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

Seedance 2.5

seed.bytedance.com

Post reply on HN
Print view View original

231–240 of 278 posts

Re: Seedance 2.5

#231

Post by benjiro29 »

tonyhart7 wrote: post #209
jjcm wrote: post #37

The quality is quite high, but an observation is the direction they're taking these models correlates heavily to the usage demand of China vs the West. Specifically, they're immensely focused on t2v for action / high effect shots. There's one human reference shot in the entire release page, and that one doesn't focus on dialog at all. For filmmakers I've talked to in the US, one of the biggest demands they have is v2…

why you also didn't consider population size different ??? China have 1.5 billion people vs US 350 ish

Repost from above:

https://en.wikipedia.org/wiki/Film_industry#Largest_markets_... ("Largest markets by box office revenue")

    United States and Canada $8,870,000,000  2025[175]
    Europe                   $8,200,000,000  2024[176]
    China                    $7,400,000,000  2025[177]
Que revenue China is the same as Europe/US. Population does not matter so much, its revenue generated by the area.

Re: Seedance 2.5

#232

Post by MrNeon »

cautiouscat wrote: post #96
sickcodebruh wrote: post #57

The sample video looks incredible and the ability to keep details consistent for so long is impressive but it _still_ looks screams of AI in every single shot. I can’t put my finger on why, something about the way the rooms are put together, the expressions of the faces, the movements… It’s very unsettling.

I think we as humans are just really good at being able to tell. CGI has been around for decades and yet we can still look at it and say "yeah that was CGI". Uncanny valley I guess?

>and yet we can still look at it and say "yeah that was CGI".

All of it?

Re: Seedance 2.5

#233

Post by galkk »

WhrRTheBaboons wrote: post #197
galkk wrote: post #158

Earlier quoted context omitted.

It will be solved. Walleted, vertically integrated Apple can do a digital authenticity signature and guarantee that the video you see was at least shot on iPhone, and I won’t be surprised if it will happen soon. This + some lidar info will be enabled and everybody except minority will say that this is a great thing to have.

until models will be trained to fake that too

I’m not talking about watermarks, I’m talking about hardware digital signature.

Re: Seedance 2.5

#234

Post by nbardy »

jjcm wrote: post #37

The quality is quite high, but an observation is the direction they're taking these models correlates heavily to the usage demand of China vs the West. Specifically, they're immensely focused on t2v for action / high effect shots. There's one human reference shot in the entire release page, and that one doesn't focus on dialog at all. For filmmakers I've talked to in the US, one of the biggest demands they have is v2…

The faces look amazing. They can post train for identity preservation easy, that is just an adapter

Re: Seedance 2.5

#236

Post by jarjoura »

jjcm wrote: post #37

The quality is quite high, but an observation is the direction they're taking these models correlates heavily to the usage demand of China vs the West. Specifically, they're immensely focused on t2v for action / high effect shots. There's one human reference shot in the entire release page, and that one doesn't focus on dialog at all. For filmmakers I've talked to in the US, one of the biggest demands they have is v2…

Google and xAI only allow 720p when you start adding references, and neither allow you to sync to audio.

This is entirely because of deepfakes.

Apparently Seedance has a version without guardrails if you are a licensed production company. I suspect Google Omni would too, but I don't know anyone who uses that model.

Re: Seedance 2.5

#237

Post by ragequittah »

Jtarii wrote: post #222
hattimaTim wrote: post #219

Earlier quoted context omitted.

Marketers are creating legit ads with these.

I'm sorry but who cares. Ads are just spam but "official".

Seems to be about $800 billion cares. Personalized advertising takes on a whole new meaning if people get to make ads on the fly. I hate everything about it but it's no doubt coming.

Re: Seedance 2.5

#238

Post by galkk »

cautiouscat wrote: post #227
npilk wrote: post #224

Earlier quoted context omitted.

To play devil's advocate: couldn't people generating AI image/video take the time to consider what they want to generate, decide how to prompt it, ask for edits to have subjects pose certain ways, and specify lighting and color grading (or tweak it afterwards with the same tools digital photographers use)? Wouldn't that give them an 'essence' in the same way?

Some of the post generation editing, for sure. For the rest of it in my opinion, they're not actually physically doing any of it. There's no friction. Friction is usually viewed as a negative thing but in the arts the constraints that we have are what make it more interesting. You don't get that accidental abberation in the shot (okay maybe you do if you count an extra finger ;). You don't get to talk back and forth…

I don’t understand that. Even in my short clips that I do for my kiddo I iterate a lot. Yes, right now it’s mostly prompt and reference images but soon (I think/hope) it will be same tools that are used in real life - things like ability to draw on frames and tell what/how to change etc. reshoots exist in real life for a reason.

Look into photo editing where ai based tools can remove subjects, do generative fills and even transform 2d to 3d/video. Same will come to other ai art too. And this will again differentiate amateurs “prompt 3 lines” and pros “I have vision and I will pursue it using tools available to me”. AI is secondary here.

As I said before - ai helps to raise the floor, not ceiling.

Re: Seedance 2.5

#239

Post by tgv »

superloika wrote: post #235

Requiem for photographic evidence.

Why do people keep making these models? Don't they care about the effects this has? It's either that, or they do it on purpose. Just let people push clips about foreigners eating dogs, Jews dry-bleeding babies or the police choking another black man on TikTok, and sit back. "In the midst of chaos, there is also opportunity."

Re: Seedance 2.5

#240

Post by ragequittah »

Nition wrote: post #77
zmmmmm wrote: post #31

Earlier quoted context omitted.

The thing is, it has the same essential property that you don't control it any more. The problem with these tools is the lack of fine grained control means there is no room for you to express your individual creative input. Your total input is a few sentences of a prompt, then the AI did all the creative part. If the creative input is what you enjoyed, it is actually not that much more (or even less) here than it was…

Another aspect is that if you struggle and struggle and create something great, people can see the work you put in, and you'll visibly stand above the rest who don't have the same level of perseverance. But if anyone can create big studio quality visuals with a prompt, how do you stand out? "When everyone's super, no-one will be."

The same way you stand out now. There are giant budget movies out there that are total trash compared to, say, The Odyssey. One that comes to mind that even had Matt Damon is The Great Wall. Such a huge budget and stacked cast (Pedro Pascal, Willem Dafoe) but a garbage movie. The tools being available to you isn't what make a you super.
Post reply on HN