The competition in text-to-video tools is heating up, but a key challenge remains: achieving desired results without exhausting resources. Runway, for instance, often consumes all your credits before producing something usable, even if you stick to their guidelines. Hailuo AI shows better consistency, while Sora Turbo sounds promising with potentially more mature generations. Progress is clear, but there’s still a way to go in perfecting these tools.
Sora is here
961–970 of 1001 posts
Re: Sora is here
#962Earlier quoted context omitted.
On one hand, I like to think that society is getting trained to recognize AI and distrust it. But at the same time my retired boomer parents are over for the holidays and I catch them watching youtube videos completely oblivious to the fact it's an AI voice and just reading an LLM generated script with B roll for eye candy. Often times it's just stolen auto generated captions from larger creators regurgitated by an A…
AI voices have gotten scarily good. They are easy to recognize because most creators use the same voices with the same intonations and don't care to cut out the mistakes. But if you don't recognize the voice it takes a couple sentences to discern that it's AI even with an ear trained on the difference. But it is funny to see how much stuff gets uploaded with zero quality control and still gets traction. These models…
I've heard enough slop using the ElevenLabs voices that I can recognize them almost immediately now. But you're right. Higher end models with less familiar voices are harder to notice. One consistent failing is that they are always too perfect. No mistakes or signs of cuts to edit out where a human VA would have made a mistake. Its all very smooth and perfect. As if they nailed it in the first shot. Once the cheap/free models manage to fix that then we are in real trouble. Also, some really lazy slop creators don't bother to fix issues with pronunciation. But that's not the fault of the model really.
Re: Sora is here
#963Many people say: > these things will get bigger and better much faster than we can learn to discern I would like to ask “Why?” Clearly, these models are just one case of “NN can learn to map anything from one domain to another” and with enough training/overfitting they can approximate reality to a high degree. But, why would it get better to any significant extent? Because we can collect an infinite amount of video?…
> But, why would it get better to any significant extent? Two years ago, the very best closed-source image model was unable to represent anything remotely realistic. Today, there's hundreds of open source models that can generate images that are literally indistinguishable from reality (like Flux). Not only that, there's an entire collection of tools and techniques around style transfer, facial reconstruction, pose c…
I get that. But, let’s say you have a glass, you fill it to one third, then to half, then to three quarter, then to full. Can you expect to fill it beyond full?
Not every process has an infinite ramp.
It seems frontier labs have been throwing all the compute and all the data they could get their hands on at model training for at least the past 2 years. Is that glass a third full or is it nearly full already?
Is the process of filling that particular glass linear or does the top 20% of the glass require X times as much water to fill as the bottom 20%?
Re: Sora is here
#964Earlier quoted context omitted.
And in the UK and Switzerland unfortunately https://help.openai.com/en/articles/10250692-sora-supported-...
is it a regulatory issue?
I suspect so, OpenAI is subject to the EU AI Act [0]. Last time they released the Advanced Voice Mode it also took some time before it became available in the EU. Not sure why UK and Switzerland are delayed as well, they are not in the European Union.
[0] https://openai.com/global-affairs/a-primer-on-the-eu-ai-act/
Re: Sora is here
#965Earlier quoted context omitted.
> Why is having power concentrated in few hands better? Because most people are dangerous morons. I don't think most people should be allowed to operate a car, let alone the most powerful tool for misinformation that has ever existed
I mean, you're clearly not wrong, but how do you propose implementing your worldview without doing even more harm to humanity? The only thing worse than a powerful, dangerous tool in the hands of the masses is a powerful, dangerous tool controlled exclusively by powerful, dangerous people. (Cue the usual moronic analogies involving thermonuclear weapons...)
The problem of distributing access to dangerous things like AI and weapons has been a problem that humans have faced for a few thousand years. Governments are instituted among men, deriving their just powers from the consent of the governed. If there was a formula for good governance, we'd have fewer problems, but generally I believe in democracy, transparency, and liberalism.
Re: Sora is here
#966Is there information when it will be available in other countries, like Germany for example?
What do you mean? “Sora is here” is not enough?
Sorry for the sarcasm but I’m just tired of this fuck Germany attitude by certain companies.
Re: Sora is here
#967Earlier quoted context omitted.
What education do you specifically think is necessary for people with average IQs all over the world to not be fooled by this, given that they are aware that videos can easily be faked in 2024? A high school degree? A bachelors?
>>given that they are aware that videos can easily be faked in 2024? That's a ridiculous assumption. In my experience no one outside of tech circles is even remotely aware that this kind of thing is possible already.
You think that the average member of the public isn't aware that videos can be faked with AI, or non-AI special effects, and your source of data for this is "your own experience"? Really?
My family is mostly working class in an economically depressed part of the Virginia/West Virginia coal country, and every single one of them is aware of this. None of them work in tech, obviously. None have college degrees.
I maintain that the attitude driving this paternalistic, censorious attitude is arrogance and condescension.
A prime example of how broadly aware the public all over the world is of AI faked videos was the reaction in the Arab world to the October 7th videos posted by Hamas. A shocking (and depressing) percentage of Arabs, as well as non-Arab Muslims all over the world, believed the videos and pictures were fakes produced with AI. I don't remember the exact number, but the polling I saw in November showed it was over 50% who believed they were fakes in countries as disparate as Egypt and Indonesia.
Re: Sora is here
#968Earlier quoted context omitted.
>>given that they are aware that videos can easily be faked in 2024? That's a ridiculous assumption. In my experience no one outside of tech circles is even remotely aware that this kind of thing is possible already.
You think that the average member of the public isn't aware that videos can be faked with AI, or non-AI special effects, and your source of data for this is "your own experience"? Really? My family is mostly working class in an economically depressed part of the Virginia/West Virginia coal country, and every single one of them is aware of this. None of them work in tech, obviously. None have college degrees. I mainta…
>>isn't aware that videos can be faked with AI, or non-AI special effects
These two are very different things. My family believes all kinds of videos on the internet are fake. None of them have any idea what a tool like Sora can do. The gap between "oh this was probably special effects" to "you have to notice pixels shimmering around someone's hand to tell" is enormous.
>>My family is mostly working class in an economically depressed part of the Virginia/West Virginia coal country, and every single one of them is aware of this.
Your working class family has time to keep up with the advancements in generative AI for video? They have more free time than I do then. If we're sharing anecdotes about families then my family is from Polish coal country and their idea of AI is talking to your car and it responding poorly.
>>I maintain that the attitude driving this paternalistic, censorious attitude is arrogance and condescension.
I'm confused - who is displaying this "censorious" attitude here?
>> and your source of data for this is "your own experience"? Really?
Yes, really. I mean do you have anything else? You are also quoting things from your own experience.
Re: Sora is here
#969Earlier quoted context omitted.
This entire thread reeks of "I'm smart enough to know that videos can be faked, but Jethro in the trailer park isn't because he's just a plumber, and therefore this tech needs to be censored or else Jethro might believe stuff that makes him vote in a way I don't like" going on here. While the average person overestimates their own intelligence, the average techy dramatically underestimates the intelligence of the ave…
Actually you overestimate the general public's ability to discern what's real or not. On top of that, most people don't even care if it's real. This is exactly why Trump won. Example: if a gen ai vid of a politician doing some crazy crime came out. Even if it were proven fake, people would start questioning everything and still act as if the politician were guilty
"This is exactly why Trump won"
See the part of my comment you are replying to where I specifically stated that the motivation for all of this is that "Jethro doesn't vote the way I want him to". You've proven my point.
The censorious attitudes on HN were non-existent before Trump won in 2016. I know this for a fact. I've had my account on here since 2012, after 2 years of being just a reader.
Meanwhile, you overestimate how immune to misinformation and lies the average HN techy is. Just a few years ago, the majority of people on here believed, with utter conviction, that the bat-borne coronavirus lab in Wuhan had absolutely no connection with the bat-borne coronavirus epidemic that started in Wuhan and that only bigots and ignoramuses could draw such a conclusion. I experienced this whenever I brought up the blatantly obvious, common sense connection in these same comment threads in late 2020 or into mid 2021. The absolutely absurd denial of common sense by otherwise intelligent people was reminiscent of trying to talk to a religious fundamentalist about evolution while pointing at dinosaur fossils and having them continue to deny what was staring them in the face.
Re: Sora is here
#970Earlier quoted context omitted.
The point is if you are the artist and have something in your head. It’s the same problem with image editing. I am sure you have experienced this.
There is no problem unless you insist on reflecting what you had in mind exactly . That needs minute controls, but no matter the medium and tools you use, unless you're doing it in your own quest for artistic perfection, the economic constraints will make you stop short of your idea - there's always a point past which any further refinement will not make a difference to the audience (which doesn't have access to the…
>There is no problem unless you insist on reflecting what you had in mind exactly.
Not disagreeing, just noting: this is not how [most?] people's minds work {I don't think you're holding to that opinion particularly, I'm just reflecting on this point}. We have vague ideas until an implementation is shown, then we examine it and latch on to a detail and decide if it matches our idea or not. For me, if I'm imagining "a superhero planting vegetables in his garden" I've no idea what they're actually wearing, but when an artist or genAI shows me it's a brown coat then I'll say "no something more marvel". Then when ultimately they show me something that matches the idea I had _and_ matches my current conception of the idea I had... then I'll point out the fingernails are too long, when in the idea I hadn't even perceived the person had fingers, never mind too-long fingernails!
I'd warrant any actualised artistic work has some delta with the artists current perception of the work; and a larger delta with their initial perception of it.