If we go from DALL-E 3, it won't be nowhere near competitive while they have the superior ground. Generating a high quality 1024x1024 image with costs around ~$0.002, but $0.08 on DALL-E 3 (20x more expensive per-image). For videos with very high computational needs (since each frame needs to be temporally consistent, you need huge GPUs to serve this) I'm expecting this to be so much more expensive than its competitors (Pika or SVD1.1)
Yes, but the point is that in a few years, there won't be a difference. Those clickbait accounts already exist for AI generated images. How many impressionable or young people have been fooled into believing history that never happened?
More importantly, how can these accounts subtly direct the generations to instill modern ideology or politics into "historical" images, giving them historical credibility? Think of all the subtly white supremacist "retvrn" accounts, for example, falsely recontextualizing inventions and accomplishments to support their ideology.
We all need to be thinking much more creatively and cynically about how these tools will be abused. The technology will get better. The people who want to abuse it will get smarter. And your capability to distinguish fake information is likely much worse than you believe - to say nothing of younger people who have less context and experience to form a mental "immune system".
>How many impressionable or young people have been fooled into believing history that never happened?
I would say, all of them. Since the dawn of history. Actually, far before, as treachery certainly precedes speech itself by a few million years in the struggle to survive game.
Just to take a contemporary western (mostly?) thing: how did it went last time you looked straight into the eyes of kids to reveal them Santa Clauss is a lie and yes almost all adults in their society are into that evil conspiracy? And what about the adult around you deeply attached to their national myths, not even mentioning all the folklore around their afterlife beliefs?
But don’t worry, everything is going to go well, I promise and you know you can trust me. :)
only if you consider "historical footage" to exclusively mean the "[original] historical footage [stored in archiving]" versus e.g. "historical[ly accurate] footage"
if "historical" is going to be used subjectively with no further qualifying statements then the meaning of "history" will be subjucated to the context it's being presented in, I don't see it's use here as contradictory
These samples look pretty amazing. I'm curious the compute required to train and even deploy something like this. How would it scale to making something like a CGI Pixar movie?
Agreed. It's amazing how much of a head start OpenAI appears to have over everyone else. Even Microsoft who has access to everything OpenAI is doing. Only Microsoft could be given the keys to the kingdom and still not figure out how to open any doors with them.
Except they keep trying to shove AI into everything they own. CoPilot Studio is an example of how laughably bad at it they are. I honestly don't understand why they don't contract out to OpenAI to help them do some of these integrations.
Every company is trying to shove AI into everything they own. It's what investors currently demand.
OpenAI is likely limited by how fast they are able to scale their hiring. They had 778 FTEs when all the board drama occurred, up 100% YoY. Microsoft has 221,000. It seems difficult to delegate enough headcount to all the exploratory projects of MSFT and it's hard to scale headcount quicker while preserving some semblance of culture.
The only official statement from Micorosft is "While details of our agreement remain confidential, it is important to note that Microsoft does not own any portion of OpenAI and is simply entitled to share of profit distributions," said company spokesman Frank Shaw.
Microsoft doesn’t have access to OpenAI’s research, this was part of the deal. They only have access to the weights and inference code of production models and even then who has access to that inside MS is extremely gated and only a few employees have access to this based on absolute need to actually run the service.
AI researcher at MSFT barely have more insights about OpenAI than you do reading HN.
Microsoft said that they could continue OpenAI's research with no slowdown if OpenAI cut them off by hiring all OpenAI's people, so from that statement it sounds like they have access.
No. They have early access. Example: MSFT was using Dall-e Exp (early 3 version) in PUBLIC, since February of 2023.
In the same month, they were also using GPT4 in public - before OpenAI.
And they had access to GPT4 in 2022 (which was when they decided to create Bing Chat, now called Copilot).
All the current GPT4 models at MSFT are also finetuned versions (literally Creative and Precise mode runs different finetuned versions of GPT4). It runs finetuned versions since launch even...
I'm almost speechless. I've been keeping an eye on the text-to-video models, and if these example videos are truly indicative of the model, this is an order of magnitude better than anything currently available.
In particular, looking at the video titled "Borneo wildlife on the Kinabatangan River" (number 7 in the third group), the accurate parallax of the tree stood out to me. I'm so curious to learn how this is working.
holy cow, is that the future of gaming? instead of 3D renders it's real-time video generation, complete with audio and music and dialog and intelligent AI conversations and it's a unique experience no one else has ever played. gameplay mechanics could even change on the fly
Even otherwise, and no matter how good the screen and speakers are, a screen and speakers can only be so immersive. People oversell the potential for VR when they describe it as being as good as or better than reality. Nothing less than the Matrix is going to work in that regard.
Yep, once your brain gets over the immediate novelty of VR, it’s very difficult to get back that “Ready Player One” feeling due to the absence of sensory feedback.
If/once they get it working though, society will shift fast.
There’s an XR app called Brink Traveler that’s full of handcrafted photogrammetry recreations of scenic landmarks. On especially gloomy PNW winter days, I’ll lug a heat lamp to my kitchen and let it warm up the tiled stone a bit, put a floor fan on random oscillation, toss on some good headphones, load up a sunny desert location in VR, and just lounge on the warm stone floor for an hour.
My conscious brain “knows” this isn’t real and just visuals alone can’t fool it anymore, but after about 15 minutes of visuals + sensory input matching, it stops caring entirely. I’ve caught myself reflexively squinting at the virtual sun even though my headset doesn’t have HDR.
Yeah, but I mean who knows why. I know some people can't, my GF is one of them.
I've often wondered if im ok with it because im used to the object on head stuff (like 25 odd years of motorcycle riding/ergo helmet wearing) and close up, high fov coverage fast past gaming? (I play on a 32" maybe 70 cms from the eyes give or take.)
> I am prone to sea sickness. Maybe it is related.
I'd think it might be given my understanding of why illness in many is triggered. It's odd because I never got sick from it, but i've seen others get INCREDIBLY ill in two different ways.
1. My GF tried to use simple locomotion in a game and almost vomited as an immediate reaction
2. A friend who was fine at first, but then randomly started getting very slowly ill over a matter of like an hour, just getting more and more nausea after the fact.
It's unfortunate, because due to lack of bad feelings/nausea/discomfort etc, I love VR. I equally from those around me can see no real path forward for it as it stands today though because of those impacts and limitations.
That being said, maybe they get smaller, lighter, we learn to induce motion sickness less, I dunno. I'm not optimistic.
DLSS is essentially this, isn't it? It uses a low quality render from the game and then increases the fidelity with something very similar to a diffusion model.
For games like 2D/3D fighting games where you don't to generate a lot of terrain, the possibilities of randomly generating stages with unique terrain and obstacles is interesting.
The diffusion is almost certainly taking place over some sort of compressed latent, from the visual quirks of the output I suspect that the process of turning that latent into images goes latent -> nerf / splat -> image, not latent -> convolutional decoder -> image
Where is the training material for this coming from? The only resource I can think of that's broad enough for a general purpose video model is YouTube, but I can't imagine Google would allow a third party to scrape all of YT without putting up a fight.
You can still have a broad dataset and use RLHF to steer it more towards the aesthetic like midjourney and SDXL did through discord feedback. I think there was still some aesthetic selection in the dataset as well but it still included a lot of crap.
In terms of following the prompt and generating visually interesting results, I think they're comparable. But the resolution for Sora seems so far ahead.
Worth noting that Google also has Phenaki [0] and VideoPoet [1] and Imagen Video [2]
The big stand out to me beyond almost any other text video solution is that the video duration is tremendously longer (minute+). Everything else that I've seen can't get beyond 15 to 20 seconds at the absolute maximum.
Let's hold our breath. Those are specifically crafted hand-picked good videos, where there wasn't any requirement but "write a generic prompt and pick something that looks good", with no particular requirements. Which is very different from the actual process where you have a very specific idea and want the machine to make it happen.
DALL-E presentation also looked cool and everyone was stoked about it. Now that we know of its limitations and oddities? YMMV, but I'd say not so much - Stable Diffusion is still the go-to solution. I strongly suspect the same thing with Sora.
The examples are most certainly cherry-picked. But the problem is there are 50 of them. And even if you gave me 24 hour full access to SVD1.1/Pika/Runway (anything out there that I can use), I won't be able to get 5 examples that match these in quality (~temporal consistency/motions/prompt following) and more importantly in the length. Maybe I am overly optimistic, but this seems too good.
Credit to OpenAI for including some videos with failures (extra limbs, etc.). I also wonder how closely any of these videos might match one from the training set. Maybe they chose prompts that lined up pretty closely with a few videos that were already in there.
Lack of quality in the details yes but the fact that characters and scenes depict consistent and real movement and evolution as opposed to the cinemagraph and frame morphing stuff we have had so far is still remarkable!
That particular example seems to have more a "cheap 3d" style to it but the actual synthesis seems on par with the examples. If the prompt had specified a different style it'd have that style instead. This kind of generation isn't like actual animating, "cheap 3d" style and "realistic cinematic" style take roughly the same amount of work to look right.
In the past the examples tweeted by OpenAI have been fairly representative of the actual capabilities of the model. i.e. maybe they do two or three generations and pick the best, but they aren't spending a huge amount of effort cherry-picking.
> Stable Diffusion is still the go-to solution. I strongly suspect the same thing with Sora.
Sure, for people who want detailed control with AI-generated video, workflows built around SD + AnimateDiff, Stable Video Diffusion, MotionDiff, etc., are still going to beat Sora for the immediate future, and OpenAI's approach structurally isn't as friendly to developing a broad ecosystem adding power on top of the base models.
OTOH, the basic simple prompt-to-video capacity of Sora now is good enough for some uses, and where detailed control is not essential that space is going to keep expanding -- one question is how much their plans for safety checking (which they state will apply both to the prompt and every frame of output) will cripple this versus alternatives, and how much the regulatory environment will or won't make it possible to compete with that.
> I suspect given equal effort into prompting both, Sora probably provides superior results
Strictly to prompting, probably, just as that is the case with Dall-E 3 vs, say, SDXL.
The thing is, there’s a lot more that you can do than just tweaking prompting with open models, compared to hosted models that offer limited interaction options.
While Sora might be able to generate short 60-90 second videos, how well it would scale with a larger prompt or a longer video remains yet to be seen.
And the general logic of having the model do 90% of the work for you and then you edit what is required might be harder with videos.
Most fictional long-form video (whether live-action movies or cartoons, etc) is composed of many shots, most of them much shorter than 7 seconds, let alone 60.
I think the main factor that will be key to generate a whole movie is being able to pass some reference images of the characters/places/objects so they remain congruent between two generations.
You could already write a whole book in GPT-3 from running a series of one-short-chapter-at-a-time generations and passing the summary/outline of what's happened so far. (I know I did, in a time that feels like ages ago but was just early last year)
> I think the main factor that will be key to generate a whole movie is being able to pass some reference images of the characters/places/objects so they remain congruent between two generations.
I partly agree with this. The congruency however needs to extend to more than 2 generations. If a single scene is composed of multiple shots, then those multiple shots need to be part of the same world the scene is being shot in.
If you check the video with the title `A beautiful homemade video showing the people of Lagos, Nigeria in the year 2056. Shot with a mobile phone camera.` the surroundings do not seem to make sense as the view starts with a market, spirals around a point and then ends with a bridge which does not fit into the market.
If the the different shots generated the model did fit together seamlessly, trying to make the fit together is where the difficulty comes in. However I do not have any experience in video editing, so it's just speculation.
The CGI industry is about to be turned upside down. They charge hundreds of thousands per minute, and it takes them forever to produce the finished product.
Sarah is a video sorter, this was her life. She graduated top of her class in film, and all she could find was the monotonous job of selecting videos that looked just real enough.
Until one day, she couldn't believe it. It was her. A video of of her in that very moment sorting. She went to pause the video, but stopped when he doppelganger did the same.
I suspect it's less about being puritanical about violence and nudity in and of themself, and more a blanket ban to make up for the inability to prevent the generation of actually controversial material (nude images of pop stars, violence against politicians, hate speech)
Put like that, it's a bit like the Chumra in Judaism [1]. The fence, or moat, around the law that extends even further than the law itself, to prevent you from accidentally commiting a sin.
I am guessing a movie studio will get different access with controls dropped. Of course, that does mean they need to be VERY careful when editing, and making sure not to release a vagina that appears for 1 or 2 frames when a woman is picking up a cat in some random scene.
We can't do narrative sequences with persistent characters and settings, even with static images.
These video clips just generic stock clips. You cut cut them together to make a sequence of random flashy whatever, but you still can't do storytelling in any conventional sense. We don't appear to be close to being able to use these tools for the hypothetical disruptive use case we worry about.
Nonetheless, The stock video and photo people are in trouble. So long as the details don't matter this stuff is presumably useful.
I wonder how much of it is really "concern for the children" type stuff vs not wanting to deal with fights on what should be allowed and how and to who right now. When film was new towns and states started to make censorship review boards. When mature content became viewable on the web battles (still ongoing) about how much you need to do to prevent minors from accessing it came up. Now useful AI generated content is the new thing and you can avoid this kind of distraction by going this route instead.
I'm not supporting it in any way, I think you should be able to generate and distribute any legal content with the tools, but just giving a possible motive for OpenAI being so conservative whenever it comes to ethics and what they are making.
I've been watching 80s movies recently, and amount of nudity and sex scenes often feels unnecessary. I'm definitely not a prude. I watch porn, I talk about sex with friends, I go to kinky parties sometimes. But it really feels that a lot of movies sacrificed stories to increase sex appeal — and now that people have free and unlimited access to porn, movies can finally be movies.
I know it's Runway (and has all manner of those dream-like AI artifacts) but I like what this person is doing with just a bunch 4 second clips and an awesome soundtrack:
Yeah, it would be way better if they just released it right away, so that political campaigns can use AI generated videos of their opponents doing horrible/stupid things right before an election and before any of the general public has any idea that fake videos could be this realistic.
you joke, but the hobbling of these 'safe' models is exactly what spurs development of the unsafe ones that are ran locally, anonymously, and for who knows what purpose.
someone really interested in control would want OpenAI or whatever centralized organization to be able to sift through the results for dangerous individuals -- part of this is making sure to stymie development of alternatives to that concept.
I do wonder why OpenAI chose the name "Sora" for this model. AI is now going to have intersectionality with Kingdom Hearts. (Atleast you don't need a PhD to understand AI.)
Sora means sky in Japanese, their reasoning is akin to "the sky's the limit".
> The team behind the technology, including the researchers Tim Brooks and Bill Peebles, chose the name because it “evokes the idea of limitless creative potential.”
It is honestly quite concerning just how good these videos look.
Like you can see some weird artifacts, but take one of these videos, compress it down to a much lower quality and with the loss of quality you might not be able to tell the difference based on these examples. Any artifacts would likely be gone.
Given what I had seen on social media I had figured anything remotely real was a few years away, but I guess not...
I guess we have just stopped worrying about the impact of these tools?
Imagine someone combining this with the Apple Vision Pro...many people will simply opt out of reality and live in a digital world. Not that this is new, but I'll entice a lot more people than ever before.
Presumably the Post-atomic horror set back technology for a while, so we should be able to expect TNG-level technology before the war. This also explains why Kirk's Enterprise uses datatapes.
Had the same thought. Seems like we’re entering the era of generative AI and mixed reality in a very real way very soon.
As much as I love the technology, I’m really not looking forward to this becoming ubiquitous. Time and time again we’ve allowed technological progress to outpace our ability to weight the societal pros ands cons.
Smartphones and the rise of image-heavy social media has rapidly changed social norms. Watch a video of people out in public 20 years ago: no screen to distract them at bus stops, concert events, or while eating dinner with friends. And if that seems trite, consider how well correlated the rise in suicide rates is with the popularity of these technologies.
Not sure if this makes me a luddite or if the feeling is common in this crowd.
but you cannot walk/feel it, just watching. It's still a huge gap to reality, less so, but you will still feel it's fake very vividly because those senses are missing.
Maybe some sort of implants that can generate senses. Would be 100s of years because you can say simulate weight/pressure and pinpoint accuracy if feeling friction.
Watching it is enough for a lot of people. Watching 1080p first person extreme sport videos on youtube is almost too compelling to me. I have to turn it off because it feels addictive.
in their research paper it says "These capabilities suggest that continued scaling of video models is a promising path towards the development of highly-capable simulators of the physical and digital world, and the objects, animals and people that live within them*." they are well set on that happening
Almost certainly troves of stock footage. The type of exaggerated motion seen in these examples is very reminiscent of stock footage. And it is heavily textually annotated for search.
Yeah, you just can't let all media, all the cost and hard work of millions of photographers, animators, filmmakers, etc be completely consumed and devalued by one company just because it's a very cool technical trick. The more powerful these services become the more obvious that will be.
What OpenAI does is amazing, but they obviously cannot be allowed to capture the value of every piece of media ever created — it'll both tank the economy and basically halt all new creation if everything you create will be immediately financially weaponized against you, if everything you create goes immediately into the Machine that can spit out a billion variations, flood the market, and give you nothing in return.
It's the same complaint people have had with Google Search pushed to its logical conclusion: anything you create will be anonymized and absorbed. You put in the effort and money, OpenAI gets the reward.
Again, I like OpenAI overall. But everyone's got to be brought to the table on this somehow. I wish our government would be capable of giving realistic guidance and regulation on this.
Not when intelligence is cheap and highly abundant. Perfecting general robotics as an improvement on humans will be quick. The upper limit of strength and consistency is much higher.
Today in the real world AI can replace very little of designers, programmers, etc. Lots of potential and extrapolation, sure. but hasn't happened. What has actually been produced by AI has been panned as not quite ripe yet.
Same with robotics. Lots of potential, but hasn't happened yet. If you read the description, Sora, is based out of trying to simulate the physical world to solve physics based problems. Something that would be perfect for the next leap in robotics.
Less glibly, no matter how good you are at following instructions, tearing out a wall filled with water than can destroy your home, fiberglass insulation that can damage your lungs and electrical wiring that can kill you will never be something I’d recommend a layman do. No matter how good the ai tutorials are.
This is the beginning of the end for many of them too. Look at the opening line of the page:
> We’re teaching AI to understand and simulate the physical world in motion, with the goal of training models that help people solve problems that require real-world interaction.
Text-to-video is just the flashy demo that everyone can understand after exposure to text-to-image. Once the model can "simulate the physical world in motion" it's only a few steps away from generic robotic control software that can automate a ton of processes that were impossible before.
Humans still have the benefit of dexterity and precise muscle control but in the vast majority of cases robots can overcome those limitations with better control software and specialized robotic end effectors. This won't soon replace someone crawling under a house or welding in awkward positions, but it could for example replace someone who flips burgers or does manual labwork.
This could eliminate the limiting factor for automating many manual processes. (ruh-roh)
Think about it. Sora demonstrate AI can understand real world physics to a scary degree.
If you use Sora like models to imagine what actions needed to be taken, then realize it, well, the only thing left is to create an arm/fingers that can took action, then you are done.
It's bimodal. AI can automate a lot of low level knowledge work, but as wide and deep as its knowledge is, it is also incredibly superficial when it comes to logic and creativity. What it's going to do is hollow out the middle class, as creative people who know how to wield AI will become wealthy while the majority of white collar workers are forced into trades.
A major follow up to GPT-4 later this year is rumored to be (far) superior at logical reasoning than GPT-4. What's likely to happen if that becomes real?
That might let it encroach more into some fields like law where it's almost good enough already. Shitty time to be a junior lawyer, firms are going to hire and promote people not for their legal skills but for their ability to manage/attract clients.
In general though, I don't think the extra reasoning ability is going to enable it to displace that much farther than it already will, GPT lives in a box and responds to prompts. When it's connected to multiple layers of real-time sensor data and self-directing, that'll be another story.
There were independent efforts to create AI agents since last year as well. AutoGPT and BabyAGI iirc. They didn’t go far probably because the LLM used was not good enough for that.
Somehow, according to that logic, and in general the logic of all AI danger hysteria, humans have no agency in determining what the limits of what AI is fed and of its use and abuse.
Some humans do - the investors and executives in AI tech companies (and the legislators who theoretically could regulate them) , who all stand to make a lot of money from every one of the "AI danger hysteria" scenarios, and are therefore highly motivated to bring them to fruition.
The rest of us have no choice. Despite millions of artists, animators, etc. all being resoundly opposed to AI art, the models that infringe their work are still allowed to exist, and it seems they're fighting a losing battle.
A lot of people are being "hysterical" because a lot of people don't have a choice.
To be clear, the problem of these scenarios is tightly intertwined with the problem of unfettered capitalism and wealth inequality in general. Food and shelter require money, and we get money by working a job. If millions of jobs disappear overnight, then of course millions of people are going to be distressed over no longer having ready access to food and shelter.
The idea of "just getting another job" doesn't scale to the destruction of entire industries employing tens of millions of people. This is how depressions are made.
The idea of "the depression will end someday" is not only not necessarily true as wealth inequality skyrockets, but is also cold comfort to the people who will lose their houses and for some, lives, due to the disruption.
A different economic system could perhaps allow us to appreciate these technological advances without worrying about them displacing our ability to live. But the American political system consistently and firmly rejects any ideas not rooted in social darwinist capitalism.
For your sake, I hope your resume is very impressive.
If millions of jobs disappear overnight it means AI is amazingly good, which means people will also have AI empowerment on a whole new level as open source trails companies by 1-2 years. Everyone will just order their AI "take care of my needs", maybe work along with it. You got to agree that we already have some amazing open models and they are only getting better - that empowerment will remain with us in times of need.
"Companies employing people" will be replaced by "people employing AI". Open models are free, small, fast, trainable and easy to use. They capture 90% of the value at 10% the cost, and are private.
"Companies employing people" getting replaced by anything is pretty dangerous in an economic system where employment is synonymous with having food and shelter. It won't matter that AI could help me keep a to-do list or generate pretty videos if I don't have a job or income.
What we're looking at is a massive decrease in the relative economic value of the average human's work. If the economic value of a hundred people is less than what the company can produce with a single human operator running AI models, then those 100 people are economically worthless, and don't get to eat.
We drastically need to tax the usage of AI models on the huge windfall they're about to create for their operators, and use that to fund universal basic income for those displaced. Generally speaking, as automation and wealth disparity skyrocket, UBI will be required to maintain any semblance of the society we currently have. I am incredibly pessimistic about the chances of that happening in any real way though.
It's automating some of the craftsmanship part, which is substantial, but in a sense, it also threatens the creative part.
It's already very tempting for large entertainment businesses to create lazy remakes as it involves less risk. Automating creative jobs will create a shift at the production level but also on the receiving end: the public.
that would never happen because someone owns the robots and rich people can afford more robots than poor people and rich people aren’t rich people if poor people aren’t poor
I would agree. While we are seeing all this creative work get automated by AI, how big of an impact would that really have on the economy?
Fully-functional autonomous driving will have a much larger economic impact - and that's just the first area where autonomous robots will come into our lives.
I'm kind of excited to see how scifi authors will tackle the generative AI revolution in their novels.
As of now, the models still need large amounts of human produced creative works for training. So you can imagine a story set in a world where large swathes of humanity are regulated to being basically gig workers for some quadrillion dollar AI megacorp where they sit around and wait to be prompted by the AI. "Draw a purple cat with pink stripes and a top hat" and then millions of freelance artists around the world start drawing a stupid picture of a cat because the model determined that it had insufficient training data to produce high quality results for the given prompt. And that's how everyone lives their lives....just working to feed the model but everything consumed is generated by the model. It's rather dystopian.
I would say it's very profitable in terms of ideas...if you put the work. The problem is that most main-market sci-fi is not about ideas, but about cool special effects and good vs bad guys.
> As of now, the models still need large amounts of human produced creative works for training.
That will likely always be the case. Even 100% synthetic data has to come from somewhere. Great synopsis! Working for hire to feed a machine that regurgitates variations of the missing data sounds dystopian. But here we are, almost there.
Agreed, by some definitions, specifically associating unrelated things, models are already creative.
Hallucinations are highly creative as well. But unless the technology changes, large language models will need human-made training substrate data for a long time to operate.
I have a novel I've been working on intermittently since the late 2000s, the central conflict of which grew to be about labor in an era of its devaluation. The big reveal was always going to be the opposite of Gibson's Mona Lisa Overdrive, that rather than something human-like turning out to be AI, society's AI infrastructure turns out to depend on mostly human "compute" (harvested in a surreptitious way I thought was clever).
I've been trying to figure out how to retool the story to fit a timeline where ubiquitous AI that can write poems and paint pictures predates ubiquitous self-driving cars.
if it was actually AI, instead of a stochastic parrot, we could ask it to design robots that could do the manual labor that we still have to do, because we haven't been able to design robots to do the manual labor.
Unfortunately, LLMs aren't intelligent in any way, so you cannot ask them to synthesize any kind of second-order knowledge.
This is why they won't take away the creative work, either. They are fundamentally incapable of creating anything new.
Come on, don't you see that the capability to understand the physical world that sora demonstrates is exactly what we need to develop those household robots? All these genAI products are just toys because they are technology demonstrators. They're all steps in the way to AGI and androids.
sensorimotor control is imo not at all the bottleneck. Teleoperated androids could do lots of useful things right now, but the AI is lacking to automate them.
well I let's say you want to make the coffee and we split that task into roughly two subtasks. The first is to imagine what motions are necessary to do that. How does the coffee cup have to move, how does your hand have to move to grasp it, etc. The second part is to find a way to use your muscles or actuators to execute those imagined actions.
I claim that the first part is the more difficult one and where we have the bottleneck currently. Furthermore, generative video AI is exactly the kind of thing that would give a model an understanding of what kinds of things have to happen in order to make coffee.
> AI is automating the creative, intellectual work and leaving the rest to us.
Indeed, there is a risk it completely devalues creative jobs. That's ironic. Even if you can still use AI creatively, it removes the pleasure of creating. Prompting feels like filling Excel sheets while also feeding a pachinko machine.
Machines have replaced a lot of blue-collar jobs alright. It's just that most of it happened during the Industrial Revolution, so we aren't even aware of all the shitty (and not-shitty-but-obsolete-nonetheless) jobs that used to exist.
I really don't think that's true. Essentially the argument is that these models are more or less just outputting the work of others. Work already done- not theoretical future work, which is what people usually criticize new technologies for.
The question here is really about whether it's sufficiently transformative, or whether that's even the right standard to be applied to generated media.
Yup! Technology is powerful. It impacts people's lives.
I love tech, but if you take the stance that it's okay to hurt people for the sake of technical progress, you get into some very dark and terrible places...
It doesn’t help that the tech industry is trying to make it seem black and white. Like you’re either endlessly optimistic and let tech run rampant or you’re a depressing doomer pessimist. We should reject this framing whenever possible.
> it's okay to hurt people for the sake of technical progress
That's a strawman. The real view is that protecting jobs that are made extinct by technology and automation is historically a bad idea because it leads to stagnation and poverty. It's better to let people lose their jobs, and for those people to find other jobs, while supporting them with a social safety net while they make the transition. Painful for them but unfortunately very necessary for a prosperous society.
Instead of using this outrage and energy to push a political will to grant something that benefits everyone forever, we should use it to grant something that helps prop up a few people in dying industries so that they can stifle innovation which would lead to a creative revolution?
What no one is asking is: 'it this makes it easy for anyone to be an artist, a director, a musician... what are we going to get, and will it be worse than what we have now?
> What no one is asking is: 'it this makes it easy for anyone to be an artist, a director, a musician... what are we going to get, and will it be worse than what we have now?
Everyone is asking this.
But that's also not the only question. The one you're ignoring here is: If these tools enable one artist to do the work of a hundred, what happens to the other 99?
AI boosters have as yet offered no satisfactory answer for this question. Given the intimate involvement some of them have with politics at the national and global level, this absence constitutes reasonable grounds for suspicion that no answer is intended or forthcoming, and that suspicion is what's asking here to be addressed.
> If these tools enable one artist to do the work of a hundred, what happens to the other 99?
Not really -- as people have gotten more efficient at their jobs, we tend to just produce more/better things, not impoverish a bunch of people. If one person can day (8 hours) making a shoe by hand, and one person can make a shoe in an hour using a shoe making machine, then we don't have one less shoe maker, we have two people making 16 shoes a day. As an effect, shoes are now much cheaper, so they aren't only worn by rich people. If the one-shoe-per-day maker refuses to use a shoe making machine, he or she can upsell their 'hand crafted' shoes to rich people who want to distinguish themselves.
Believe me, I am not a 'free market fixes everything' person, at all, but in these cases, that is how it has worked since the industrial revolution. This is not a new process (automation making a task much more accessible/efficient) and this is not a new complaint (what happens to the people who made a living doing task).
Change is scary -- and everyone has the right to be afraid of an uncertain future, but I can't recall an instance of the regressive approach actually working to allay the fears of those who imposed it. Yet, we all see huge reminders of how our lives have been improved by making hard things easier and accessible to more people.
The argument as presented so omits even the possibility of harm being done anyone in this process as to seem as if it seeks to foreclose the thought at root.
It would not surprise me if anyone called this pollyannaish, or even Panglossian.
You don't really touch at any point in your argument on even the possibility someone might be harmed, in the process of entire segments of the labor market being automated. Why is that?
It is assumed is anything with any kind of scale that harm with occur.
Did anyone get harmed when photography was used to supplant portraits? Did anyone get harmed when mail started getting sent by rail instead of horse? Did anyone get harmed when air travel became possible? Did anyone get harmed when we supplied electric power to homes?
I have an idea -- why don't you propose a solution to AI ruining creative jobs and we can apply that standard to it.
Price in the externality. The multiple of US GDP that OpenAI currently seeks in funding should certainly suffice to fund UBI, and if that slows down OpenAI's development of new capabilities, then that should still be preferable to the alternative of OpenAI being enjoined from doing business until that is done.
Of course you may respond that this is unrealistic, which it is; it requires a government capable of acting via regulation in defense of its citizens, and so nothing like it will be done.
The social safety net component of your idea is both extremely important and not at all likely in the modern ultra-capitalist, "even healthcare is socialist extremism" political atmosphere.
Maybe mass unemployment will create a sea change in that mentality, but most of the people who's opinions need to be changed will probably just laugh at "the elites" getting screwed over.
I think dismantling creative fields like this is completely different from automating manual labor in a way that makes humanity more prosperous. I don't see what the upside is of this -- it's not making creative work better, it's devaluing creative work and disenfranchising creatives.
Is it? What is another example of a technological leap that made a certain class of workers redundant while also continually relying on the output of these same workers to be feasible in the first place?
The current batch of LLMs is in the same class of technological revolutions as Napster and The Pirate Bay. Immensely impactful, sure, but mostly because of theft of value from elsewhere.
New data can still be created using AI and curation, couldn't it? New works, incorporating AI or not, still enjoy copyright protections that one can monetize by selling access to that specific work.
Things that can be easily reproduced already have little value, the people who produce those things have adapted to focus on brand, and that's just how it's going to be from now own.
Reminds me of an interview with a Korean pop music producer I watched 15 years ago.
South Korea had a high % of broadband penetration earlier than many Western countries, and as a result physical CD sales crashed very hard, and very quickly. So he asked himself, what's the most analog good I could sell? It's people. And went the pop idol / personality marketing route with great and lasting success.
I dont think they're saying its up to tech companies to decide what has value, more that the development of new technology itself ends up deciding for the rest of the world how things are valued.
It's been this way for 10,000 years since the invention of the wheel. New inventions change how things are valued by making it easier for people do more work with less time.
The creators who create media can also use these tools to create more media faster, as can novices. It's not like OpenAI literally eats the media, never to be shared with the world again.
Oh I see, they're not eating the media, just extorting the creators into paying OpenAI in perpetuity to use the tool derived from their own work, or face becoming uncompetitive with their peers who do use it. What if landlords, but for media creation, and they don't even have to pay for the land in the first place. That's fine then.
> pay a subscription to OpenAI in perpetuity in order to remain competitive with their peers
This is how technology works in general and should not be vilified. Someone comes up with a better way to do things (in this case bringing creative ideas to life) and charges a premium on top of that for their efforts. If the current wave of creators doesn't like it, then they should instead make something people want more than what their competition has to offer.
Either way, this is why local open source models are critical, so that everyone can benefit without needing to pay any single party.
If a company were founded tomorrow which allowed you to stream unlicensed TV shows and movies for a monthly subscription, undercutting Netflix and Amazons licensed streams, that wouldn't be described as "a better way to do things" just because their customers prefer it for being cheaper and easier because all the content is in one place. The difference between that and what OpenAI is doing is just degrees of abstraction, either way they're deriving value from others work without compensating them, and actively undermining the ongoing creation of the work they're appropriating, while simultaneously relying on the ongoing creation of that work to keep feeding their machine.
IP law has yet to decide whether my interpretation of the situation is correct in the legal sense, but I find it impossible to see "ChatGPT absorbs the work of writers/journalists and sells a superficially reworded version without attribution or compensation" as anything but theft obfuscated behind lots of fancy math. It's only going to get worse if LLMs end up displacing traditional search engines, so one day you'll publish an article and get exactly one impression from GPTBot which then turns around and figuratively copies your homework.
Forgive me for thinking that it may be difficult for independent artists to compete against the trillion-dollar groundbreaking plagiarism machine that is actively plagiarizing their work faster than they can produce original work, without consequence, and suffocating them under a deluge of generated works.
This is an extremely different difference of scale, which does constitute a meaningful difference from prior technologies.
It’s difficult for independent artists to live as independent artists today, even without the specter of a “trillion-dollar groundbreaking plagiarism machine”[0]. So far, we’ve still been producing original work, primarily because it’s what we do even when we’re not making money from it. It’s a blessing and a curse.
This is not to dismiss the concern. I simply wanted to state that artists will find ways to keep moving the creative bar forward.
[0] I really like this turn of phrase, thank you for sharing it.
I create media for a living, painstakingly creating stuff from scratch in 3D. This tool will not help me, it will help clients avoid ever having to contract me. The main beneficiaries of this are holders of capital
People on HN like to split hairs and make muddled juxtapositions about human rights and AI model capabilities. But this is something that people and governments around the world would have to reckon with very quickly, since the rate at which generative AI technology is advancing, there could be hundreds of millions of people who’re unemployed and have no way to find work.
The quickest way to address this would be an extremely high tax rate on any generative AI model, say 500%, while the government figures out what’s the best way to sustain an economy (such as UBI) with a diminishing set of consumers as more people are pushed towards unemployment.
I suspect what was meant is something like 500% VAT, where if a generative AI charges a customer $6, then $5 goes to the taxman and $1 to the AI company.
Taxes are meant to capture some of the profit that is made by a business entity. You could use a local model, but if you sell some kind of product or service, the tax would be levied on you. Not declaring that properly, of course, is tax evasion :)
I see the validity of this concern in the short term, but long term I feel like this is a bit doomsday. I don't want anyone's livelihood to get shafted, but realistically I see this as lowering the barrier to creating videos / proofs of concept--which is a good thing (with a lot of caveats and asterisks).
>Yeah, you just can't let all media, all the cost and hard work of millions of photographers, animators, filmmakers, etc be completely consumed and devalued by one company just because it's a very cool technical trick.
Oh man, how I miss it when ice was hauled from the Arctic in boats.
You recognize the difference, right? Modern freezers don’t rely on people shipping ice from the Arctic. Generative AI does rely on people continuing to create media.
Sorry no. If there was even the remotest possibility that everyone could be brought to the table, none of these would even exist.
Training a massive model like this is a risk, and no one is going to take that risk without some reward. You can complain OpenAI is going to too much of the value, but its value that would have otherwise never existed. It's value.
I am getting sick of these "people can't be allowed to make their own nice things easily, because of a pugnacious (and very online) interest group that wants to keep getting money" takes.
Typical argument against technological progress "We should ban technology to stop it doing what humans can do in a fraction of time and resources".
Can see this create an explosion of new Content from aspiring Film, Story tellers and cut scenes from Game creators that previously never would have the budget or capabilities to be able to see their ideas through to creation.
If we had a safety net where career progression and time/money invested in training was unnecessary to sustain life, then maybe. Until then it feels like a bit of allowing a few people to plunder and own the collective output of millions.
This moment seems like trade guilds revolting against free craftsmen. What AI is essentially doing is learning skills from people according to their works and then helping everyone according to their needs. It's more rad than open source.
This is not plunder, it is empowerment. Blocking generative AI would be a huge power grab for copyright owners. They want to claim ideas and styles, and all their possible combinations.
Gen AI need only ensure it never reproduces a copyrighted work verbatim. Culture doesn't work if we stop ideas from moving freely.
I agree that preventing technology from dispersing generally prevents the creation of wealth. However, given our current economic structure, the downside in instability of a livelihood has dramatic effects on swaths of people who were unlucky enough to be disrupted -think of the dramatic costs of retraining, healthcare access, the high costs of diminished earnings, inability to accrue wealth and retire. Perhaps we could socialize these costs, but we don't and are unlikely to do so.
Another issue to look at is the lack of ownership of the tools of your trade. In a context where many use AI models to competitively produce, hosts of AI models essentially own the access to your trade - thereby able to charge a toll, or privilege certain behaviors for any who strive to make living with these tools. (of course this is happening now with plenty of software products). The ultimate trajectory of this is not democratization of a toolset, but a transfer of wealth from labor to capital. And keep in mind that the labor share of income has been steadily declining for half a century.
The creation of wealth from AI ultimately depends on the strength of democratic and pluralistic institutions that safeguard ownership of your trade, democratized access to capital, and safeguards of welfare in the environment of creative destrcution. Otherwise you wind up with the cotton gin.
“Many were increasingly of the opinion that they'd all made a big mistake coming down from the trees in the first place, and some said that even the trees had been a bad move, and that no-one should ever have left the oceans.“
We all stand on the shoulders of giants. Yes, I want artists and other creators to be compensated fairly for any work that they contribute into training datasets, but outside of that there is no moral responsibility AI creators should feel towards those whose potential careers would be impacted.
They aren't. Every person is free to use AI or not.
Go blame your fellow consumers if you don't like the fact that they prefer AI.
These are choices that everyone makes. AI companies alone aren't forcing everyone to use their cool new tools. Instead, thats a decision that 10s of millions of people are making every day.
Research on creativity and competition points to this. Essentially, creativity occurs when there is some expectancy of increasing competitiveness. However when the expectancy of value capture from your effort becomes less clear, or diminished, creativity stops altogether.
(as pointed out in the "Freakonomics" episode highlighting this reaserch)
Interestingly a lot of movies flopped in 2023 not because of bad visuals, but because their writing was bad. Hence, I believe the demise of the movie industry is overstated. I can see completely new forms of entertainment coming out of this. Probably Youtube will be the biggest winner as the social network with the highest monetization and reach.
Does it not just shift where we (as people) perceive value? If the cost of content drops to effectively zero, it seems reasonable that we would not value it so highly. If so, it does not mean that people do not value anything, but it may mean we start associating value with new or different things. While this may disrupt industries, I do not think we have an ethical or legal duty to those industries to remain profitable.
GREAT response imo, I’ll try to remember this concise phrasing. I think this highlights that people aren’t worried so much about changes coming to them as consumers, and are much more worried about what “industries no longer remaining profitable” means for them as a laborer.
capture the value of every piece of media ever created
In what way does “I have a computer that can make movies” mean “I have captured the value of every piece of media ever created?” What do you mean by “value”? In my biased view, this amazing new technology couldn’t possibly be a better time to fix our insane notions of property, intellectual or otherwise
I don't disagree with your basic sentiment, but it's worth pointing out that, on some level, the *entirety of artificial intelligence* is not much more than a "cool technical trick."
Never ever will there be everyone at the table. This is not how the Internet works. It is not how the world and humanity work. If OpenAI doesn't do it, the next big player will. China will. Maybe it'll soon not even need China because it'll be so easy to deploy.
There is no stop now. It's too late for that. Time to think about the full development and how we'll handle that. How we as people will be able to exist next to it. What our purpose in the world is supposed to be. What the purpose of "value" is. What the purpose of "economy" or "the market" is.
It doesn't really matter, because if this is possible then it will not be exclusive to OpenAI for long. It's simply just something that can exist. There will be open source versions of everything lagging 1-2 years behind or something.
Do you feel the same about the hard work of knocker-uppers having been devalued by the invention of the alarm clock? Or is it just the (relatively) highly paid intellectual workers that "cannot be allowed" to be replaced with machines?
You can't regulate it because it will just be outsourced to another country.
Nope, we are headed towards deflation. Families that need only a single worker to support everyone, and even support extended family, and less time working overall.
It's worth remembering that "intellectual property" is an entirely artificial and fairly recent construct. Humanity did fine for thousands of years without it, and I'm not going to shed too many years if OpenAI blows it up.
Are you against records? Because the technology to record songs and play them back at your leisure killed an entire industry of live performers / instrumentalists?
The call for live music drastically shrank when it became trivial for any business or residence to play music on command.
Are you against automatic language translation? I can positively guarantee that the training data that they used to be able to create significantly better translation models was not authorized for that purpose.
The entire translator industry has been steadily shrinking ever since the invention of automatic language translation.
Etc etc etc.
There's obviously two aspects of this complex social issue right now.
1. Whether or not the usage of publicly available media as training data is legal/ethical.
2. Whether or not the output of these types of generative systems (even if they're trained on "ethical" training data) which may result in the displacement of many jobs is legal/ethical.
I'm neither for nor against AI (LLM, diffusion, video, etc), but if you are going to take a stance, then you have to be consistent in your view.
You don't get to cherry pick - I don't want to see you using chatGPT, copilot, stable diffusion, DALL-E, midjourney, sora, etc.
It's weird that a call for generative AI to be more equitable towards the people whose creative work powers it is being interpreted as somehow being against tech, against AI, or that I think technological advancement should never make jobs obsolete.
Countdown to when studios licensing this for "unlimited" episodes of your favorite series.
There was Seinfeld "Nothing, Forever" AI parody, but once the models improve enough and are cheap enough to deploy, studios will license their content for real and just have endless seasons.
Or even custom episodes. Imagine if every episode of a TV show was unique to the viewer.
Custom AI commercials would be very interesting. Instead of seeing strangers enjoying the benefits of the product, it shows you. A car commercial would show you driving, etc.
Commercials and TV episodes could have a basic "story arc" and then completely customized to the viewer.
Think about the simpson's or something. Imagine that the story of the episodes were kept, but you could swap in the characters and locations. So for instance if you lived in Nashville TN, all the simpson's episodes could be generated to show the settings as Nashville instead of Springfield.
Then you could have the AI switch out the characters to be people you want. Maybe you want to replace Lisa with an AI Simpsons version of you. Mayor Quimby with Nashville's actual mayor, etc.
> Custom AI commercials would be very interesting. Instead of seeing strangers enjoying the benefits of the product, it shows you. A car commercial would show you driving, etc.
I think it'd kind of defeat the point - I can't imagine a person that'd want their likenesses to be used to market to them. It'd be a disaster. Setting swaps are more realistic, though at the point where things get good enough for that to be possible, we may just see completely on-demand newly generated media instead of modifying what already exists.
If I saw myself onscreen telling myself to buy a product I've never seen or used: I would not buy that product or use that service. It feels violating to have your image used against your best interests (of not being manipulated to be capitalism's bitch) like that.
That is a hell-scape (to me).
Inserting yourself into shows... that's feels different, but my gut tells me advertisers will corrupt that idea quickly. Product placement...
Could you show any example of that pipeline? I'm trying to think about technology not using which would result in being cancelled, but can't come with anything
One understated aspect of AI Seinfeld is that it took many steps to differentiate it from the actual Seinfeld and create its own identity, such as the 144p visual filter and the random microwave. Those tweaks added to its charm.
If someone tried to do AI Seinfeld again in 2024, many would criticze it for not being realistic enough now that the tools to do so are now available.
I assume you would still be able to do that, just better? Like pixel art. Super Mario Bros. 3 look great despite being 36 years old. Contrast this with 3D games for the original PlayStation that have aged poorly.
I'm not sure there would be much demand for purely custom/individualized episodes beyond the novelty and maybe for fun with a group of friends. Most of the reason people watch TV or movies is for the shared experience that you can discuss with others. It could definitely drive down production costs though, hopefully HBO uses it to eventually redo Game of Thrones post season 4
Well there is always your AI girlfriend and AI friend group with the AI generated podcast breaking down the episode. (jk, sort of)
> Most of the reason people watch TV or movies is for the shared experience that you can discuss with others
I wouldn't say that. Most of the reason people watch TV is to kill time.
To be honest, I find my discussions with friends about TV shows on the decline just because of the fact that everyone is watching there own thing. So many shows and people watch them at their own pace. so most of the discussions go like this "Hey have you seen that new Netlix show X?" "No I haven't, maybe I'll check it out". Or "Oh yeah, i saw that a year ago, Its good but I don't remember the details".
Before Streaming when you had a set schedule for TV, it was way easier to discuss things because people were forced to watch programs on a certain day and there was more limited content. This led to "water cooler" conversations about what the previous nights show.
I bet if you graphed (discussions had about tv shows) / (hours watched of tv shows) that graph would trend down.
Think about little kids. My niece watches cocomelon all day long. She doesn't need to discuss it with anybody. She just wants an unlimited stream.
> I wouldn't say that. Most of the reason people watch TV is to kill time.
How annoying to see something amazing and then not be able to find anyone who also experienced it that you can ... what word mean's commiserate but in a positive way?
I'm thinking now about the astronauts that walked on the Moon and had only the few others. I think one of the astronauts bemoaned having gone to this amazing place, like some kind of wild vacation, but not being able ever to return.
You can just talk to your AI companion about it. If you involve another human there's always a chance somebody might be slightly bored or inconvenienced, so we want to avoid that.
Indeed. That is why in our family we watch broadcast or timeshifted tv and no netflix. Still it is hard to find other families like that so little tv stuff to talk about at work during lunch.
Same about music. In good ol' days, one would meet a friend to listen to cool new music together, share CDs with mp3s etc
It's actually really weird. I wanted to buy my niece some CDs for Christmas to discover 90s music, but kids don't listen music from CDs anymore. They don't have devices even. Should I buy her a Spotify gift card and send her links to Spotify via Whatsapp? It's so strange.
That would not work because that's now how people work. People watch/play media to connect to others. How can you talk about anything to anyone or have any shared culture when other people will never see what you see?
Movies, books, games, are a collective culture, not an individualist one. I don't know about you, but when I like an experience, I want to share it with others.
To be blunt about it, I can't help but imagine that the people who make such comments (and I've seen quite a few recently) are just complete philistines. They're the same people who can't draw, write, play, sing, design, or anything else and yet think they know what's good.
It's almost as if they think the purpose of art or entertainment is to stimulate some particular part of the brain and everything else between that and the screen/speakers/canvas/whatever is just an inconvenience that ought to be dispensed with as soon as technology allows.
That's entirely orthogonal to the issue it was addressing.
The point is that it doesn't matter how close the two can become (indeed, we're already pretty much there); people will always want to read stuff written by actual people (or at least a thinking being) than something purely generated by a model with no other grounding in reality.
> Imagine if every episode of a TV show was unique to the viewer.
This is the bit I don't think will happen, at least in big quantities. Half the fun of watching a popular series is being able to discuss it with epople afterwards!
I honestly expected video generation to get stuck at barely consistent 5 second clips without much movement for the next few years. This is the type of stuff I expected to maybe be possible towards the end of the decade. Maybe we really are still at the bottom of the S curve which is scary to think about.
I have a book I've written (first three parts available free at https://www.amazon.com/Summer-of-Wonders/dp/B0CV84D7GR). Is there some way to feed this to the tool and get an animated version out? Or this with some other tool(s)?
2,312 comments
[ 0.24 ms ] story [ 413 ms ] threadI hope there is at least some cherrypicking here. This also seems like some shots fired at some of the other gen video startups
this is the opposite of history
More importantly, how can these accounts subtly direct the generations to instill modern ideology or politics into "historical" images, giving them historical credibility? Think of all the subtly white supremacist "retvrn" accounts, for example, falsely recontextualizing inventions and accomplishments to support their ideology.
We all need to be thinking much more creatively and cynically about how these tools will be abused. The technology will get better. The people who want to abuse it will get smarter. And your capability to distinguish fake information is likely much worse than you believe - to say nothing of younger people who have less context and experience to form a mental "immune system".
I would say, all of them. Since the dawn of history. Actually, far before, as treachery certainly precedes speech itself by a few million years in the struggle to survive game.
Just to take a contemporary western (mostly?) thing: how did it went last time you looked straight into the eyes of kids to reveal them Santa Clauss is a lie and yes almost all adults in their society are into that evil conspiracy? And what about the adult around you deeply attached to their national myths, not even mentioning all the folklore around their afterlife beliefs?
But don’t worry, everything is going to go well, I promise and you know you can trust me. :)
Social media is really good at separating content from context, things like this will distort people's understanding of history.
if "historical" is going to be used subjectively with no further qualifying statements then the meaning of "history" will be subjucated to the context it's being presented in, I don't see it's use here as contradictory
One wonders how you might gain a representation of physics learned in the model. Perhaps multimodal inputs with rendered objects; physics simulations?
OpenAI is likely limited by how fast they are able to scale their hiring. They had 778 FTEs when all the board drama occurred, up 100% YoY. Microsoft has 221,000. It seems difficult to delegate enough headcount to all the exploratory projects of MSFT and it's hard to scale headcount quicker while preserving some semblance of culture.
I don't think what you're saying is correct though, either. All the early news outlets reported 49% ownership:
https://en.wikipedia.org/wiki/OpenAI#:~:text=Rumors%20of%20t...
https://www.theverge.com/2023/1/23/23567448/microsoft-openai...
https://www.reuters.com/world/uk/uk-antitrust-regulator-cons...
https://techcrunch.com/2023/01/23/microsoft-invests-billions...
The only official statement from Micorosft is "While details of our agreement remain confidential, it is important to note that Microsoft does not own any portion of OpenAI and is simply entitled to share of profit distributions," said company spokesman Frank Shaw.
No numbers, though.
Do you have a better source for numbers?
AI researcher at MSFT barely have more insights about OpenAI than you do reading HN.
In the same month, they were also using GPT4 in public - before OpenAI.
And they had access to GPT4 in 2022 (which was when they decided to create Bing Chat, now called Copilot).
All the current GPT4 models at MSFT are also finetuned versions (literally Creative and Precise mode runs different finetuned versions of GPT4). It runs finetuned versions since launch even...
In particular, looking at the video titled "Borneo wildlife on the Kinabatangan River" (number 7 in the third group), the accurate parallax of the tree stood out to me. I'm so curious to learn how this is working.
[Direct link to the video: https://player.vimeo.com/video/913130937?h=469b1c8a45]
If/once they get it working though, society will shift fast.
There’s an XR app called Brink Traveler that’s full of handcrafted photogrammetry recreations of scenic landmarks. On especially gloomy PNW winter days, I’ll lug a heat lamp to my kitchen and let it warm up the tiled stone a bit, put a floor fan on random oscillation, toss on some good headphones, load up a sunny desert location in VR, and just lounge on the warm stone floor for an hour.
My conscious brain “knows” this isn’t real and just visuals alone can’t fool it anymore, but after about 15 minutes of visuals + sensory input matching, it stops caring entirely. I’ve caught myself reflexively squinting at the virtual sun even though my headset doesn’t have HDR.
The DK1 I could wear for like 1 minite before feeling sick, so they are getting better ...
I am prone to sea sickness. Maybe it is related.
Yeah, but I mean who knows why. I know some people can't, my GF is one of them.
I've often wondered if im ok with it because im used to the object on head stuff (like 25 odd years of motorcycle riding/ergo helmet wearing) and close up, high fov coverage fast past gaming? (I play on a 32" maybe 70 cms from the eyes give or take.)
> I am prone to sea sickness. Maybe it is related.
I'd think it might be given my understanding of why illness in many is triggered. It's odd because I never got sick from it, but i've seen others get INCREDIBLY ill in two different ways.
1. My GF tried to use simple locomotion in a game and almost vomited as an immediate reaction
2. A friend who was fine at first, but then randomly started getting very slowly ill over a matter of like an hour, just getting more and more nausea after the fact.
It's unfortunate, because due to lack of bad feelings/nausea/discomfort etc, I love VR. I equally from those around me can see no real path forward for it as it stands today though because of those impacts and limitations.
That being said, maybe they get smaller, lighter, we learn to induce motion sickness less, I dunno. I'm not optimistic.
For games like call of duty or other hyper realistic games it very likely will be.
A large part of fighting games is the style.
The cost difference of just making bespoke art and tuning an AI system to generate it for you may not be worth it (at least right now.)
https://youtube.com/watch?v=P1IcaBn3ej0
From a few years ago, where the game is rendered traditionally and used as a ground truth, with a model on top of it that enhances the graphics.
After maybe 10-15 years we will be past the point where the entire game can be generated without obvious mistakes in consistency.
Realtime AI dialogue is already possible but still a bit primitive, I wrote a blog post about it here: https://jgibbs.dev/blogs/local-llm-npcs-in-unreal-engine
See also: https://en.wikipedia.org/wiki/Vanilla_Sky
Worth noting that Google also has Phenaki [0] and VideoPoet [1] and Imagen Video [2]
[0] https://sites.research.google/phenaki/
[1] https://sites.research.google/videopoet/
[2] https://imagen.research.google/video/
DALL-E presentation also looked cool and everyone was stoked about it. Now that we know of its limitations and oddities? YMMV, but I'd say not so much - Stable Diffusion is still the go-to solution. I strongly suspect the same thing with Sora.
They're literally taking requests and doing them in 15 minutes.
But all to be said, it is no less impressive after this new demo
and i still prefer Dalle-3 to SD.
Sure, for people who want detailed control with AI-generated video, workflows built around SD + AnimateDiff, Stable Video Diffusion, MotionDiff, etc., are still going to beat Sora for the immediate future, and OpenAI's approach structurally isn't as friendly to developing a broad ecosystem adding power on top of the base models.
OTOH, the basic simple prompt-to-video capacity of Sora now is good enough for some uses, and where detailed control is not essential that space is going to keep expanding -- one question is how much their plans for safety checking (which they state will apply both to the prompt and every frame of output) will cripple this versus alternatives, and how much the regulatory environment will or won't make it possible to compete with that.
Strictly to prompting, probably, just as that is the case with Dall-E 3 vs, say, SDXL.
The thing is, there’s a lot more that you can do than just tweaking prompting with open models, compared to hosted models that offer limited interaction options.
there are mini-people in the 2060s market and in the cat one an extra paw comes out of nowhere
Most fictional long-form video (whether live-action movies or cartoons, etc) is composed of many shots, most of them much shorter than 7 seconds, let alone 60.
I think the main factor that will be key to generate a whole movie is being able to pass some reference images of the characters/places/objects so they remain congruent between two generations.
You could already write a whole book in GPT-3 from running a series of one-short-chapter-at-a-time generations and passing the summary/outline of what's happened so far. (I know I did, in a time that feels like ages ago but was just early last year)
Why would this be different?
I partly agree with this. The congruency however needs to extend to more than 2 generations. If a single scene is composed of multiple shots, then those multiple shots need to be part of the same world the scene is being shot in. If you check the video with the title `A beautiful homemade video showing the people of Lagos, Nigeria in the year 2056. Shot with a mobile phone camera.` the surroundings do not seem to make sense as the view starts with a market, spirals around a point and then ends with a bridge which does not fit into the market. If the the different shots generated the model did fit together seamlessly, trying to make the fit together is where the difficulty comes in. However I do not have any experience in video editing, so it's just speculation.
Sarah is a video sorter, this was her life. She graduated top of her class in film, and all she could find was the monotonous job of selecting videos that looked just real enough.
Until one day, she couldn't believe it. It was her. A video of of her in that very moment sorting. She went to pause the video, but stopped when he doppelganger did the same.
Don't get overly excited until you can actually use the technology.
Good luck generating anything similar to an 80s action movie. The violence and light nudity will prevent you from generating anything.
1. https://en.m.wikipedia.org/wiki/Chumra_(Judaism)
They're trying to be all-around uncontroversial.
These video clips just generic stock clips. You cut cut them together to make a sequence of random flashy whatever, but you still can't do storytelling in any conventional sense. We don't appear to be close to being able to use these tools for the hypothetical disruptive use case we worry about.
Nonetheless, The stock video and photo people are in trouble. So long as the details don't matter this stuff is presumably useful.
I'm not supporting it in any way, I think you should be able to generate and distribute any legal content with the tools, but just giving a possible motive for OpenAI being so conservative whenever it comes to ethics and what they are making.
https://youtu.be/JClloSKh_dk
https://youtu.be/upCyXbTWKvQ
Or Meta will do it for them.
> That’s why we believe that learning from real-world use is a critical component of creating and releasing increasingly safe AI systems over time.
"We believe safety relies on real-world use and that's why we will not be allowing real-world use until we have figured out safety."
someone really interested in control would want OpenAI or whatever centralized organization to be able to sift through the results for dangerous individuals -- part of this is making sure to stymie development of alternatives to that concept.
That'll fix it.
I have no idea, just guessing...
> The team behind the technology, including the researchers Tim Brooks and Bill Peebles, chose the name because it “evokes the idea of limitless creative potential.”
Like you can see some weird artifacts, but take one of these videos, compress it down to a much lower quality and with the loss of quality you might not be able to tell the difference based on these examples. Any artifacts would likely be gone.
Given what I had seen on social media I had figured anything remotely real was a few years away, but I guess not...
I guess we have just stopped worrying about the impact of these tools?
As much as I love the technology, I’m really not looking forward to this becoming ubiquitous. Time and time again we’ve allowed technological progress to outpace our ability to weight the societal pros ands cons.
Smartphones and the rise of image-heavy social media has rapidly changed social norms. Watch a video of people out in public 20 years ago: no screen to distract them at bus stops, concert events, or while eating dinner with friends. And if that seems trite, consider how well correlated the rise in suicide rates is with the popularity of these technologies.
Not sure if this makes me a luddite or if the feeling is common in this crowd.
What OpenAI does is amazing, but they obviously cannot be allowed to capture the value of every piece of media ever created — it'll both tank the economy and basically halt all new creation if everything you create will be immediately financially weaponized against you, if everything you create goes immediately into the Machine that can spit out a billion variations, flood the market, and give you nothing in return.
It's the same complaint people have had with Google Search pushed to its logical conclusion: anything you create will be anonymized and absorbed. You put in the effort and money, OpenAI gets the reward.
Again, I like OpenAI overall. But everyone's got to be brought to the table on this somehow. I wish our government would be capable of giving realistic guidance and regulation on this.
But in reality it seems like the opposite is going to be true. AI is automating the creative, intellectual work and leaving the rest to us.
AI is cheaper than a high paid designer, developer, writer, etc.
A robot is more expensive than a human laborer.
It's really funny to see the squirm from those thinking truckers would be automated away, not them.
Not when intelligence is cheap and highly abundant. Perfecting general robotics as an improvement on humans will be quick. The upper limit of strength and consistency is much higher.
It is currently more expensive to build a robot for many tasks than it is to have a human do it.
> Perfecting general robotics as an improvement on humans will be quick.
It has not been nor is there any indication it will be.
Same with robotics. Lots of potential, but hasn't happened yet. If you read the description, Sora, is based out of trying to simulate the physical world to solve physics based problems. Something that would be perfect for the next leap in robotics.
There's no physical task that robots have replaced humans for me.
Hell, even the roomba sucks (pun intended) and my wife has to pick up the slack.
https://www.youtube.com/watch?v=kCKN8k-OFG8
Less glibly, no matter how good you are at following instructions, tearing out a wall filled with water than can destroy your home, fiberglass insulation that can damage your lungs and electrical wiring that can kill you will never be something I’d recommend a layman do. No matter how good the ai tutorials are.
> We’re teaching AI to understand and simulate the physical world in motion, with the goal of training models that help people solve problems that require real-world interaction.
Text-to-video is just the flashy demo that everyone can understand after exposure to text-to-image. Once the model can "simulate the physical world in motion" it's only a few steps away from generic robotic control software that can automate a ton of processes that were impossible before.
Humans still have the benefit of dexterity and precise muscle control but in the vast majority of cases robots can overcome those limitations with better control software and specialized robotic end effectors. This won't soon replace someone crawling under a house or welding in awkward positions, but it could for example replace someone who flips burgers or does manual labwork.
This could eliminate the limiting factor for automating many manual processes. (ruh-roh)
If you use Sora like models to imagine what actions needed to be taken, then realize it, well, the only thing left is to create an arm/fingers that can took action, then you are done.
In general though, I don't think the extra reasoning ability is going to enable it to displace that much farther than it already will, GPT lives in a box and responds to prompts. When it's connected to multiple layers of real-time sensor data and self-directing, that'll be another story.
https://www.theinformation.com/articles/openai-shifts-ai-bat...
There were independent efforts to create AI agents since last year as well. AutoGPT and BabyAGI iirc. They didn’t go far probably because the LLM used was not good enough for that.
The rest of us have no choice. Despite millions of artists, animators, etc. all being resoundly opposed to AI art, the models that infringe their work are still allowed to exist, and it seems they're fighting a losing battle.
A lot of people are being "hysterical" because a lot of people don't have a choice.
To be clear, the problem of these scenarios is tightly intertwined with the problem of unfettered capitalism and wealth inequality in general. Food and shelter require money, and we get money by working a job. If millions of jobs disappear overnight, then of course millions of people are going to be distressed over no longer having ready access to food and shelter.
The idea of "just getting another job" doesn't scale to the destruction of entire industries employing tens of millions of people. This is how depressions are made.
The idea of "the depression will end someday" is not only not necessarily true as wealth inequality skyrockets, but is also cold comfort to the people who will lose their houses and for some, lives, due to the disruption.
A different economic system could perhaps allow us to appreciate these technological advances without worrying about them displacing our ability to live. But the American political system consistently and firmly rejects any ideas not rooted in social darwinist capitalism.
For your sake, I hope your resume is very impressive.
"Companies employing people" will be replaced by "people employing AI". Open models are free, small, fast, trainable and easy to use. They capture 90% of the value at 10% the cost, and are private.
What we're looking at is a massive decrease in the relative economic value of the average human's work. If the economic value of a hundred people is less than what the company can produce with a single human operator running AI models, then those 100 people are economically worthless, and don't get to eat.
We drastically need to tax the usage of AI models on the huge windfall they're about to create for their operators, and use that to fund universal basic income for those displaced. Generally speaking, as automation and wealth disparity skyrocket, UBI will be required to maintain any semblance of the society we currently have. I am incredibly pessimistic about the chances of that happening in any real way though.
It's already very tempting for large entertainment businesses to create lazy remakes as it involves less risk. Automating creative jobs will create a shift at the production level but also on the receiving end: the public.
Fully-functional autonomous driving will have a much larger economic impact - and that's just the first area where autonomous robots will come into our lives.
As of now, the models still need large amounts of human produced creative works for training. So you can imagine a story set in a world where large swathes of humanity are regulated to being basically gig workers for some quadrillion dollar AI megacorp where they sit around and wait to be prompted by the AI. "Draw a purple cat with pink stripes and a top hat" and then millions of freelance artists around the world start drawing a stupid picture of a cat because the model determined that it had insufficient training data to produce high quality results for the given prompt. And that's how everyone lives their lives....just working to feed the model but everything consumed is generated by the model. It's rather dystopian.
That will likely always be the case. Even 100% synthetic data has to come from somewhere. Great synopsis! Working for hire to feed a machine that regurgitates variations of the missing data sounds dystopian. But here we are, almost there.
1. Interacting with the randomness of the world
and
2. Thinking a lot, going in loops and thought loops and seeing what they discover.
I don't expect them to need humans forever.
Hallucinations are highly creative as well. But unless the technology changes, large language models will need human-made training substrate data for a long time to operate.
I've been trying to figure out how to retool the story to fit a timeline where ubiquitous AI that can write poems and paint pictures predates ubiquitous self-driving cars.
Unfortunately, LLMs aren't intelligent in any way, so you cannot ask them to synthesize any kind of second-order knowledge.
This is why they won't take away the creative work, either. They are fundamentally incapable of creating anything new.
Yes that’s my whole point…
I claim that the first part is the more difficult one and where we have the bottleneck currently. Furthermore, generative video AI is exactly the kind of thing that would give a model an understanding of what kinds of things have to happen in order to make coffee.
Indeed, there is a risk it completely devalues creative jobs. That's ironic. Even if you can still use AI creatively, it removes the pleasure of creating. Prompting feels like filling Excel sheets while also feeding a pachinko machine.
The question here is really about whether it's sufficiently transformative, or whether that's even the right standard to be applied to generated media.
I love tech, but if you take the stance that it's okay to hurt people for the sake of technical progress, you get into some very dark and terrible places...
That's a strawman. The real view is that protecting jobs that are made extinct by technology and automation is historically a bad idea because it leads to stagnation and poverty. It's better to let people lose their jobs, and for those people to find other jobs, while supporting them with a social safety net while they make the transition. Painful for them but unfortunately very necessary for a prosperous society.
This is the part that no one is expecting to see actually happen, though. Without that addressed, your argument is sound but footless.
What no one is asking is: 'it this makes it easy for anyone to be an artist, a director, a musician... what are we going to get, and will it be worse than what we have now?
Everyone is asking this.
But that's also not the only question. The one you're ignoring here is: If these tools enable one artist to do the work of a hundred, what happens to the other 99?
AI boosters have as yet offered no satisfactory answer for this question. Given the intimate involvement some of them have with politics at the national and global level, this absence constitutes reasonable grounds for suspicion that no answer is intended or forthcoming, and that suspicion is what's asking here to be addressed.
Not really -- as people have gotten more efficient at their jobs, we tend to just produce more/better things, not impoverish a bunch of people. If one person can day (8 hours) making a shoe by hand, and one person can make a shoe in an hour using a shoe making machine, then we don't have one less shoe maker, we have two people making 16 shoes a day. As an effect, shoes are now much cheaper, so they aren't only worn by rich people. If the one-shoe-per-day maker refuses to use a shoe making machine, he or she can upsell their 'hand crafted' shoes to rich people who want to distinguish themselves.
Believe me, I am not a 'free market fixes everything' person, at all, but in these cases, that is how it has worked since the industrial revolution. This is not a new process (automation making a task much more accessible/efficient) and this is not a new complaint (what happens to the people who made a living doing task).
Change is scary -- and everyone has the right to be afraid of an uncertain future, but I can't recall an instance of the regressive approach actually working to allay the fears of those who imposed it. Yet, we all see huge reminders of how our lives have been improved by making hard things easier and accessible to more people.
It would not surprise me if anyone called this pollyannaish, or even Panglossian.
Did anyone get harmed when photography was used to supplant portraits? Did anyone get harmed when mail started getting sent by rail instead of horse? Did anyone get harmed when air travel became possible? Did anyone get harmed when we supplied electric power to homes?
I have an idea -- why don't you propose a solution to AI ruining creative jobs and we can apply that standard to it.
Of course you may respond that this is unrealistic, which it is; it requires a government capable of acting via regulation in defense of its citizens, and so nothing like it will be done.
Maybe mass unemployment will create a sea change in that mentality, but most of the people who's opinions need to be changed will probably just laugh at "the elites" getting screwed over.
Hurt is a very subjective word in this context, how many people do you think the invention of the steam engine hurt? Or the electricity?
The upside is that creative works are completely democratized.
Now, anyone, with very little effort is fully empowered to create creative works on their own and there is no barrier to entry.
Yes, empowerment and democratization harms people who's livelyhood depends on disenfranchisement.
The current batch of LLMs is in the same class of technological revolutions as Napster and The Pirate Bay. Immensely impactful, sure, but mostly because of theft of value from elsewhere.
The factories that replaced the artisans were only made possible by the work of the artisans forging the way.
AI is nothing like anything we’ve seen, and is truly unique in the dangers it poses to the world.
South Korea had a high % of broadband penetration earlier than many Western countries, and as a result physical CD sales crashed very hard, and very quickly. So he asked himself, what's the most analog good I could sell? It's people. And went the pop idol / personality marketing route with great and lasting success.
Nonsense. Also, my point is that it shouldn't be up to tech companies to unilaterally decide what has value.
It's been this way for 10,000 years since the invention of the wheel. New inventions change how things are valued by making it easier for people do more work with less time.
Instead it is up to the consumers.
If consumers choose to give money to AI company, and not to artists, then in the eyes of the consumer those artists do not have value.
This is how technology works in general and should not be vilified. Someone comes up with a better way to do things (in this case bringing creative ideas to life) and charges a premium on top of that for their efforts. If the current wave of creators doesn't like it, then they should instead make something people want more than what their competition has to offer.
Either way, this is why local open source models are critical, so that everyone can benefit without needing to pay any single party.
IP law has yet to decide whether my interpretation of the situation is correct in the legal sense, but I find it impossible to see "ChatGPT absorbs the work of writers/journalists and sells a superficially reworded version without attribution or compensation" as anything but theft obfuscated behind lots of fancy math. It's only going to get worse if LLMs end up displacing traditional search engines, so one day you'll publish an article and get exactly one impression from GPTBot which then turns around and figuratively copies your homework.
This is an extremely different difference of scale, which does constitute a meaningful difference from prior technologies.
This is not to dismiss the concern. I simply wanted to state that artists will find ways to keep moving the creative bar forward.
[0] I really like this turn of phrase, thank you for sharing it.
That's not what extortion is. Stop abusing language.
You can deliver more content, faster, cheaper.
The quickest way to address this would be an extremely high tax rate on any generative AI model, say 500%, while the government figures out what’s the best way to sustain an economy (such as UBI) with a diminishing set of consumers as more people are pushed towards unemployment.
How are you going to stop me from doing that?
Even the free and open source stuff will destroy industries and you can't confiscate everyone's consumer gamer PCs.
Taxing the big guys doesn't save creative industries. It's a lost cause.
The only thing I can imagine is like limiting people’s compute power
But even then they’d just go do it in another country or use an online service based in another country
There isn't a way to "capture" the value from that.
Even if you aren't directly selling AI assets to someone else, people simply using AI themselves will still destroy industries.
Good luck confiscating everyone's graphics cards. The cat is is out of the box already.
> the tax would be levied on you
No it wouldn't. AI is already everywhere. Its game over. You aren't going to be able to track basically anyone who is using local models or other AI.
Oh man, how I miss it when ice was hauled from the Arctic in boats.
Training a massive model like this is a risk, and no one is going to take that risk without some reward. You can complain OpenAI is going to too much of the value, but its value that would have otherwise never existed. It's value.
Who's "you"?
Automate away the lower classes all you want, just don't touch the white collar class, that's a heckin' nono.
Can see this create an explosion of new Content from aspiring Film, Story tellers and cut scenes from Game creators that previously never would have the budget or capabilities to be able to see their ideas through to creation.
This is not plunder, it is empowerment. Blocking generative AI would be a huge power grab for copyright owners. They want to claim ideas and styles, and all their possible combinations.
Gen AI need only ensure it never reproduces a copyrighted work verbatim. Culture doesn't work if we stop ideas from moving freely.
Another issue to look at is the lack of ownership of the tools of your trade. In a context where many use AI models to competitively produce, hosts of AI models essentially own the access to your trade - thereby able to charge a toll, or privilege certain behaviors for any who strive to make living with these tools. (of course this is happening now with plenty of software products). The ultimate trajectory of this is not democratization of a toolset, but a transfer of wealth from labor to capital. And keep in mind that the labor share of income has been steadily declining for half a century.
The creation of wealth from AI ultimately depends on the strength of democratic and pluralistic institutions that safeguard ownership of your trade, democratized access to capital, and safeguards of welfare in the environment of creative destrcution. Otherwise you wind up with the cotton gin.
Yeah, this is why "AI creators" shouldn't be the ones unilaterally deciding how this all plays out.
Go blame your fellow consumers if you don't like the fact that they prefer AI.
These are choices that everyone makes. AI companies alone aren't forcing everyone to use their cool new tools. Instead, thats a decision that 10s of millions of people are making every day.
Copyright should have ended decades ago. It has accomplished nothing but harm.
(as pointed out in the "Freakonomics" episode highlighting this reaserch)
https://direct.mit.edu/rest/article-abstract/102/3/583/96779...
https://freakonomics.com/podcast/can-a-i-take-a-joke/
Means for us :(
That's one big trick, almost magical.
There is no stop now. It's too late for that. Time to think about the full development and how we'll handle that. How we as people will be able to exist next to it. What our purpose in the world is supposed to be. What the purpose of "value" is. What the purpose of "economy" or "the market" is.
Exiting times.
Nope, we are headed towards deflation. Families that need only a single worker to support everyone, and even support extended family, and less time working overall.
https://www.gnu.org/philosophy/not-ipr.en.html
https://www.gnu.org/philosophy/not-ipr.en.html
The call for live music drastically shrank when it became trivial for any business or residence to play music on command.
Are you against automatic language translation? I can positively guarantee that the training data that they used to be able to create significantly better translation models was not authorized for that purpose.
The entire translator industry has been steadily shrinking ever since the invention of automatic language translation.
Etc etc etc.
There's obviously two aspects of this complex social issue right now.
1. Whether or not the usage of publicly available media as training data is legal/ethical.
2. Whether or not the output of these types of generative systems (even if they're trained on "ethical" training data) which may result in the displacement of many jobs is legal/ethical.
I'm neither for nor against AI (LLM, diffusion, video, etc), but if you are going to take a stance, then you have to be consistent in your view.
You don't get to cherry pick - I don't want to see you using chatGPT, copilot, stable diffusion, DALL-E, midjourney, sora, etc.
There was Seinfeld "Nothing, Forever" AI parody, but once the models improve enough and are cheap enough to deploy, studios will license their content for real and just have endless seasons.
Or even custom episodes. Imagine if every episode of a TV show was unique to the viewer.
Commercials and TV episodes could have a basic "story arc" and then completely customized to the viewer.
Think about the simpson's or something. Imagine that the story of the episodes were kept, but you could swap in the characters and locations. So for instance if you lived in Nashville TN, all the simpson's episodes could be generated to show the settings as Nashville instead of Springfield.
Then you could have the AI switch out the characters to be people you want. Maybe you want to replace Lisa with an AI Simpsons version of you. Mayor Quimby with Nashville's actual mayor, etc.
I think it'd kind of defeat the point - I can't imagine a person that'd want their likenesses to be used to market to them. It'd be a disaster. Setting swaps are more realistic, though at the point where things get good enough for that to be possible, we may just see completely on-demand newly generated media instead of modifying what already exists.
That is a hell-scape (to me).
Inserting yourself into shows... that's feels different, but my gut tells me advertisers will corrupt that idea quickly. Product placement...
If anything it was a turn off and I was confused how they knew where I worked.
I used to get ones that said “Comcast user you are insecure” and stuff.
Majority Report spoke to one of the negotiators and national board member:
https://www.youtube.com/watch?v=E62k1ZsY1IU
If someone tried to do AI Seinfeld again in 2024, many would criticze it for not being realistic enough now that the tools to do so are now available.
> Most of the reason people watch TV or movies is for the shared experience that you can discuss with others
I wouldn't say that. Most of the reason people watch TV is to kill time.
To be honest, I find my discussions with friends about TV shows on the decline just because of the fact that everyone is watching there own thing. So many shows and people watch them at their own pace. so most of the discussions go like this "Hey have you seen that new Netlix show X?" "No I haven't, maybe I'll check it out". Or "Oh yeah, i saw that a year ago, Its good but I don't remember the details".
Before Streaming when you had a set schedule for TV, it was way easier to discuss things because people were forced to watch programs on a certain day and there was more limited content. This led to "water cooler" conversations about what the previous nights show.
I bet if you graphed (discussions had about tv shows) / (hours watched of tv shows) that graph would trend down.
Think about little kids. My niece watches cocomelon all day long. She doesn't need to discuss it with anybody. She just wants an unlimited stream.
How annoying to see something amazing and then not be able to find anyone who also experienced it that you can ... what word mean's commiserate but in a positive way?
I'm thinking now about the astronauts that walked on the Moon and had only the few others. I think one of the astronauts bemoaned having gone to this amazing place, like some kind of wild vacation, but not being able ever to return.
It's actually really weird. I wanted to buy my niece some CDs for Christmas to discover 90s music, but kids don't listen music from CDs anymore. They don't have devices even. Should I buy her a Spotify gift card and send her links to Spotify via Whatsapp? It's so strange.
Movies, books, games, are a collective culture, not an individualist one. I don't know about you, but when I like an experience, I want to share it with others.
It's almost as if they think the purpose of art or entertainment is to stimulate some particular part of the brain and everything else between that and the screen/speakers/canvas/whatever is just an inconvenience that ought to be dispensed with as soon as technology allows.
https://future-sf.com/fiction/1700/
The point is that it doesn't matter how close the two can become (indeed, we're already pretty much there); people will always want to read stuff written by actual people (or at least a thinking being) than something purely generated by a model with no other grounding in reality.
This is the bit I don't think will happen, at least in big quantities. Half the fun of watching a popular series is being able to discuss it with epople afterwards!