RIP TikTokers

Happy Tuesday! Here’s a joke for you. Why did the programmer quit their job? Because they didn’t get arrays. Bud dum tsh! Sorry…

Here’s some news: 

🌴 The first GTA 6 trailer is here after being leaked yesterday.

💬 Meta is quietly ending messaging between Instagram and Facebook.

🚀 NASA plans to land a spacecraft on a world 800 million miles away.

Okay, that’s hyperbole. But what if instead of having to record a carefully choreographed dance routine, you could just upload a single image and let technology handle the rest? 

MagicAnimate, a new diffusion-based human animation framework built by ByteDance (the owners of TikTok) is getting a lot of hype on social media for doing exactly this.

MagicAnimate takes any static image and turns it into a full-blown animation. Imagine the Mona Lisa, and now imagine if da Vinci had the technology to record her dancing. The team behind MagicAnimate made that a reality.

Want to know how it works? According to the team, “MagicAnimate employs a video diffusion model and an appearance encoder for temporal modeling and identity preserving” for the sample image. For the animation process, it uses a “simple video fusion strategy that produces smooth video transition during inference.”

I’m not sure what all of that means either, but the results speak for themselves. 

ByteDance isn't the only company working on turning images into video. Alibaba is building a tool called Animate Anyone that will be released in the future.

So the next time you want to show off your dancing skills, skip the recording, snap a photo and try…

• This extension turns mentions of Elon Musk into “Wannabe Stark.”
• Hitab is an AI-powered tab manager that looks like a Mac dock.
• Spellar AI is a copilot for helping you become better at speaking.

• Lottielab makes it easier to animate your Figma designs.
• Respell uses AI to automate knowledge work.
• SuperDuperDB lets you build AI apps without having to move your data.