I created an image with ChatGPT, turned it into a video with Melius, and added a bit of typography to make a video around the line:
“y’all actually know what this is for?”
Thanks to Melius, which actually understands what I’m asking for, even someone like me who doesn’t edit videos at all was able to get a pretty decent result.
Can a 30 second doodle really become something you'd ship? 🤨 Apparently with ChatGPT the answer is yes...
I went in skeptical: rough sketch, no love, ran it through the new Images 2.5 model and this came back. Lighting and depth nailed almost instantly, some details still needed a couple markup edits before it felt really good! ✏️
It makes me think, where's the line between a helpful shortcut and cutting corners?
Can I call this productive procrastination? 👀
Was playing with kinetic type while working on another post and liked these too much not to share them separately.
Made in my aftrglo
No timeline. No keyframes. Free to try ✨
Your turn to play!
1M views in 3 days for OpusClip launch.
Obviously not just the video - it's the team, the timing, the distribution push. But I'm genuinely proud to be the part that moves that fast.
Delivered in 14 days.
Have a product launch coming up? Feel free to DM me.
The launch piece makes the product story feel tangible—the checkmark sequence and kinetic type build a clear sense of momentum without losing the human proof behind the result. Great example of motion carrying the narrative.