OpenAI Will Shut Down Sora Video Platform
cross-posted from: https://lemmy.world/post/44699253 [https://lemmy.world/post/44699253] > This is clearly a sign that the product failed to draw in enough customers and its viability was overhyped. > > Hopefully, it is the start of the AI bubble bursting.
Go go China !
Bops the tankie.
Like, I have a Chinese LLM loaded right this second and follow them closely, but holy moly. Curb your enthusiasm.
Anyway, OpenAI has plenty of compute to train a Sora 2 if they want, but apparently they don’t. My guess is some combination of:
They couldn’t figure out a more efficient architecture, like you speculated. I buy that. OpenAI’s development is way more conservative than you’d think, and video generation is inherently intense, especially if Sora 1 is the baseline.
…Maybe they looked at metrics, saw Sora is mostly used for spam, scams, or worse, and pulled the plug for liability reasons?
They’re focusing on short-term profitability, as other commenters mentioned.
They’re focusing on short-term profitability hemorrhaging slightly less money, as other commenters mentioned.
Quick nitpicky fix. The only way they don’t lose shit tons of money is if they shut down completely.
OpenAI said it will discontinue Sora, the generative-AI video creation platform it launched in late 2024, without providing a reason for the decision.
That is the strongest indication this is the beginning of the end for the AI bubble. Sora burned a ton of processing power, with no clear value proposition, just to keep the hype cycle going a little longer. Shutting down without explanation leaves the most likely one: they are out of helium to pump into the balloon. And if that balloon isn’t inflating, it’s deflating.
It’s not and probably the opposite.
When Sora launched it was way ahead. Seedance 2’s release was notably better than any of the other video gen models, Sora included.
The market is getting commoditized because there’s no moat and OpenAI hasn’t led on pretty much any release for a while now other than Sora, which they’re probably falling behind on now.
This is the opposite of a burst from a tech standpoint, even if OpenAI as a company starts to pop.
TL;DR: This is likely happening because the tech accelerated across the industry in ways OpenAI can’t catch back up to, not because it’s lagging.
Upvoted for a different perspective, but I suspect it ends in the same place.
OpenAI is kept solvent by investor capital, and capital is kept flowing by the perception of OpenAI being the market leader. Seedance being a better model, enough to cause OpenAI to exit the market, still ruptures the perception of value. In a market with no clear profitability path, that’s ground falling away.
It also can’t be simply commoditized because generations (I’m sure even Seedance) are expensive and still not good enough for production use, even if 50% of their consumer base might boycott if a major studio even did use it in production. Commoditization can’t occur when there’s still no economically self-sustaining, market-acceptable “good enough” product. Without that, even if the leader changes, it’s a race between lemmings (sorry) off the cliff.
Yep - they briefly led in video gen but quickly were overtaken by other groups. There are even open source local models that perform really well now.
They could conceivably catch back up but how does that help them when their priority is chasing the AGI/ASI dragon?
No, it’s called the basic business model for tech companies since years. Sadly.
A bubble popping would be when people start asking for their ROI or sell.
It still is for the creators there. Instead of browsing the algorithm I start on the subscriptions page, to only see uploads from people I actually want to.
There’s sometimes complaints about “I thought you were dead” when the channel has been uploading regularly the entire time. People just never got recommended the videos despite hitting all the buttons.
For example, did you know both Physics Girl and Tom Scott have returned this month - hopefully a sign that the world can still heal.
Some unsolicited add-on recommendations -Ublock origin - beyond the addblocking, I use the picker tool to filter all the extra sections like “news”, “trending” “you might like” etc. Unhook - toggles to disable a bunch of features like comments, home screen, end screen etc. Enhancer for Youtube - Themeing and a bunch of extra settings like setting defaults for each video. speed, volume, resolution, fill screen (which is different than full screen), PIP while you scroll comments. (The author just did a rework, so it can be a little bugged sometimes - reinstalling it fixed it for me last time it went wonky.
Wait, hold on. Enhancer for Youtube is back on Firefox? Correct me if I’m wrong, but IIRC the dev didn’t update the extension for firefox for a long time due to some policies mozilla have with extensions, and IIRC it has been stuck at 2.5 something whereas the chromium version gotten the 3.0.
I’ve had to make do with another youtube extension since it keep bugging out.
this is great news!
There’s sometimes complaints about “I thought you were dead” when the channel has been uploading regularly the entire time.
Every once in a while, if you have the “notifications” on for a creator it’ll just randomly turn it off. I have a creator that I pay $1/mth subscription to, which you would think YouTube would take as a suggestion that I like that channel and show me when they upload something new. Nope! Instead the algorithm thinks I want to watch a fucking Neo Nazi musician.
I’d bet money that the Disney deal falling through was because OpenAI couldn’t guarantee that Sora couldn’t be used to generate porn of their characters, since attackers will almost certainly always find new prompt injections.
Surprise surprise, it’s a giant fucking black box that you can never have complete control over.
The act of breaking the AI and the guardrails around it is certainly a kind of art, I think. That’s rebellion, satire, a spit in the face of corporate fascism. The actual content that’s produced is rather irrelevant.
But for argument’s sake, if the core essence of art is to be a celebration of human creativity, then to pretend that typing a two-sentence prompt into a computer program and then slapping your signature on whatever it spits out is “art” is just absurd.
This topic invites some odd takes on the nature of art, but ‘the text is irrelevant’ is a new one.
How many sentences does someone have to type before we say an idea is theirs and the tool is how they express it? I’m genuinely confused by how many people insist a visual medium, transmitting a concept from one person’s brain to the brains of an audience, can never be art. If someone spends days tweaking a thousand words to get the perfect combination of fifteen unforgivable fetishes, then whatever you’re looking at is their soul laid bare. Even if all they did was look at a thousand slot-machine pulls to pick one that has the right vibe, that’s akin to nature photography. Those artists usually did nothing to influence their subject. They just observed it real nice.
Consider this image. It’s full of outright errors and questionable details, yet it conveys a complex subject in a clever way. People have chided ‘the model doesn’t know what a bouncer is,’ but it doesn’t have to, for a human being to use a bouncer to communicate an opinion.
How many sentences does someone have to type before we say an idea is theirs and the tool is how they express it? […] If someone spends days tweaking a thousand words to get the perfect combination of fifteen unforgivable fetishes, then whatever you’re looking at is their soul laid bare.
Except no one ever publishes the prompt they used to create the “art” so there’s no evidence that they put in that amount of effort. In fact, if they did, I would argue that would be art because there would be evidence of their creativity in the writing. But then they could just… you know… publish the writing. The AI adds nothing to the mix, it really only takes away.
Even if all they did was look at a thousand slot-machine pulls to pick one that has the right vibe, that’s akin to nature photography. Those artists usually did nothing to influence their subject. They just observed it real nice.
I mean, all that really shows is how little respect you have for photographers, that you think that’s at all equivalent.
AI image generation is amazing for replacing stock photos, and not bad at replacing clipart and porn images.
AI video generation is ok at replacing very simple videos without continuity or physics, but their only real applications are for spreading misinformation or mindless scrolling, there’s just no real way to get anyone to pay for them.
That’s aside from the fact that sora could’ve been great for generating generic stock footage/b-roll, but the way they implemented it was to generate a script, then audio, then video, which meant that it really struggled to generate anything without a focal point, ie what it would actually be useful for.
… but their only real applications are for spreading misinformation …
There’s a ton of that going around with sora, but for all I know it could be a smalll group of people. According to their pricing page, 140 bucks a month normally (50% off right now lol) will get you almost 5000 videos a year. Seems plenty to spread a bunch of shit.
In the dotcom era, the push was to create lots of free services. Once you had enough users, you wanted to see how many would be willing to pay for it. There was a formula that justified getting more investment (it varied by domain). Back then, almost nobody other than Amazon survived the hard shaking of the tree.
We may be coming up to the point where customer acquisition through free service ends. Whatsever is left standing will move to the next round.
Everybody else gets dropped on the floor.
FallenWalnut
paraphrand
merc
‹Hexa«Back›
CodeBlooded
Ricky Rigatoni
EatMyPixelDust