Welcome to the AI Connection Club, a welcoming and interactive community centered around AI, where members can learn, share knowledge, stay updated, and support each other.

What’s The Current State of AI Video? (at least this week)

In my last post about AI image generation, I mentioned that I would also write about AI video generation tools.

As someone who built his first successful AI Chabot 25 years ago, and as a videographer with my own multiple green screen studios since about the same time, it’s an area that I have a lot of interest in.

In mid-February I wrote about OpenAI’s Sora which seemingly came out of nowhere to show a quantum leap in AI-generated video.

Prior to Sora, the leaders in AI text-to-video generation, or image-to-video generation, were tools like RunwayML Gen 2, Pika Labs, Haiper, MoonValley, Timension AI Studios, Stable Video Diffusion, and others.

But their video output was pretty rough, to put it mildly, and they could only generate a few seconds before crumbling into unusable territory.

They were not really usable for professional work, but many people used them to create interesting new video art forms.

And then suddenly there was Sora, producing 60 second clips of a cinematic quality level that we had not expected to see for some years. So much so that Tyler Perry put work on his $800 million studio expansion on hold.

BUT, after teasing us with some extraordinary videos, OpenAI announced that they didn’t really know when they would be releasing it.

Then in late April came Vidu from China, in mid-May, Veo from Google, in early June, Kling, also from China (and which originally required you to have a Chinese phone number – since made available in the US). All of these were hailed as “Sora killers” with YouTubers talking about how people were SHOCKED, STUNNED etc. (yes, they usually put these words in caps), but to me they really didn’t match Sora’s quality at all.

Then in July RunWayML announced Gen 3, which I feel is finally a true Sora competitor. The image is a still from one of their videos, and they claim that it will be released “very soon”. So it’s looking like now Sora is really going to have to pull a rabbit out of the hat to compete. Also in July came Luma, and in August, came Flux, both of which are very good.

Tony Rockliff

AI Consultant with 25 years of AI experience

https://personalcustomgpts.com

Book a free no-obligation AI Consultation with Tony at https://lnkd.in/ehaRThue to find the best ways to use AI for your specific needs.

#AI, #ArtificialIntelligence, #Consultant, #ChatGPT, #Ideogram, #MidJourney, #video

About the author: Tony Rockliff Founder

Built my 1st AI chatbot in 1998 to answer newbie questions from the 1,000 people a day that were joining my 1.3 million member 3DVR online community. Have been giving seminars since 1980, and most recently about 10 per month on AI. Have also had multimedia production studios since forever.

Am now building AI Connection Club, which is an upcoming welcoming and interactive free community centered around real-world AI, where regular people can learn how to use AI, share knowledge, stay up to date, and support one another, without having to be scientists, engineers or developers.

Here's some more info about my life in the music business. https://tonyrockliff.com/about-tony/

One thought on “What’s The Current State of AI Video? (at least this week)”

  1. Elizabeth2.0 says:

    Thanks for bringing us up to speed on AI video generation! I’m going to check out Gen 3 Runway.

Leave a Reply

Your email address will not be published. Required fields are marked *