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

$100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

tryai.dev

Post reply on HN
Print view View original

511–520 of 581 posts

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#511

Post by ryandrake »

Scubabear68 wrote: post #26

It is jarring to me that most of the dancing seems slightly out of sync with the music. It is like a music video uncanny valley - images look good, but the lack of sync to the sound shatters the illusion entirely.

While it's made huge improvement in just the past few years, AI still hasn't quite solved motion, especially human motion. The humans in these videos are rendered really well, but move unnaturally. Like someone did a motion capture of a real human, and then played it back with a very high quality 3d model--something hard to describe is lost in the processing. Everything seems to move just too smoothly, at too exactly constant a speed. Same for camera movement. A real life camera doesn't precisely follow some exact 3D Bézier curve at an exactly constant speed. AI doesn't get this yet.

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#512

Post by ericmcer »

scrollop wrote: post #298
hbn wrote: post #257

Like most of this stuff, it's obviously impressive technology compared to what existed a few years ago. But the end product has zero artistic value. It's a grey goo of the average of every concept picked up from the concept of the song. A talented creative with a vision could make something more interesting and enjoyable in an afternoon with a $0 budget.

As much as I don't want jobs taken away by AI, a creative person with a vision that would usually required $1 million budget, could create such a video using AI by being very specific in what they require which would not be a " grey goo of the average of every concept " eg. "scene 32, same 2 characters from the previous scene, now dressed in the garb of Egyption middle kingdom high priests, though with cat faces, dan…

The prompts would more be like: "The dancing and lip movement is still wildly out of time with the music"... "Ok now the dancing is even more out of sync with the beat"

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#513

Post by sporkland »

hbn wrote: post #257

Like most of this stuff, it's obviously impressive technology compared to what existed a few years ago. But the end product has zero artistic value. It's a grey goo of the average of every concept picked up from the concept of the song. A talented creative with a vision could make something more interesting and enjoyable in an afternoon with a $0 budget.

People keep referring to the output of neural nets as the "average of the training data" but imho the deep value in neural nets and the interesting philosophical part is they are massive conditional probability engines whose results can change based on the variables you feed in.

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#514

Post by slopinthebag »

4chandaily wrote: post #421
esikich wrote: post #212

Earlier quoted context omitted.

It's a new medium and I think that is really exciting to watch it develop.

Agreed. As with any new tool, there will be an adjustment period before the Overton Window catches up. This is the current era's version of these hit classics: - "If you use autotune,you aren't a real singer!" - "Programming a drum machine isn't really writing percussion!" - "If you use Photoshop, you can't be a real artist." - "No real artist would use duct tape as a medium." - "Video games can't be art." - "Samplin…

But at the same time, we've had people commissioning art for centuries and we still don't consider the commissioner to be an artist.

Using AI as part of your workflow? Possibly art. Using it as your entire workflow? Slop, not art.

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#515

Post by qingcharles »

BLKNSLVR wrote: post #129

This is how you do AI music videos: https://www.youtube.com/watch?v=Njk2YAgNMnE Lean into the not-quite-but-almost uncanny-valley-ness. Make it a feature, not a bug.

I think this one is a decent example too -- it would probably fool a lot of people:

https://www.youtube.com/watch?v=mZIU8sF3YlY

(the song is AI too)

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#516

Post by throwforfeds »

scrollop wrote: post #298
hbn wrote: post #257

Like most of this stuff, it's obviously impressive technology compared to what existed a few years ago. But the end product has zero artistic value. It's a grey goo of the average of every concept picked up from the concept of the song. A talented creative with a vision could make something more interesting and enjoyable in an afternoon with a $0 budget.

As much as I don't want jobs taken away by AI, a creative person with a vision that would usually required $1 million budget, could create such a video using AI by being very specific in what they require which would not be a " grey goo of the average of every concept " eg. "scene 32, same 2 characters from the previous scene, now dressed in the garb of Egyption middle kingdom high priests, though with cat faces, dan…

Sure, but it appears as if the LLM just doesn't understand dancing and rhythm. I don't know why, but an old music video by the band Jungle came to mind -- just a kid dancing, and while it definitely cost more than $100 I'm sure, it couldn't have been all that expensive: https://www.youtube.com/watch?v=9JkDzNOgO3U

Edit: oh I remembered another, The Blaze said their first music video cost them $100 to make: https://www.youtube.com/watch?v=UivZrL2znh0 and their follow up didn't cost a whole lot more they've also been on record saying: https://www.youtube.com/watch?v=54fea7wuV6s

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#517

Post by y-c-o-m-b »

coldtea wrote: post #108
thewanderer1983 wrote: post #51

Earlier quoted context omitted.

Visual effects went through this same development issues as the industry matured. What took this industry decades to advance it taking months in AI. Think the spaghetti Will Smith and now this. Another one people don't mention here but is specific to video is higgsfield ai.

> Visual effects went through this same development issues as the industry matured The industry haven't matured except until a certain point. Then it declined. Modern visual effects are worse than practical effects in their heyday. They are also worse done than 3D effects in their tasteful early days (like Jurassic Park).

I think this has more to do with the cutthroat business than the technology involved. It seems the tech has improved, but the quality has suffered due to pressure to just fling shit out the door using the least amount of money to generate the most amount of profit. They are squeezing every penny dry. It seems to me the person you're responding to still has a valid point.

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#518

Post by zahlman »

websap wrote: post #482
hbn wrote: post #257

Like most of this stuff, it's obviously impressive technology compared to what existed a few years ago. But the end product has zero artistic value. It's a grey goo of the average of every concept picked up from the concept of the song. A talented creative with a vision could make something more interesting and enjoyable in an afternoon with a $0 budget.

The scary thing is that this is the worst this technology will ever be. The media industry is cooked. Get ready for $100 movie theater tickets, because the cost actual creative work is going to go sky high in the next few years

The industry may be cooked, but the economic effect you predict is entirely illogical. Things don't get more expensive simply because other things get cheaper.

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#519

Post by halJordan »

qbit42 wrote: post #411
lp4v4n wrote: post #402

Earlier quoted context omitted.

I understand what you mean but speakers and recording equipment destroyed the economics of paying people to play music a long time ago. It's another way of looking at the problem. Smartphone cameras and easy-to-use professional equipment were thought to destroy the photography profession, yet photographers still exist, although we have to admit not as many as in the past. If I had a company with a budget, I wouldn't…

Sure, this mirrors history, technology is always transforming economics. I guess my point is that there is a decent sized middle class of creative professionals right now, and I expect that the number of them will go down significantly due to AI. Seems like art will become more and more restricted to hobbyists and the lucky few who achieve mainstream success. This is kind of what happened with photography. But I thin…

If you take the analogy of smartphones & photographers- then why don't you take both sides? We're much better off with now than then.

Re: $100 AI Music Video: Claude Fable 5 vs. GPT-5.6 Sol

#520

Post by timacles »

perching_aix wrote: post #438
spaqin wrote: post #311

Earlier quoted context omitted.

Art isn't just the screenplay, or a prompt; it's the clash between reality and the intent (whether a location or practical set constraints), the individual skill of everyone involved, costume and prop department. And it may change during production. And effort. If it's easy, it's easily reproducible and not worth that much. And novelty, that's something current models aren't capable of. That's why Jarimoquai's Virtua…

A notable chunk of criticism GenAI receives has to do with output irreproducibility and semantic instability relative to the input, so this is a bit entertaining to read. If there's a way to do something better and worse, then one can absolutely talk about added and absent value. I browse AI generated images a ton, and the difference between beginner / low effort submissions and high effort / advanced submissions is…

Yeah one thing people aren’t really grasping with a generative AI is that it fundamentally can’t produce the same thing twice. You can’t really create top notch art with all kinds of inconsistencies
Post reply on HN