one.

Re: Veo 2: Our video generation model

#288

Post by exodust »

fernly wrote: post #152

Superficially impressive but what is the actual use case of the present state of the art? It makes 10-second demos, fine. But can a producer get a second shot of the same scene and the same characters, with visual continuity? Or a third, etc? In other words, can it be used to create a coherent movie --even a 60-second commercial -- with multiple shots having continuity of faces, backgrounds, and lighting? This quote…

> "what is the actual use case of the art?"

Not much. Low quality over-saturated advertising? Short films made by untalented lazy filmmakers?

When text prompts are the only source, creativity is absent. No craft, no art. Audiences won't gravitate towards fake crap that oozes out of AI vending machines, unrefined, artistically uncontrolled.

Imagine visiting a restaurant because you heard the chef is good. You enjoy your meal but later discover the chef has a "food generator" where he prompts the food into existence. Would you go back to that restaurant?

There's one exception. Video-to-video and image-to-video, where your own original artwork, photos, drawings and videos are the source of the generated output. Even then, it's like outsourcing production to an unpredictable third party. Good luck getting lighting and details exactly right.

I see the role of this AI gen stuff as background filler, such as populating set details or distant environments via green screen.

Re: Veo 2: Our video generation model

#289

Post by taneq »

arjie wrote: post #283
yurylifshits wrote: post #163

Earlier quoted context omitted.

There's another important contender in the space: Hunyuan model from Tencent My company (Nim) is hosting Hunyuan model, so here's a quick test (first attempt) at "pelican riding a bycicle" via Hunyuan on Nim: https://nim.video/explore/OGs4EM3MIpW8 I think it's as good, if not better than Sora / Veo

I was curious how it would perform with prompt enhancement turned off. Here's a single attempt (no regenerations etc.): https://www.youtube.com/watch?v=730cb2qozcM If you'd like to replicate, the sign-up process was very easy and I was easily able to run a single generation attempt. Maybe later when I want to generate video I'll use prompt enhancement. Without it, the video appears to have lost a notion of direction.…

I mean, you didn’t SAY riding forwards…

Re: Veo 2: Our video generation model

#290

Post by eddd-ddde »

ttul wrote: post #281
ks2048 wrote: post #266

Earlier quoted context omitted.

Are Apple and other phone/camera makers working on ways to "sign" a video to say it's an unedited video from a camera? Does this exist now? Is it possible? I'm thinking of simple cryptographic signing of a file, rather than embedding watermarks into the content, but that's another option. I don't think it will solve the fake video onslaught, but it could help.

I think this will be a thing one day, where photos are digitally watermarked by the camera sensor in a non-repudiable manner.

This is a losing battle. You can always just record an AI video with your camera. Done, now you have a real video.
Post reply on HN