Open AI Releases the BEST AI Video Generator BY FAR. Sora Text to Video
Vložit
- čas přidán 14. 02. 2024
- Open AI Just Changed the World. Again.
▼ Link(s) From Today’s Video:
Sora: openai.com/sora
► MattVidPro Discord: / discord
► Follow Me on Twitter: / mattvidpro
-------------------------------------------------
▼ Extra Links of Interest:
✩ AI LINKS MASTER LIST: www.futurepedia.io/
✩ General AI Playlist: • General MattVidPro AI ...
✩ AI I use to edit videos: www.descript.com/?lmref=nA4fDg
✩ Instagram: mattvidpro
✩ Tiktok: tiktok.com/@mattvidpro
✩ Second Channel: / @matt_pie
-------------------------------------------------
Thanks for watching Matt Video Productions! I make all sorts of videos here on CZcams! Technology, Tutorials, and Reviews! Enjoy Your stay here, and subscribe!
All Suggestions, Thoughts And Comments Are Greatly Appreciated… Because I Actually Read Them.
-------------------------------------------------
► Business Contact: MattVidProSecond@gmail.com - Věda a technologie
BTW I literally clicked record seconds after I saw this release. This is my genuine reaction to world changing technology!
Looking forward to watching this
More news like this and the Earth will split in half
Yeah man this is wild wild wild!
The 24 year old eyeball on the website 😮
Mind blown 🤯
With these types of announcements, you realize that OpenAI isn't just sitting around doing nothing. They surely have several projects that haven't been released yet
Wait until we see the robots.
Sure, but I’m thinking about GPT-5 not an entire AI video model! And one that is THIS GOOD? Open AI never used to move at this pace.
If they could use 1% of their human ressources to not have a shameful web app and mobile app 😅
And they've got tons of GPU servers too
I think gpt-5 in it's multi modality requires alot of high quality specialty models so if the training model is this good imagine how much better it'll be as a multi model model with this model having trained it in part as a collection of experts @MattVidPro
Google: introducing 1 million context window.
Openai: hold my 🍺
Hold on to your 📃 fellow 🎓
@@typicalglitchers Are you seen what i'm seen?
This seems far more interesting
@@joec136gpt 5 will have an infinite context window
@@creed4788 That would be cool. An infinite or extremely long context window is something AGI will need because it will have to take into account years of new information.
What's so insane is that even the "Bad" generations they show still look photorealistic...like if I saw a wolf pup multiplying from thin air, that's definitely what it would look like lol exciting stuff!
It still struggles with the logic and physics, but the visual realism is just about there.
Imagine showing those puppies to medieval people.
Now this actually SHOCKED THE INDUSTRY 😂
Hahahahahaha 😂😂😂😂😂
I think you also mean this CRAZY new SECRET technology STUNNED and surprised EVERYONE! lol. I'm so tired of seeing those clickbait titles all the time. I need to remember to click "don't recommend channel."
what clickbait? his title says everything what he described in the video.
it was actually shocking for once!!!! OMG
@@RainingMan1001 If you're asking me, I'm not talking about this video, but ones that put SHOCKED and STUNS in the title of every video they make. Matt is great, as is the other Matt. lol.
This CANNOT be real. If it is then i have never been this impressed by anything before in my entire life
It has to be really cherry picked lmao, no way they casually have a model that can make videos like this, even if it is cherry picked, still hard to believe
@@zrakonthekrakon494cherrypicked or not, it’s still AI creating video.
I was the most impressed when I first saw DALL-E. This is just the natural evolution of that. Maybe just a little sooner than we expected though.
And the videos are seconds long imagine generated hole 30 minutes video would take forever you’ll think.
@thekrakon494 Let's assume you're right (and you probably are) -- that means that what we're seeing here represents "the worst this is ever going to be."
As someone who has lived through the advent of FMV on PCs, I'm as blown away by this as I was the first time I saw a home computer playback full motion video with sound. It was totally, completely, utterly mind boggling.
My point is that in just 30 years, the concept of watching FMV isn't even a novelty anymore; it's the norm. We watch 4K full length movies transmitted wirelessly to our home televisions over the internet. In 1993, a computer couldn't playback more than 30 seconds of video at 240p, and even then it had to source the video from a CD-ROM.
That said, as mind blowing as Sora AI seems now, just remember that in as little as 2 to 5 years, what you're seeing here now will seem primitive.
I'm shocked like an 11th century man watching television
Same. In fact I would sayeth these exact words, " What Sorcery is this, Soothsayer??!!"
The consistency of this is just LEAPS AND BOUNDS above any other AI video generators. That's actually insane...
Timestamps:
0:00 Intro
1:12 Snowy Tokyo city clip
2:08 Openai talks about safety
2:26 Wooly mammoths clip
3:33 Cinematic movie trailer clip
5:14 Papercraft coral reef clip
5:40 Animated fluffy monster with candle clip
6:41 Women in Tokyo clip
8:40 Community reactions
8:54 OpenAI Sora blog post
9:41 Coastline clip
10:06 More clips
10:44 Red team access
11:17 Even more clips
16:37 Weaknesses
17:35 More about safety
18:15 Even more clips
19:14 Research techniques
20:57 Matt's final thoughts
If I were a movie studio or animation studio, I’d be freaking out right now
Im an EX VFX artist, I moved on as soon as Midjourney started getting good.
Animation studios definitely. Movie studios I think can survive a bit longer since even though these videos look photorealistic, there's no sound + I doubt these AI actors can have good acting abilities (yet). It'll be awhile until they can generate sound + good acting abilities. For animation, the visuals are over - all you need is someone skilled in sound design + maybe a good writer then you can make a quality animation film.
Just give it one more year.
@@Yic17Gaming
@@Yic17Gaming 1-I think the dialogue will be the hardest to nail down, and 2- I just took a few clips from Sora and added sounds to them on my channel. The environment sounds are pretty easy to create
@@MicahYeah Yeah, even if you add voice - the lipsync will be an issue. We'll have to wait for a plug-in that adds lipsync to the audio.
This is only in less than one year after video generators existed, imagine in two years from now 🥶🥶
It will Freeze my balls off
a couple years not one year, but yes, the improvement is rapid
Imagine 10
the wild thing about this is, even when it messes up it doesn't look bad it looks trippy AF
Prompt: Pika executives trembling in their seats, brainstorming on a new moat to dig
😂😂😂
I'm sure they will be involved.
😂😂
Pika doesn't have executives. It's a small team.
To be fair you can use Pika today but Sora is locked for now.
I thought the title was clickbait, and wow that is insane. Everyone thought OpenAI was falling behind but this shows they are still far ahead of anything anyone else is doing.
I was telling my friends nearly 7 months ago that within a year to two years they're going to make it so that you can jump on a service like Netflix, and just tell it what kind of movie you want to see and it'll walk you through a series of basic questions to help guide it in generating that movie or TV episode. And then it'll just generate it for you. This is absolutely astounding to see that kind of temporal an image stability.
And people are worried about studios using AI, we are just going to use it at home.
Isn't it perhaps too early for that? The ability to script a movie and generate coordinated, consistent video and audio for hours seems like something AGI caliber.
@@Joaquin25333 I don't know if that's going to be too far behind. You're right it is a larger project to create a movie but honestly most of it can be scripted and you would just have chat GPT right the storyline and script and tie it into audio generation, for music and voices, and then something like this to just start developing the scenes. It would utilize something like what's in stable diffusion where it's able to use reference material for style and character consistency.
It won't be Netflix, it'll be some startup in a garage renting out server space in exchange for subscription fees.
In such a scenario why would you need Netflix? Just make your own content or watch other peoples creations on an open platform.
As a videographer I feel like selling my lenses and camera and just buying NVIDIA Cards 😞
No don’t sell your lenses and camera but YES buy yourself NVIDIA cards.
I just bought more lenses for photography. Look, passion is passion, regardless of the technology. If videography is specifically what you’re passionate about then follow that.
@@MattVidProTech like this tends to cause you to lose your passion though 😢
Its over
If AI can replace you, you're not being authentic enough. Your authenticity is what makes your product unique, what makes your product creative and different from others.
There will always be an audience for phenomenal works. You have to think outside the box, do something AI cant.
Oh man, in the future, we could literally create our own movie just by giving prompts
18-24 months
I give it 5-10 years before we can generate AAA video games at home.
Yes, but while also putting millions of artists out of work.
@@mirandataylor6385Are you an artist?
But so much censorship
I love how they talk about "safety" when we're all really just waiting for a version with the safety turned off.
Exactly…
It can't be stopped; it's coming eventually. It seems a far way off now, but in 10 years we will look back and see how inevitable it was
Little boys are just not ready.
@@xAgentVFXAyo???
@@starmorpheus custom porn. Not good
Wouldn't be surprised to learn that this new model was imagined and coded by unrestricted GPT5.5
I share the utter disbelief regarding how big of a jump in quality this is.
If this model really does work as well as shown at this point then 2024 might be the last year we'll be able to tell if any type of media has been AI generated or not.
kind of scary not knowing what the truth is.
Indeed. I am a bit conflicted right now, because I desperately want to use it myself, but I'm not sure if they really should put it out before the US presidential elections...
Why
I wonder, what if they are using GPT-5 to help them do these mindblowing stuff?
The crazy thing is how the surroundings are reflected from her sunglasses. In the closeup is possible to see the street that was in the picture the shot before. Somehow this model is completely aware of the scene and the elements. This is also a big leap towards AGI
The example videos are shockingly good. The days of $200 million movies are nearly over. This is the first nail in the coffin.
I'm worried that we'll be flooded with a load of slop, actual filmmakers will stil most likely still dominate the field due to skill alone.
@ecognitio9605 we are already loaded with crap. It's going to be 100x worse with AI
@@ecognitio9605 a great joke you don't even know the creative possibilities of billions of people.
AI movies will need a coherent style to them, one of the important things AI still has ought to iron out, is *consistency,* and then potentially... maybe artists will be able to properly express themselves without it all looking like AI slop. i'm a hard believer that this *can* work with the right tools and given the proper time. but as how things are looking right now, i'm not too worried about the videos or short films. we already have plenty to worry about AI images, and it's been one hell of a ride...
@@AltKaxREAL Laziness will dominate over human ingenuity, hardwork and artistic mastery.
other companies releasing AI video gen tech:
OpenAI: lol
Somehow even though I've already seen image generation AI go from garbage to photo-realism in such a short space of time it's still mind-blowing seeing something so good coming out for video already. This is going to change video creation forever. I've always (different account to this one) done simple videos sitting in front of a camera without any B-roll because I don't really have the time and energy to get a bunch of it for every video. This could change that and allow me to make interesting well edited videos without having to shell out for stock footage or record my own.
Edit: I actually had to double check that it wasn't April already because some of these clips are so good they had me legitimately questioning if this was a troll.
So agree with your edit
I wonder which string of code they moved to make it look better haha
The woman walking in Tokyo street and the wooly mammoths are simply mind-blowing.
6:20 pretty nuts the candle even illuminates its hand when it gets closer
That’s a crazy level of detail I easily missed.. that’s insane
That Ray tracing is now AI generated too
Remember when 2 years ago you were blown away by still images? Now there's this... like wtf. Where will we be in another 2-4 years?! Scary!
Actually, Chat GPT 3 and all the image generation stuff was less than two years ago. It was about 16 months ago.
@@ChainedFei Image generation stuff was like 3-4 years ago.
This is a long way from Will Smith eating pasta 😎
The pasta test for this model is still undone!
I just shouted in shock at the close up of the japanese woman in sunglasses. Absolutely wild
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 10 years you can bet it will look decent. Oh, and i forgot the videogame industry too. I mean it's happening now.
Next year, more likely.
you'll be able to create so much more with much more ease!
I'm just waiting for someone to train a LORA on mesh designing.
@@nowayhose-mi2on I dislike corporate studios too, never worked for them, but the problem is that A.I. companies are stealing portfolio data from artists without their consent. Also imagine a future where all entertainment will be generated by all the people with a phone. No more actors, directors, animators, composers. The skill will drop massively. The output of all of this will be just pure spam. Nobody will be able to sell it because it will be free.
As a plumber, this is all very interesting 🤔
Remember just how huge the anti-AI panic was on Twitter with the release of DALLE-3? Then imagine how big it will be with this one
And wait until AI achieves sentience!!!!
@@AmazingArendsThat will never happen. A.I. only does what its told, whether that's deciding the "best opinion to give" based on the majority, or being coded to give lesser common opinions. Everything is regulated and it has no mind of it's own.
@@freestylegamingartist8192 So you're not one of the "AGI is right around the corner" people. Fair enough. 😊
I was expecting that things like this would be possible maybe 5 years from now.
Open AI is 5 years ahead
This is how exponential progress works … think months nor years … and soon, weeks then days
This one actually deserves a SHOCKING title.
A Click bait title cannot do it justice :)
I love how OpenAI says "if" they decide to incorporate the technology into a model for the public. OpenAI, if you don't ever decide to release a video model to the public, open source will do it eventually and you will be left behind in the dust. The technology has been unveiled and it's only a matter of time before the public has access to it.
It's only a matter of when, not "if" "Stable Video" jumps to this quality.
They surely can’t release this to the public. Imagine sexual or even gory contents to the extreme are massively generated. We are doomed if they don’t have 100% control on this monster.
I strongly disagree
What I wonder is how far behind the open-source video gen systems will lag Sora. You know, the ones that have no content filtering whatsoever? I think open source is 6 to 9 months behind the best Midjourney and Dall-E 3 content, in the still image space. Maybe open source will lag further in the video arena, more like 2-3 years? If it's 3 years, then by 2028 maybe there'll be open source video gen that has absolutely no content control and beats even what we saw today from Sora...
@@niveketihw1897 Even if it's five years away, it's coming soon. There will be no way to stop an open source model that produces unfiltered, flawless video. It's coming, and soon. People are going to create horrible stuff, and there will be no way to stop them; there will just be laws to prevent its public distribution
I love how new AI video generators keep getting more consistent. Thank you for sharing!
So often, as CZcamsrs, we say things like "This changes everything." Usually, it doesn't. But this time. This is different. What we are seeing here is the beginning of a new era in filmmaking. The things we will be creating will truly be shocking. OpenAI just hit it out of the park. 💯. Thanks for jumping on this so quickly, Matt.
Not only that. What we see here is the beginning of AI really understanding the physical world we are living in. This has huge implications for every aspect of our lives, not only movies. But as someone with ambitions in movie making, I am beyond hyped for that aspect.
@@missoats8731 We are truly seeing the future unfold before our eyes. Exciting times to be alive.
The AI movie industry might have the same process as the actual one. First, we'll have silent films, then we'll have talkies, and then the special effects will get better and better until we have stunning full-length feature films entirely generated by AI. As a writer, that makes me very excited because I won't have to worry about finding the right actors, making changes because of not being able to find perfect people, creative differences, and Hollywood racebending all of my characters.
No one will care about your characters when they can be generated effortlessly
Stop being so enthusiastic until you've actually made a lemon like character wearing sunglasses
The lemon should try eating spaghetti on the beach
wow, it's shocking. I could not expect this quality this year, maybe in 2-3. The next step is making a screenplay with sketches, camera movements, description of scenes, and even feeding it with pictures of characters, scenes, concept arts, etc, the style. and bumm, you have a movie.
Well, you'd have a silent movie 😆
@@missoats8731 haha, you are f... wrong. Search for Lipsync :)
Actually over for us 3d artist
Ikr, that surfing scene would take at least a week to create with years of knowledge and expensive build to simulate and render. Even the fur is simulated, not just the ocean.🤯
Not really, original content will always have a demand, eventually to train and fine tune AIs that can further work with the models, human labor will still be needed to refine and adjust when necessary, but the demand will def die down A lot
@@hinter9907 stop coping, just look at the growth rate, it's truly over.....
AI won't be able to know the idea you have in your head. You'll be able to ask an AI to give you a model, then you can use your skills to fix it to match what you had in your head.
No need to do annoying shit like topology, rigging, or getting the hair to look right. Just have the AI do that part. Then throw that model into video generator like this to ensure it gets the right creature, otherwise, it might not get it how you wanted by text alone.
You'll be ahead of the curve and able to properly fix problems, as well as have better ideas, with your skills.
Video games will be amazing in the future.
I'm literally crying right now, this is so beautiful, the day we all were waiting for has arrived.
We're almost there
As someone in the industry, I am absolutely terrified.
Good, 5 years down the row and people regardless of their political background will be able to turn their views thoughts and stories into a 3 hour long movie for everybody to see in a matter of days, unlike right now where woke liberal lobby has the hegemony over entertainment part of the popular culture
I sh*t you not, I literally got chills when I saw the sheer power that this model is capable of
Hats off to OpenAI 👏👏👏👏 the future truly is now!
Haha yeah just consume whatever product is put in front of you! Buy the Apple vision pro because you were told to!
@@KayakoProfileapple vison for bozos, of you love your eyes stay away from that sh!t
I'm really interested to see if this will develop into something that you could give it a book, and it'll create a whole movie. Doesn't seem that far fetched now.
The "safety" policies will prevent that from ever happening.
I been telling that to my children is going to happen for many years now, Not far away now my guess before the end of 2025.
@@vainezaiven6677hope that doesn't go on into the later half of the AI era...doubt it, though. Maybe the open source stuff?
@@dniilii Yeah, open source will be our best bet for "safety"-free programs. They'll just take longer to come out (and they'll be less user-friendly when they do).
CZcamss servers are so fucked when this goes public lmao
The raytraced light on the monster eyes is mind blowing honestly
I guess we now know how Nvidia is gonna improve their Ray tracing and path tracing tech in the future
AI EVOLUTION SPEED IS MINDBLOWING 🤯🤯Thankyou matt for conveying all news in so detail!!
it was just a few weeks ago that google presented its video generation AI which was a slap in the face as the progress was crazy and you now have open ai which gives you a second slap in the face with its new AI
WTF ?! when I first saw it my mind got blown up so much, I was in total shock and awe at what I was seeing, I couldnt believe my eyes, I thought it's not really AI text to video, I honestly expected this level of coherency and realism to exist with text to video AI models only maybe in like 3-4 years and it's already in existence, what the ? are we having a mini - singularity of the AI field or what? what's going on? mother of god, this is way and beyond too good to exist already.
Welcome to the singularity
Exactly my response
@@itsdakideli755 this is not even AGI ley alone ASI and the future, What a time to be alive.
@@itsdakideli755 :) Now we should imagine what the real singularity would be like in the far future if it's already feeling like it.
I believe we are currently in the "human-assisted singularity" phase. I do not doubt that AI is being utilized to improve present and future AI systems.
It's fantastic to see you being so authentic, mezmerized and shocked/happy at the same time! I love your work! Thank you!
From the first second clip of your video and smile on it, without audio and without seeing a title, I knew there was something big coming!
I'm so happy Matt rushed out a video lol. I've been freaking out at the sight of this. Reminds me of the feeling of wonder I encountered when I first used GPT-3, back when it was new.
I love the idea that I could make my own Snyder cut versions of movies. I'm gonna do some weird shit with this I promise :P
This is amazing - thanks for posting -- Ai video is going to be such a massive game-changer for so many reasons
This is why mattvidpro is my favourite YT Channel in the AI space.
You really know how to capture the feeling and express it in the most natural way.
Love you man.
Also, love OpenAI.
Been with you since the early dalle-3 days.
Thanks for making this video so that I didn't have to watch a video with an annoying clickbait thumbnail.
Hooray for MattVidProAI!
Finally, no more flickering images in AI video. Can't wait for this to go public. I very much could use it for a visual novel I'm creating.
Up until last night the plan was to head back to Australia probably late next year, put a small crew and cast together, and use various tools including AI to shoot one of the feature films I've been writing over the last two decades.
This release has now moved that timeframe up by about a year.
See you in like, November, Melbourne.
As a 3D artist, it's so over :D
O yea
? Become content creator lol, believe me alot of people gonna remain consumers than creators doesn't matter how easy it becomes
Wow, just wow. this model is absolutely insane. Can't wait to start using it.
This is more proof that open ai is still ahead of the game
any doubt? still no one reached gpt 4
@@sherpyaGemini pro 1.5
No Doubt
@@sherpya gemini/bard is better now
@@deaddycruel not sure just faster
I can't wait for an updated version of the AI wizards messing up a fast food restaurant, but made with Sora lol
Wizardposting is here to stay
I guess I'm going to quit studying animation 😭
Yeah. Like, literally. That job is dead.
@@Post-Truth_Cephalopodlike pretty much every single job out there? Atleast physical jobs are safe for now
@@gelopisaminasotaripitanasi5921 Most jobs, for sure. Though I believe a lot of physical jobs will be killed quite a bit sooner than we think. That said, there are some "creative" jobs that will hold out for quite a while, like theoretical physics, literary writing, the less commercial side of civil planning, etc.
At least you have a cool hobby when we are all out of a job
"thats how it always starts with the oooo's aaaaahhhh's but then theres the running. And screaming....."
Well, thanks for solving the "what am I watching tonight" conundrum.
Love this comment 😂😂
😂😂😂😂☠️☠️
@@Zugzug2011 oh you and me both my man lol. Excellent choice
This is INSANITY! 😮🎉 I was not expecting this at all! 2024 is going to be insane. Who knows what the world will be like 10 years from now!
Holy crap! I literally had to check if it was April 1st and thought OpenAI was pranking us. This isn’t just an improvement; it’s several generational leaps ahead. Unexpected, unprecedented and unbelievable! Like you said, if OpenAI casually has this kind of tech behind the scenes, what other mind-blowing realities await us in the very near future?
5:56 Did you notice that the monster has a tiny candle embedded in their hand?
Total character consistency and physical world logic will be standard soon thanks to Open AI.
They've 100% achieved AGI internally and they're breadcrumbing us
While trying to engineer/lobotomize it so that it serves the governments purposes.
I'm sure the super secreqt multimodal models they have would actually stop the planet for a day if leaked....
They are definitely ahead of everyone
actually was one of my thoughts on how they advanced THIS fast, they were informed on a better process
Agi won't be a thing in any time soon
Completely MENTAL. Just saw there whole demo page, and DAMN. I though this quality would only arrive in a few years and here we are already... Incredible, TRULY.
The consistency is crazy. Open Ai is going to start taking on the movie industry shortly with these advancements.
The amount of safety guardrails that will be needed to be put on this so it doesn't produce anything graphic will be insane
They'll tell you that's the kind of safety they are working on, of course, but what should concern us most about closed-source AI is that "red team" == left-wing activists, working to restrict a lot more than just mature content.
@@jg_ultraThis right here. They want to create a god to make their unworkable, irrational reality into a reality. Woe be to everyone.
The scary thing is that if something like this becomes open source, you’re completely screwed.
Eventually there will be uncensored video generations of this quality so why worry about it so much
@@joec136 My greater concern is the gaslighting that comes from describing so many of these models as open source. Open source models, provided without all input data and parameters, are still a black box, still guided by the red team (the name of which is perhaps bold but definitely unsurprising, given how many of them are unabashed commies.) There are a great many people who will see/hear/read “open source model” and think “Yay! I can trust this to be unbiased, because open source!” when nothing could be further from the truth.
People have been laughing at me when I been saying that soon you can ask AI to make you TV series and movies based on any book or story.. and you can add yourself as the main character etc.. maybe view it through the eyes of the main character.. maybe add some xxx to it.. or make partly a game.. .. I expected to see something this good by the end of this year! it is mid February!
bro… i just don’t even have words for this… it’s just shocking… imagine where we’ll be in 6 months… 😂
2hrs ago on release I was super excited! Now my stomachs upset and got that weird feeling of walking down an unknown alley at night lost in a foreign big city 😮💨 Open Ai definitely has Agi in the sandbox training.
I just hope that the same that occurred with Dalle-2 that few months later appeared Stable Diffusion to challenged it (with great success), now happens the same with some open source proposal to disrupt the AI video generation scene.
best part is in a few months we will get another huge advancement like this at this rate
Months? 😂
Simultaneously it will replace your teacher, doctor and advocate. 👍♥️
I really like your enthusiastic, genuine reaction 🙂
My jaw is dropping too. Rushed to CZcams to see who’s posted
I just went to their web site for Sora
Watching the video examples on the OpenAI website honestly made me cry.
That second space guy is beautiful!
That really is impressive. First AI thing in a while that has really felt like a genuine level up.
Altman is playing chess, the next Steve Jobs for sure.
The 3d kangaroo was really impressive and scary. As someone still trying to do 3d, I'm definitely feeling a 'dey terk er jerbs' moment, but I'm happy it's happening! Especially since this is just the beginning! It's only going to get more impressive in a short time.
That was fast, Matt.
I started recording 4 min after launch 🤣
Your reactions are worth the price of admission alone!!! LOVE this! I can't WAIT to try this out. Thanks for sharing.
Just a reminder: when OpenAI says "safety," our ears should hear "censorship."
It'll be safe in the same way as DE3, which is to say they'll make a game out of their content regulation system that forces people to get more creative in their prompting.
Well I guess it's necessary or it'll just crash and burn sooner rather than later, but I'll take it. I'll be able to use this to make insane music videos
No, there's definitely safety concerns.
Listen 'haha we all want to make porn' or have a character in a movie get shot in the head etc, but there is legitimate, society changing concerns here. This isn't a fear-mongering remark.
You'll get what you want eventually, but be glad OpenAI takes careful approaches.
@@IcyLucario I do agree with this.
But I also expect that OpenAI will take the "safety" guidelines way too far.
I can imagine a future in which chatgpt 5 could give us video tutorials about absolutely anything .
I love your enthusiasm!!!! I reacted the same way, and people I know just don't get it. Thank's for the video!!❤
It really looks like a lucid dreaming sequence. We might be closer to the concept of consciousness with this models than we thought...
Ok, this ended the debate about aliens. OpenAi has alien technology.
AI will BECOME the "alien intelligence!"
Amazing video Matt!! This is super exciting, thank you for sharing this with us
I kept on checking your channel for hours waiting for this video to see your reaction haha ( commenting before watching your reaction)
Text to Video game, while gaming ... with AI NPC and everything, if you have seen an ISEKAI, you could play something like this but it wouldn't be a novel or manga/anime, it will be the perfect game, it will happen, it will be incredible, it's already almost possible
So much better than last year. Just imagine next year
The understanding of cinematic aesthetics (camera movement, framing, slow motion, depth of field, film grain, lighting etc.) Sora possesses is mind-blowing. To use this right now in a professional context I suspect one will have to steer away from its weaknesses and try multiple generations to really nail a shot but it's very rare these days for any shot in a film to last a full 60 seconds. Obviously things like lip-syncing and consistent characters across multiple prompts will be needed to get it to the point where you could make a traditional feature film but AI models have so much footage available of humans speaking that lip-syncing should be something it masters quickly. Can't believe we got here this fast.
I imagine a future where you can give an AI an idea for a film, and it will create the entire video, script, character voices, and even the music. Once it's done, you can share it online. Exciting times ahead!
Yeah this future you speak of seems to be less than five years away at this rate.
OMG! This is so, FREAKING AMAZING! Great work!
I can’t wait!
Can you make an ai of yourself, like Dreambooth and Stable Diffusion?
Matt is on the case! coming in with the news 4 minutes after announcement. Thanks, Matt!