The World of AI Generative Video
The world of AI generative video just keeps getting crazier and crazier with new announcements coming out all the time. Really cool research and more flexibility and controllability over the videos we create. It’s crazy because some of the predictions that I made in this 2024 predictions video have already come true.
2024 Predictions
There’s one prediction I made that I couldn’t have been further off on. I don’t think in 2024 we’re going to have a tool where I type in a prompt and it spits out a 30-minute episode of South Park for me. But I do think we’re going to see much higher quality AI video along with longer generations in that higher quality.
Recent Developments
On January 23rd, just three weeks after my predictions video, I completely changed my stance and course corrected. I made a comment that I don’t think AI is going to be creating full-length shows and films this year. However, I was wrong. The next two months have been wild with mind-blowing AI video advancements.
Sora and P Labs
Starting with Sora, the realism and physics of this tool have impressed many. P Labs also introduced a new feature called lipsync, allowing AI-generated videos to talk based on audio files uploaded. These advancements have taken AI video generation to a whole new level.
Emo by Alibaba Group
Alibaba Group’s Emo technology is another groundbreaking development. It can generate expressive portrait videos with audio, capturing emotions and movements in an incredibly realistic way. The ability to convey emotions through AI-generated videos is truly impressive.
LTX Studio and Anamorph
Companies like LTX Studio and Anamorph have also made significant strides in AI video generation. LTX Studio can create entire storyboards and films from prompts, offering a comprehensive video creation platform. Anamorph’s technology can reorder scenes to create unlimited versions of one film, providing a unique viewing experience for each audience.
In conclusion, the world of AI generative video is evolving rapidly, with new tools and capabilities emerging constantly. The future of AI video generation looks promising, and we can expect even more innovative and creative developments in the coming years.
There are some instagram accounts that serve perfectly generated 30-second single shot clip video in the form of reels (@feed.carribeanmodels , @feed.modelz , etc , etc) .. so yeah .. imagine having 30 seconds for each shot in a generated movie instead of the 2,5-5 seconds that most people are working on piecing together. I wouldn't be surprised if by the end of 2024 we will see the same kind of numbers on social media for AI video as we are currently seeing for AI stills.. And using embeddings and lora's in video production as LTX .. we'll that's a nobrainer isn't it. I guess the next thing is a tool where you generate your main characters and then generate videos with those characters .. everything could be done today without openai's insane money and hardware .. so it's going to be a hell of a year for AI video 😀
AI is going to change the world, but it will never replace human brains and creativity.
I said it was going to be "game over" sooner than later when it came to AI video… I was right but I didn't expect it was going to be THIS soon.
Did you miss the Nvidia event or did I miss your video about that event? I think what Nvidia revealed would accelerate everything like crazy.
The EMO project is interesting, but I feel it would be helped if there were a transcript to help identify the words being spoken for lip-sync, I see in many instances, the final consonants are dropped from the lip-sync because of poor recordings, or ambiguous sounds. Also when hearing SH, F, TH and other fricatives, for example, our minds fill in the expected sound based on context, while the lip-sync often gets it wrong.
I'm really into the effects you use. It's cool to see how you use AI to make your videos better, and the quality keeps getting better with the tech.
7:53 I could see a company trying out 100+ slight variations of AI generated commercial just to probe customer bases (making even more positions “redundant”, but I think we’re way past that debate)
OMFG!! That EMO project is absolutely NUTS! Soooo impressive. Jaw dropping! I can only imagine what is not in public domain
If you want to be an independent, neutral journalist, you shouldn't collaborate financially with AI companies.
Do songs not affect my videos because I'm not monetized in the first place? I WAS monetized before terrorists and the Paul brothers ruined it for everyone, and game music had that effect on my videos (which is stupid if I can make money from the games themselves). But I never actually made any money regardless other than the few cents I made from being with the Freedom network. But they didn't give a fuck about me, so I got no attention.
We are sooooo f*cked!
Nice video. I'm thinking to watch it again, and follow if there's any difference. 🙂
Can you use it to remake "The Last Jedi" please?
Please don't use these clickbait, sad/shocked/crazy faced thumbnails. You're better than this.
My predictions: AGI before summer of this year. Up to 5 min video in 3-4 months.
Reduction of the coding workforce by 30 percent by the end of the year. ChatGPT 3.5 level AI without internet on cell phones this year.
Disappointed that you're going to work for an Israeli startup Lightricks, while a brave soul like Aaron sacrificed his life for Palestine. Don't lose your humanity.
industry SHOCKED!
Hey Matt, just FYI, I create the 1st feature AI movie last year and got it into actual movie filmfest not AI film festivals. We don't put it on YouTube because of some of the restrictions for some film festivals and I didn't get permission from Stars Wars, my movie is called Dj Gregu it's over a hour long all AI footage, and music. Last year another animation my team did the old-school way got 19 film awards. So it's good to see folks catching up. Hit me if you want to know more
Good on you for admitting when you get it wrong!
AGI no later than the end of 2025.
4:06
its kindda funny she has ears of different colors.
Please talk about shinefy!
same shirt from your prediction video? it's like you went back into the future!
It’s just to difficult to keep up with the latest best practices of AI. If you’re a creative writer, focus on your story. People will still want to watch a good storyline.
Wow, my messages have typos in it. I didn't write it like that at all lol.
WOOOHOO , so proud to have taken part in Our T2 remake 50 artists or awesomeness
Naw it's all going to suck with good writers
I still find it so ironic how fervently the music copyright is protected while the copyright for the art used for learning models isn't.
Hmmmm hmmmm hmmmm crazy hands is signalling for ya
LTX Studio is another waitlist boys and girls so get over there and get your spoke in!
I just did my request, let's see how it goes. Oh and Matt, your message that I got disappeared. I was just looking for it but don't have telegram as of now. Then you for your message.
Creating a short stock video is sooo farrr from a movies, that people believing this are insane..
Even with all the text content gpt can’t even write a book, nog even anything more then a generic t very ai like blog post…
Yes it’s crazy times, but these hype kiddos are pissing me off lol 😝
Wow, great update – it is bold to make predictions in this environment but I appreciate that you do it 🙂
Thanks for this 😊
In the mid-90's I ran some "graphics" classes at the local "Film Institute" for creative types wanting to learn about what they might be able to get out of a computer for TV or Film… for a handful of students I had to show them how to handle the "mouse" first. 30 years later…
Yeah, lets remake the ending of Game of Thrones