Sora 2's Remix Feature Will Change How We Create

OpenAI's AI video model has opened the door to the promise of interactive creativity and everything looks different from here.

Today on AI For Humans:
Sora 2’s Remix Might Make You Creative
Google NanoBanana 2 Maybe Leaks
Plus, Grok Imagine Video Is Good Now

Love this newsletter? Forward it to one curious friend. They can join in one click.

Welcome back to the AI For Humans newsletter!

So I went semi-viral on the Sora app this week but it wasn’t for an original creation (you can argue how original any Sora post is somewhere else).

It was for a ‘remix’.

When this video remix where I chose to put a giant Queen Elizabeth in a block of ice (sorry) ended up getting over 1600 likes, I started to think a lot more about the power of this tool within Sora and the implications for content creation at large.

What Exactly Is a Sora Remix?

A remix on Sora is where you take someone else’s video and prompt the system to change one aspect of the original to create something new but directly connected.

It sometimes means changing the subject of a video or the words the characters say or the setting. Basically, the remix-er adds a twist on the original video.

But what makes it really interesting is that it gives social video an entirely separate axis to explore. I mean this quite literally. Instead of swiping up, you can choose to swipe right to explore the remixes of the video you land on.

This video is a good example. The original was incredibly creative and weird (two fencers using swordfish instead of foils) but if you swipe right you’ll see a number of variations like fencing with fences, fencing with Dyson vaccums, and even swordfish fencing with fences.

It’s like a bunch of parallel universes…

This means you can go deeper into one specific idea and, as you swipe through a series of remixes, you start to feel the beginning of a mini-universe being formed.

Suddenly, there are a TON of stories to tell from a single starting point.

And, even better, you can easily add your own piece of lore.

When Creation Becomes Co-Creation…

YouTube, Tiktok & Reels are mostly one-way consumption & creation video tools.

To consume, you watch a video and maybe follow the creator and then swipe or click onto the next one.

To create a video, you record a video and upload it and hope someone will watch it.

Tiktok has a few tools designed to get the user to reply via video (stitch lets you use the first few seconds of someone else’s video for example) but generally this one-way path has become the accepted way social video works.

Sora completely changes this. Every video becomes a two-way street.

Actually, it’s more like a multi-way street that can go in any direction you want.

The Remix feature lets you, the user, jump off of someone else’s brilliant idea and immediately add your own spin.

You don’t have to find the right lighting or your microphone or a series of pre-made images, you just remix that video you found funny and now you’re part of the story.

Even better, other people can see and engage on your remixed video and follow you as a creator as well.

If you read through the presentation I dropped in here a few weeks ago, you might’ve seen the following slide:

Sora 2 is a game…

Sora’s remix feature is not just a content tool… it’s a game.

It’s an entirely new interactive way to engage with the media you consume and then co-create your own.

So Why Does That Matter?

We’ve been trained for decades that internet content would be supplied by a very small number of creators and that the vast majority of people would just lean back and consume.

Even in the YouTube era, which opened the door for the entire Creator economy to blossom, there were still a fairly limited number of humans willing to put themselves in front of the camera and ask people to ‘click that subscribe button’.

Tools like Sora’s remix feature are starting to shift that balance significantly.

Because AI video is so much easier to create than traditional video, the act of creation is much easier as well. And when you, the user, create something that not only tickles you but tickles others, you start wanting to create more.

To me, this is 100% a great thing for humanity.

Of course, not all of this will be good. In fact, the vast majority of it might be bad. But, dear reader, I’m here to tell you that a lot of it is already kind of good.

I suggest you spend some time with Sora this week and instead of just swiping up, find a video you enjoy that has a fair number of remixes and swipe right.

You might just lose yourself in this weird 1980s newsroom and then add your own dumb thing for them to do or say. And before you know it, you’ve written their backstories in your head and you’re off and running.

See you next week!


- Gavin

In this week’s AI For Humans: The data is in. AIs are taking our jobs. What’s next?👇

3 Things To Know About AI Today

Uncensored Nano Banana 2 Images Leak… Or Maybe Not

The model was up and (supposedly) uncensored for just long enough for some pretty crazy examples to flood the Internet.

Suprise! It’s no longer available.

But exactly how do we know these are really from Nano Banana 2?? We don’t.

One of the craziest things about AI images now is that these could just be remarkable Photoshops and because the models are so good we believe they’re AI generated. The other hard part now is that finding sources for actual information gets trickier and trickier as people (especially on X) post for clout.

I have major doubts that the images above are real… if for no other reason that Google would likely never release an uncensored checkpoint.

But the generated screenshot and anime outputs seem believable and stranger things have happened.

The Pope is now an AI Influencer

Did you have the Pope weighing in positively on the advances in AI while speaking directly to AI devs about AI alignment on your bingo card? I did not.

But this weekend, Pope Leo XIV took to X to post a message calling for builders of hashtag AI to ‘cultivate moral discernment’ in their creations going forward.

Personally, I’m psyched that the Pope is talking about AI and bringing attention to an important aspect of AI creation. We do want an AI properly aligned with good human behavior.

As we’ve mentioned on the show, AI will be the dominant conversation in world politics over the next five to ten years and the Pope still wields a lot of power.

Is Vince Gilligan’s new show Pluribus an AI metaphor?

I’ve seen the first two episodes of Vince Gilligan’s (Breaking Bad / Better Call Saul) new show Pluribus on Apple TV and it’s VERY good.

I won’t say exactly what it’s about (no spoilers here) but there’s a TON of conversation online around how it might be an AI metaphor.

I love Vince’s work going back to his X-Files episodes and I’m not entirely convinced Pluribus is about AI (again, no spoilers). But his comments above and this narrative illustrates clearly the divide we’re seeing amongst the creative community around generative AI content.

We 💛 This: Grok Imagine Video… Is Good Now?

I saw a tweet over the weekend mentioning that Grok Imagine video had upgraded and I decided to give it a quick shot.

Um, is Grok one of the best video models now?

I have to admit that I’d mostly written off Grok Imagine after we tried it on the show a few weeks ago but I am impressed. It’s not great with sound yet (stick with VEO3 or Sora 2 for that) but it’s worth exploring.

Some other interesting examples I saw:

You can use Grok Imagine directly from the Grok website or within the X app as well (just make sure you’ve updated it).

Of course, leave it to Grok’s CEO (you know who) to make it weird.

Are you a creative or brand looking to go deeper with AI?
Join our community of collaborative creators on the AI4H Discord
Get exclusive access to all things AI4H on our Patreon
If you’re an org, consider booking Kevin & Gavin for your next event!