OpenAI Just Changed Video Forever
Вставка
- Опубліковано 14 лют 2024
- OpenAI just announced their stunning new AI Video Model, Sora. This is by far the most stunning AI Video Generator we have seen yet.
Let's dive in and check it out!
LINK:
openai.com/sora - Наука та технологія
This with multimodal input will be wild(er). I mean the moment you can specify the characters, voices, script and performance notes.... it'll be able to make a full movie
Mind...Blown...emoji. I can now very clearly see Joe Russo's claim of Full Ai generated movies in (less now) two years.
No question. I see complete scenes, even the ability to load a full screenplay into AI and having it render a movie out, perhaps by the end of 2025. Maybe sooner at the insane rate things are quickening.
i would love a full movie with its own scores and story, and you could just make anything just tell it what basic settings for the movie and it makes a whole film!
@@dave230k33kHonestly I would just use it to upscale LOTR, remaster audio, fix bloopers like Eomer's sword falling out of his sheet and maybeeee (Not sure) alter some scenes or dialogue
No Ai video generator, no matter how perfect, will make a bad story better. We must not lose touch with our stories and ideas, as overwhelming as that may be.
Imagine creating historically footage for anything and it looks perfectly real…even for the ancient past
It’s going to be awesome. I keep thinking of the future kid in high school who hands in a movie instead of a paper!
Comment like this make me wonder if people really understand how "AI" works.
In like a majority of movies, no shot is ever more than 30secs before the next shot kicks in. And Sora here can output a 60s long shot.
So basically, just with Sora here, when finally released, a filmmaker can build a film, shot by shot and scene by scene.
For instance, they can start with a prompt for an establishing 40sec shot, then a 10sec closeup shot of character A, another shot of character B, another back to A. Until all the shots are done for that scene.
They then move to the next scene and generate all the necessary shots, each under 60secs, needed for that scene as well.
Rinse and repeat till the film is done.
Exciting. So in like 2025 and beyond, such films just might be possible.
Very possible. The hard part will perhaps be to have coherence between shots, so that the content and characters look the same in shot A and shot B.
I’m in no doubt this will be solved over time though.
A film without dialogue. I mean I'm sure it will happen at some point, but at first it would be silent films. More useful for music videos
@@shredd5705 Dialog will 100% certainly be possible in the future as well. It is in fact already possible to put lip-sync on video now, and this will be further refined until you can't tell the difference. I'm willing to bet, It is within reach in just a few years.
@@martin231181 Not doubting that at all. I don't think it will even take a few years. But this alone won't make anyone a film maker. It can be used for advertisement, music videos, as supporting tool for real film making. It's going to put a lot of people out of jobs. Animators, VFX artists, b-roll crews
However, I don't think people having wide access to film making will somehow make entertainment "better". Netflix has already too much things to watch, I often give up choosing and don't watch anything. If randoms can create ton of videos and put it online, it will become a sea of bad story telling, nobody will find the good indie films after a while. Something like Kung Fury kickstarter would go unnoticed
Similar thing is happening with art, Google Image search for example is flooded with bad AI art, when you look for real art the search results have become poisoned
@@martin231181 With exponential technology growth in mind.. I'm gonna make a wild bet and say 1 year by 2025 March. I wouldn't be surprised if it is possible as soon as 3-6 months. Maybe it'll happen next month and shock everyone lol
The dirt kicked up by the Range Rover with the tree shadows was amazing.
Seriously, I have not seen anything even close to that with any ai generated vehicle footage. It’s nuts.
The one that looked like a western is the most mindblowing to me. Mostly because they wouldn't have a drone to make that shot back in the day. It's damn near time travel.
Until you start to look at all random details that either doesn't make sense or you will never see in another shot. But it's like a dream.
Keeping up with all the advances in AI is an exercise in itself! Great update!
Thank you so much! Sometimes I feel like that dude running backward on the treadmill!
@@TheoreticallyMedia do you use AI to help you? 😁
Going to start writing all my interesting dreams down from now on. “Prompt my dream into a movie” coming soon lol
That is such a great idea! I have been meaning to jot down/journal in the AM as well. Your idea might just have solidified that!
But first…coffee.
This is mind blowing. VR will not be the same with this tech O_O
VR will never be able to run this tech.
@@T_Time_never is a long time. Not serious. But I’m the mean time we can stream it to vr.
Absolutely mind blowing! I already saw the video's on the website yesterday. It gave me the same goosebumps as when I saw the first Midjourney pictures. I wonder when this will become widely available, and I wonder how quick this platform develops. I mean, imagine (lol) if they also add image + prompt to video so you can bring Midjourney images to live. There are a lot of opportunities for creative people here (like making whole music video's, and movies even). The downside is: you simply can't believe anything you see on the news or where ever anymore.
Sora is choosing to vignette those eye shots, because when you do eye closeups with real world cameras, the depth of field kicks in when you're focused on the eye (when you have the right settings). So it is probably mimicking the shots it trained on.
Yeah, I think I might have jumped the gun on that. I just noticed it in a lot of the example shots, but now that I'm viewing more of them, I think you are correct-- it is specific shots that call for it.
Every week I see notices of new advancements, and every week I say “things will be amazing in a few months.” Yet it’s reaching the point where it’s often just days before another large step is taken forward.
The year 2030, I mean 2025, will be incredible!
OpenAI just stold Runway's lunch off the table. I seriously cannot wait to get my hands on this, especially when it has image to video. The creative possibilities are endless. It will be interesting to see if we can create a consistent look and feel for full music video projects and short movies. Same aesthetic but with different locations and so on.
Was actually just checking this... I'm blown away... Thanks for your videos. Great as always!
BLOWN. AWAY!! And thank you as well!! This was a bit of a rush to get out, and I REALLY wasn't planning on making one today!
I was not expecting to see a Renaissance today
Seriously, was not on my bingo card for the month! Crazy!
@@TheoreticallyMedia I need to convince them to give me access for conveniently short form content length videos
Back in the day, when we were saying, Eh, it's not quite there yet.
Yeah, that was a few months ago!
Thanks for this video. That _really_ is insane! When you think of an entire workflow for a presentable one-minute video, this just changed everything forever. The video title is spot on!
For those glitches. I wonder if they could use an adversarial Network to detect it and compute the loss. The adversal network is simply trained on video and catching irregularities.
I'm sure by the time we see a full release (or something near to it), those glitches will be ironed out.
I'll actually be a bit bummed when all the surreal AI stuff is stamped out for good. It really can make me laugh pretty hard.
Full star trek holodeck is like 2 years away, give or take.
I also want the replicator for me Earl Gray Tea! That part of AI might be a lot longer off though!
@@TheoreticallyMedia Something tells me it'll be one of the first machines super-AGI will want to build.
I love the fact we've don't need to rely solely on stock footage for videos! I hated that fact it takes so long to find the right footage for what you are trying to convey! Now I can create something original
And you figure between this and video Inpainting? There really isn’t anything you can’t visually represent! Insane!
@TheoreticallyMedia I'm obsessed with dalle-e in chatgpt I think I may have an addition 😳
Now I have runway and I am not sleeping because I am learning that one too! HELP LMAO
love your videos
@@TheoreticallyMediauntil we get the open source version and it doesn’t give us sad dog pictures for everything 😂
Thanks for keeping us up to date with a truly awesome development for creatives!
Just saw Emad's comments over rSD, he's implied they have some sort of counter model in the works, but I suspect even if it does come out in the next 6 months it wont touch Sora, not that it has to though, so long as its open source and even half as good.
Interesting, I’ll have to check that out. I saw someone speculating that Sora is using Metahuman to generate? That’s an interesting idea.
For sure, things are REALLY shaking up for 24!
@@TheoreticallyMedia I saw someone mention UE5 being involved as well, but I don't see how. I've not seen or heard of anything announced between them.
Have to say, Sora blindsided me, I genuinely thought we'd see those features in a year or two from now. Crazy times, cant wait to get my hands on it.
I've never been more excited and concerned about an AI announcement
Do you think that anyone involved in the making of the GTA series saw that driving sequence? Wonder what they would think if they saw it?
Oh, I think we're going to see a lot of this tech hit games before we see it hit big budget films. Given how hard Nvidia is leaning into AI, I can't see it not happening!
The age of filmmaking lasted only a little over 100 years. Incredible.
Thanks for keeping us updated, if Sora is truly as good as it looks here that is mind-blowing!
This is OpenAI’s response to the Superbowl ads making fun of AI video capabilities.
It is in the best interest of whoever made those ads to disillusion people into thinking AI isn’t that great so they don’t lose their jobs.
It is in OpenAI’s best interest for everyone to realize how great AI is so they can get the 7 trillion $ they are asking for.
We are seeing giant corporations now in a massive battle for the future of media consumption, they have no idea what to do, and the people are in power with access to AI tools that can replace entire companies of media production.
There was for sure a little finger pointing and “ha-ha’ ing” in those Super Bowl spots. But I think it’ll be a very quiet day at the Ad agency tomorrow morning.
This is a real shell shocked moment.
What a revolution, and you're so right, this period of experimentation with old models is going to be totally nostalgic for all of us. It was really great to be a part of it, now a new world is opening up.
It might be interesting to remaster some of our older AI Videos once we get our hands on this. I wonder what the Sora version of Sh*t would look like?
Also, I keep thinking of video inpainting when this hits...or masking out our "actors" and green screening...or all the other post VFX stuff that can be done to generated footage. It's going to be insane!
@@TheoreticallyMedia Oh yes, that's a good idea, I'm sure I'll be sending all my old prompts into it anyway to see what kind of world it'll generate. I hope we don't have to wait three months for that. The value of current tools is falling fast. 😥
I'm blown away, and can't wait to see it come out. Any idea when Sora will be released?
The news about Sora drop like a bomb and I even share the videos with some members of my family who are not following this space like I do only because I think this is a giant leap forward that we will see the repercussions of sooner than expected.
I'm doing a talk on AI Video next week...and yeah- for those who aren't following this space? When I show this, I'm going to see some melted faces.
...I take some glee in that?
Great (quick) breakdown. The outputs are so impressive, such a leap from Pika and Runway.
I cannot believe we're here already. I expected this to be toward the end of the year! Geez, now where are we going to be in December?
@@TheoreticallyMedia .. 100% this. It's only the 16th of Feb!!, just a few weeks ago we were all excited by hyped-up updates to Pika.
A huge upgrade. WHEN can we access it? (I'm sure you or Marques Brownlee will let us know 😀) Great work, OpenAI team.
Ha! MKB will surely know before me! But, I’ve got some feelers out! I’ll let you know as soon as I touch base!
As an independent recording artist without a tonne of money to drop on videos and such, this is really exciting.
Oh, 100%! Man, you see those guitars on the wall? I know EXACTLY where you're coming from!
The quality of the imagery is amazing, and the coherence is downright mind blowing. That all said, when taken as a whole, there are definitely still a lot of things that act on you subconsciously in such a way that you would never be convinced it is real video. Yet :)
Oh for sure. We’re going to be in quite the uncanny valley for a bit. But…only a matter of time.
This is... incredible. And if rumors are to be believed to be true, February has not quite dropped its biggest banger yet. Hope to hear more soon in the next couple weeks.
March. March is going to be INSANE. Just like March of 2023. Beware those ides!
Another major landmark day. Amazing! Can’t wait to get my hands on it!
This will be perfect to get a quick establishing or insert shot while editing the final picture. Any idea if you can train this with a model of the interior of a set actor or prop?
Unknown as of now, but Stability is teasing something just as good, and if that is open source, then you should be able to!
Give up folks, every single person on the planet has just become a filmmaker. That once rare polished diamond has just turned to a dusty pebble.
500 million animators/filmmakers have just entered the chat. RIP, it was great knowing you :(
I mean, you still gotta have a great story, and something amazing going on. Signal to Noise will get pretty loud, I agree-- but I can argue that it's already pretty loud.
There's some shell shock right now for sure, but it won't be long until we're able to recognize that "look"-- and it'll still take talented filmmakers and artists to smooth it over and turn footage like this into something compelling. Cream always rises and all that..
@@TheoreticallyMedia maybe. But on the other hand maybe we are entering a new dawn where these long format movies are dying off and lots are moving onto trends/memes/virals.
Not saying 100% eradication but a hell of a lit of ppl will be able to create and the pool is getting too large and oversaturated.
The day where we play a video game with ultra realistic images is closer and closer... and developpers will be able to create these games with much less effort. Excited for the future of video games but at the same time I fear for the jobs of workers into this industry. It will also probably change the whole movie landscape including vfx. I can see a future where the viever will be able to customize their viewing experience. Like watching an old movie and getting prompts like "every character is Elija Wood" or "Sauron wins at the end" or "I want to watch LOTR but it is in the style of horror movies".
I was talking about that yesterday to an Uber driver who had no idea about what was going on in AI. I think I kinda blew his mind.
I am impressed by the advancement in temporal continuity, but I have three issues with this first iteration:
Text: Text remains a problem. I don’t think it’s a coincidence that the first scene is set in Tokyo and not New York, where the signs would be in English.
Human Realism: While the facial features are great, there is no facial movement. Additionally, when there is movement (such as the cat in bed), it becomes obvious that it’s AI-generated. Any hand movement also exhibits a familiar AI glitch.
Media Transition: There is no image-to-video or video-to-video transition to see how it combines different modes. Currently, it relies solely on text prompts.
All good points! We'll see what (and if) the final release looks like-- but, as goes the now common phrase: This is the worst it will ever look.
Finally AI video that is actually good and ALMOST useable. The stylish lady's left and right legs switch places and her ankles are weird. But dang, this is still much better than anything else. Time for Runway to run away.
I think at these generation lengths, you'll be able to get some really workable footage going that you can cut around the jank. Curious to see how much frame control we're going to have...
I say we will see a plethora of short films when this bad boy hits. This time with AI is nuts - I plan to do a project using the tools available and then something new hits thats better tech and I feel I'm wasting time if I don't use the new thing. Now this! I won't ever get a personal project done, although, if this is what we see I doubt anyone will get close anytime soon. Until MidJourney does their thing.
I cannot WAIT to see what MJ has up their sleeve. I'll say this, I feel bad for anyone who is about to drop a Video Model. OpenAI just set a massive bar. Anything less than this is going to look anemic.
ME TOO. I'm intrigued to know if MJ saw this coming. I feel they all are somewhat integrated internally with the amount of people involved and those with loose lips. If you look at the image gen comparison between MJ and DALL-E, I think MJ wins hands down - and if that is any indicator on the tech they each possess - MJ just might be better or at least as good out of the gate with text to video (and that's scary) haha. Gonna be some fun videos for you to make when MJ does roll it out! Bruce Lee Terminator should now officially be in Pre-Production! @@TheoreticallyMedia
This is insane... you can definitely create music, art and info shots with this.. if it can generate consistently
I would use the biggest cliche on this one: "It is there" .This totally changes everything.
I always get in trouble for using the "Everything Just Changed" thumbnail-- but in this case: It's TRUE!
@@TheoreticallyMedia You are right. It did change everything. Now I know why some people said a few months ago that we will see a full movie made by AI (text-to-vide)o this year.
Filmmakers will be making use of this technology. I would. This is the first channel i will be watching once this is release.
If you can live with a lot of oddities in the details and have trouble to use it for multiple scenes
So they're reaching out to artists, designers, and filmmakers to help improve the thing that will ultimately put them out of a job. Very nice lol
At least they have a job...that doesn't pay anything.
Thanks for this, Tim. Really looking forward to trying Sora.
SAME!
Between this and another company, you say? I may have misheard, but did you say there's a model that can do 60 minutes soon? Or was it seconds? Was it OpenAI?
Seconds. And that’s this model. Most of the other ones are around 4 seconds currently. But my guess is that if OpenAI just smashed it, then the others will follow soon!
The video is crazy. HOWEVER! That Pignose amp back there? I never see those things anymore. God bless!
Haha, good spot!! I LOVE that pignose! After my kids were born, I downsized a lot of my gear, and that little pignose was the only amp I had left. I love it!
Of course, in the years that followed, all my gear ballooned up again. Hah, I've got a stupid 2x12 Fender tube amp-- but, again: I find myself at the point where I'm like: "Why do I have this?"
I'm pretty much all Neural Plugins at this point!
Still, when I don't feel like plugging into a computer, I don't use the stupid big amp-- I use the Pignose! Go figure!
@@TheoreticallyMedia Bro, some of the Neural plugins are so good though. I love the Morgan amps one. I heard in a G3 QnA, Steve Vai say that he was working with Neural - a Steve Vai Archetype would be crazy. I'm hoping at some point they create an Eric Johnson one. Thanks so much for all the videos, man!
The Range Rover footage knocked me the F out. I replayed over and over freeze framing it. I hope Sora will replace Hollywood then you and I could make our blockbuster films. Also, I would love comedians take a stab at and make some outrageous parodies.
Seriously, best driving footage I have seen yet! And yeah, I would LOVE to see what some funny comedians can do with this! Haha, did you ever see Be Kind Rewind?
@@TheoreticallyMedia YES! LOVE that movie. Exactly the kind of stuff I wanna see. I wonder what Gondry would do with this. He has a very creative son, though. Maybe he might get ahold of this thing...
Would be interesting to create a massive ai open world environment, and then be able to walk around in it using your gaming console and controls...
Something i`d like to test is doing an orbit of a character and then passing the video through photogrammetry to see if we can pull 3d character data.
I've seen some people already futzing around with that-- it looks like it's possible! They didn't do characters, but rather some of the drone footage...and yeah, it works. Crazy.
Their AI model understand 3D space accurately right? so it means that maybe in the future it might even be able to be experienced in real time in virtual reality HMD with 3D depth (of another AI model that was trained to split single images to two different angles to stream it in the VR HMD), how crazy is that going to be? to be able to type some text describing a fantastic scene and be able to then right away experience it in full 3D and even be able to move our head/rotate it in its space ETC? the future potential of this tech is so crazy and incredible.
Definitely will replace stock footage but can’t see it replacing it much else
Everything just changes all the time.
That is true! Thumbnail remains accurate!
Crazy weeks, crazy time we live in, CRAZY EVERYTHING, EVERYTHING IS CRAZY.
Everything. Is. Crazy!!
Whats more shocking is the fact that Sora seems to create videos based on understandings of how physics and gravity works in reality BASED ON texts. Also, what Sora shows us is that now AI can actually render our real world in 3D dimension: This is just beyond text-to-video tool; this is rather AI is like now having arms and legs in addition to having eyes (text to image). Imagine how that level of tech can be embedded in automated cars. Now I’m actually beginning to think that Sam Altman asking trillions of dollars is not bullshit at all.
Elections are going te be crazy this year... 😀
I will say that the timing on these advancements couldn't come on a worse year. Given that the Superbowl just normalized AI Video (I mean, jokingly, but still)-- I think we'll see a LOT of in the political ads.
To be honest, I'm not looking forward to it at all...
Does anyone know when it is open for public?
1 minute we are close, if it can extend & maintain characters, narrated videos will expand as for with audio not yet, but with deep fakes this will be amazing. Once it gets to 5 minutes to 10 minutes Oh my
wonder how this will change video games - 4:18 imagine thats what the next Forza could look like. insane
Insane!!!!! Loviiit tx
absolutely amazing progress to ai world!!!!!🤩
I don't know why (yah I do) they release this teaser of Sora and not open it to the public. They are only allowing a select few "Red Teamers" to test this out.
oh, you know why. My guess is to flex on Google's Gemini 1.5. The AI Wars are really kicking into full gear!
Found footage spooky UA-cam content is gonna pop off with this
OMG its happened way faster than. i expected . exciting times
So much faster! Did not expect this until much later in the year!
yes - It looks amazing - the production quality:)
Totally insane. I did not think we'd see this level this early in the year!
Amazing !@@TheoreticallyMedia
my jaw is dropping too!!!
Mind blowing!
Brain is goo. Jaw is on the floor. This is nuts.
Amazing, thank you!
this stuff is going radically shift the market. were the average person is now fully capable a making a movie at home with a small budget.
I'm already thinking of a few scrips for series to make use this stuff. this is a massive break through.
The best one is the octopus and the crab you didn’t show. The coherency of each animal is amazing and each suction cup of the octopus is in line with extremely complex movements
Sora + Apple Vision Pro’s spatial video.
I want that. I really really want that. 🙏
When it will launch!?
"Walking around on horses" :p Poor horses! Bet they don't like people walking on them at all.
Ha! Admittedly my brain wasn’t fully on for this video! Today was supposed to be a day off!
It's interesting that they were able to develop a text-to-video model that could handle scenes with multiple characters before they made a solid text-to-image model that could do that without getting confused.
I’m legitimately stunned by this. We have yet to play with it, obviously- but if this is really what we’re looking at? I mean, this is an indie filmmakers dream come true.
@@TheoreticallyMedia I'm wondering if it might be a glimpse into their priorities. More people are probably more excited about generating their own movies than their own still images.
As for me, I just want to illustrate my books. Unfortunately, my books are action-adventure/mystery novels in the style of the classic pulp magazines of the 20th century, so the censorship policies found on every "corporate" AI program are a constant thorn in my side.
@@TheoreticallyMediawhy is their image model so much less impressive than this I wonder
Content creating gonna change!!!!!
I can only imagine that they are utilizing GPT 4 to aid this new tool. I can’t comprehend the jump without it.
I’d imagine that to be the case as well. Which, might make a heavy amount of catchup for Runway and Pika.
Although, I keep saying the that Pika has been suspiciously quiet lately. Like, I think they’ve got something brewing…
My job will be dead sooner than I think. I thought that a day, 3D software will use AI to generate a frame and manage the shot. It will be a new age for independent film creators!
It's comforting to know that domain "experts" will keep the training wheels on this so we don't hurt ourselves. Also, I can sleep better at night knowing Adobe and the NY Times conspired to add provenance C2PA metadata to the output. It's like having a nanny we didn't know we needed!
You know, I do kind of wonder if you take those outputs and then re-edit them in Premiere or Davinci, doesn’t the metadata get wiped?
From MIT Tech Review "C2PA is secured through cryptography, which relies on a series of codes and keys to protect information from being tampered with and to record where information came from. More specifically, it works by encoding provenance information through a set of hashes that cryptographically bind to each pixel... "
@@PvtGrips-vh7ti About C2PA, I checked now, on OpeanAIs own page they admit that C2PA metadata can be easily removed from Dall-E 3 pictures, by screenshotting or uploading them into social media (which usually removes all metadata from photos) why would videos be any different? OpenAI admits that it's not a silver bullet by any means. "Cyptographically bind to each pixel" doesn't change the fact that someone can just screen capture the video or throw into Premiere, and erase all metadata
@@PvtGrips-vh7ti Also from the MIT Tech Review:
"What’s more, since C2PA relies on creators to opt in, the protocol doesn’t really address the problem of bad actors using AI-generated content."
"Ultimately, the coalition’s most significant challenge may be encouraging widespread adoption across the internet ecosystem, especially by social media platforms. The protocol is designed so that a photo, for example, would have provenance information encoded from the time a camera captured it to when it found its way onto social media. But if the social media platform doesn’t use the protocol, it won’t display the photo’s provenance data."
"The major social media platforms have not yet adopted C2PA."
So, the metadata gets erased if you want to hide it. By simply taking a video capture of the video, or rendering it in Premiere with slight changes. Or uploading your fake video into Instagram or TikTok, which erases all metadata. Even with safeguards, this can be used to mislead people in various ways.
Want want want want want want want want want want want! Thanks Tim, that is literally amazing.
I just offered Sam Altman the 7 trillion he is looking for, if I can get access. I can write him a check in 2 minutes! That check will bounce super hard, but I'll have made a video by then! Worth it!
i wish someone would change youtube headlines forever
Super!
It will be great if Adobe adds something like that to after effects
Any plans for it to get img2video prompting?
Apparently it does! So excited!
@@TheoreticallyMedia Superb! Can we expect an update from you on the matter over time?
I think it will be a while before people accept watching unreal actors and some might never accept it, it will be kind if like buying organic food vs non-organic, just knowing that it isn't real humans might turn some people off but for animation it won't matter. Also famous actors will do everything in front of green screen in one day or just license themselves
Its so impressive, I hope they release it this month, but that seems unlikely.
I think we’re a lot further out than that. I’m thinking summer at the earliest. But I’m also wrong all the time!
Amazing
If you download the video from their site, it is 1080p 👀
I was in a bit of a rush to get the video out, I wanted to see FPS as well. It’s gotta be 24 right?
In the future they'll probably be able to ask it to generate a Drone footage of ancient Rome. And the AI will use its knowledge of all the structures and the art and the architecture to recreate it. It's like we're living in a simulation which we're creating as we go.
I keep thinking of when this gets hooked up to Google Earth. Need a location for a shoot? You’ve can have that exact location at any moment without any travel.
As an aspiring filmmaker and artist this is terrifying. Many people will be fired from the film industry and animation, even tho now you can spot errors, in 8-10 years you can bet it will look decent. Oh, and i forgot the videogame industry too. I mean it's happening now.
As a vet of that industry-- (and side note, I've got some stuff coming up on that next week), I totally understand where you're coming from. That said, I do encourage you to lean into it. While I don't doubt this is going to be an era of bumps and bruises, it will also be a massive opportunity to get a jump start on the future.
This isn't a 1:1 analogy, but think of it like this: It's 1995 and there is no such thing as motion graphics. You have a copy of After Effects 1.0 in your hands. What can you do with it?
this is good, I will be able to do my movie finaly .. jaja jaj .. so much waiting . to this
This is definitely next-level & scary at the same time, in a good way.😂
The person on the far left side at 03:22-23 looses his/her head though!
If you look at a LOT of the outputs, you’ll see small things like that occurring. It is for sure not 100% yet.
Best we’ve seen so far, but still not quite a reality simulator.
@@TheoreticallyMedia Yeah by far the best yet! :) In a week or two we will have even more stunning stuff, I'm sure of it... :)
Finally........ real Video AI has arrived ! I'm finally impressed!! 👏
I was legit thinking about you when I was making this video! I figured this would be the one!
@@TheoreticallyMedia
💜
@@TheoreticallyMedia
I honestly don't understand how the Open AI team has managed to pull so far away from the rest of the entire field in AI Video......... it's like they time warped into the future and brought future technology back into 2024.
Me acknowledging that I'm truly finally impressed (actually blown away) with AI Video technology is no small thing as you know. This is day 1 of no looking back as far as I'm concerned.
@@TheoreticallyMedia
I'm trying to imagine the other very, very hard working people in the field of AI Video watching these SORA demo videos........ how dumbfounded (not meant pejorative at all) they must be to know how hard they probably worked every day on their AI Video technology........ then to see how utterly primitive their modeling or technology is compared to cutting edge SORA.
I actually would feel sympathetically bad for them if they were utterly disgusted with SORA. 🫂
@@TheoreticallyMedia
On a side note, I bet that some of the training for SORA was on some of the highest end games out there. Especially the scene of the truck driving down the dirt road definitely gave me high end game graphics vibes.
But as far as I'm concerned, it works for me.
Okay your thumbnail text is true this time, Tim! 😂
Haha, still gonna get yelled at by someone for sure!
😆 🤣 😂 Yes, I kinda feel the same way and Tim knows I've been a real Debbie Downer & skeptic of the current crop of AI Video tools. But the day I'm truly impressed with AI Video has finally come!
This is most likely the first step into ai generates films. every other model could only generate partially and required labor to actually get a good generation.
This is a BIG step too. Pretty insane...now, when do we get to play with it?!
Whatever technology they used to create this model is a black box and a company secret, it’s so far more advanced than anything we’ve seen before it’s a genuine mystery. For those who are worried about the censorship don’t be, this is a wake up call and motivation for the competitors to try and keep up, even if they can’t, any idea that doesn’t show gore, sex, celebrities, or other intellectual IP is still fair game, that’s not as restricted as you might think.
Did no one notice the half horse in the drone mining town ?
Historically accurate. Life was really hard on horses back then! haha.
I studied for years, I slaved for another 2 years without getting a contract and now I'm unemployed. This is the final blow. fuck this world
Dude never give up, do your own thing, get better at it, beat AI
("write a supportive reply to a youtube comment of a person exasperated at the sora news with no reference to any sunsets or hamsters in adventure gear")
@@notanemoprog thank you for your supportive words, I appreciate the irony to use chatgpt 😀
I've calmed down now, but It's not easy.
Holy moley
We're in the middle of a true paradigm shift. Everything will change. Everything. It's eerily disturbing and deeply fascinating at the same time. What worries me the most, though, is what kind of impact it will have on human cognition in the long run. When we let AI do most of the work - especially in the creative department - we risk not acquiring skills on our own. It's the easy way out, it truly is, and we all know the risks of taking the easy way out. I don't know...I've never been this confused about what I should feel about a certain thing. AI is already having a profound effect on how I view the world. Imagine what it will be like a year from now...hell...even a few months from now!