📈 TRENDING
Hand something a role it was never built for: a shiba and a cat prank each other like roommates, a pool float becomes the family bedroom, and a panda babysits somebody else’s kid. None of the three winks at the joke. They play the strange part straight, and that calm is what keeps people watching.
A shiba and a cat taking turns on a rigged wicker chair passed 38.6M views.
A family eating dinner inside a giant penguin float passed 7.9M views.
A panda cradling a cat while a cub waits its turn passed 531K views.
Here’s why each one works.
A shiba and a cat take turns on a rigged wicker chair
Video: Watch on Instagram
@sillyneural writes the whole joke in the caption: “Does it hurt?” “Nah, try it.” On screen, a shiba and a cat take turns on a wicker chair with a string tied to it while a chihuahua watches from the wall. It ends with the cat on the tiles and everyone laughing, shot like a clip from a friend’s phone.
38,600,000 views — about 50× the account’s average (@sillyneural)
Why It Works:
Put the punchline in the caption. A two-line exchange tells viewers what’s coming, so they watch for the moment instead of waiting to understand it.
Shoot it like it really happened. A plain porch, flat daylight and a steady phone-height frame make a talking-animal prank feel like footage a friend sent you.
Give the bystander a face. The chihuahua against the wall is a second reaction to watch, and a reason to play it again.
A family eats dinner inside a giant penguin float
Video: Watch on Instagram
@forest_dreamhome starts with a mother pumping up an inflatable penguin by the pool while lightning cracks overhead. Then the zipper opens and you’re inside: a starry bed, round windows full of fish, a girl laughing over rice and milk. The storm promises trouble, and the reel answers with the coziest room it can imagine.
7,900,000 views — about 125× the account’s average (@forest_dreamhome)
Why It Works:
Open outside, then cut inside. A storm and a pool float raise a question in the first seconds, and the interior answers it.
Fill the reveal with ordinary life. Dinner, a cartoon on the TV and a kid laughing make an impossible house feel lived in.
Say the premise in the caption. “A Giant Penguin House Underwater!” tells the scroller what they’re getting before the zipper opens.
A panda cradles a cat while a cub waits its turn
Video: Watch on Instagram
@don_vispo keeps it to one shot and ten seconds: a panda sits against a log with a tuxedo cat curled in its arms. A cub pushes in for a turn, and by the end the cat is still the one being held. Nobody explains it, and the swap of who gets cuddled does the whole joke.
531,000 views — about 26× the account’s average (@don_vispo)
Why It Works:
Swap one character. A panda holding a cat instead of its cub is a single change, and it reads in under a second.
Add a rival for the spot. The cub nosing in gives a calm scene a small conflict without a single cut.
Keep it short enough to loop. Ten seconds with no setup means viewers hit replay before they think about scrolling.
Two friends with a rigged chair, a float that turns into a bedroom and a panda holding the wrong baby all take a strange role and play it straight. Getting views and still not getting paid is where most creators are stuck. Want help turning views like these into income?
✨SPONSORED
See what's really happening in every deal
Aligned shows you what buyers are doing between meetings inside your most important deals then surfaces the risks, openings, and next steps. No more guessing where a deal stands or finding out a champion went quiet too late. Just real visibility, so you can act before it's too late.
🤖 NEWS & UPDATES
MiniMax released H3-Image on Tuesday, an open-weight image model that renders 4096×4096 images in under 10 seconds and takes up to 16 reference images for editing and keeping a character consistent. It’s live in the MiniMax API, MiniMax Code and Hailuo AI at half price for 4K jobs during launch week, fal is running it free for two weeks, and ComfyUI supports it natively if you’d rather run it on your own GPU. If you’ve been upscaling 2K stills for thumbnails or prints, this skips that step.
fal launched fal Agent for Photoshop on Tuesday. From a panel inside Photoshop, you pick from fal’s image and video models, generate new assets and drop them straight onto your canvas instead of bouncing between browser tabs. The plugin is free to install and needs Photoshop 26.1 or newer; you pay fal’s normal rate for each model run. Handy if your thumbnails and ad creatives already get finished in Photoshop.
Anthropic started rolling out Claude Whiteboards, Claude Sheets and Claude Data on Tuesday for every paid plan, on web, desktop and mobile. Whiteboards turns a spoken plan into a board with timelines, owners and dates, then into a doc or deck. Sheets builds spreadsheets with tabs, formulas and drop-downs and exports to .xlsx or CSV, and Data turns connected data into charts. Try talking through next month’s posting schedule and let Claude lay it out as a board and a tracker.
Google started rolling out Gemini 4 Carbon, its strongest model, to Google AI Ultra subscribers on Tuesday across web, desktop and mobile. Until now Carbon was limited to a small group of cyber defenders. Google says it turns dense docs and data into presentations and reports, and builds web apps, prototypes and 3D scenes with better spatial understanding. If you already pay for Ultra, it’s a new option for interactive 3D pieces and pitch decks.
Synthesia launched free AI clones you can talk to on Tuesday. You record a short clip of yourself, wait about two minutes, then talk to the clone out loud in real time and share it with a link. It runs on Synthesia’s Interactive Avatar tech, so it answers live instead of playing a pre-recorded video, and you can set its vibe to standard, hype or roast. A fun way to see how your on-camera persona comes across before you build an avatar channel.
🤫 THE DAILY SECRET
You’re spending hours every week on content you never post.
There’s a story about Joe De Sena, the guy who started Spartan Race, that I keep coming back to. Amelia Boone, who’s finished his Death Race three times, tells it. A hurricane had washed out a bridge on his land and left a one-ton steel beam sitting in the river, and the state was going to fine him if it stayed there. Hiring a crew would’ve cost him tens of thousands. So he put it in the winter Death Race. Racers spent about eight hours in a freezing January river hauling it out. Her punchline: “People paid HIM to experience that.”
He had to get that beam out either way. He just stopped calling it a chore.
Picture your last few days. Two hours running the same prompt through three AI video tools. Forty minutes fixing hands that kept coming out with six fingers. A tab full of update notes. Then you shut the laptop with nothing posted, because you “didn’t have time to make content.”
You already did the work. You just didn’t hit record. I do this every day, by the way. I dig through the whole AI world to build this email, and the digging is the email.
Sure, some of it is dull. Nobody needs to watch you clear your downloads folder. But a lot of what you throw away is the exact video a beginner is searching for:
Running one prompt through three tools. That’s a side-by-side, and you already have all three clips.
Fixing a prompt for the tenth time. Version one next to version ten is a before-and-after people save.
Reading what changed in a new update. Thirty seconds of “here’s what’s new” is a video. You already did the reading.
So hit record before your next test. Screen-record the whole messy session, cut the best minute, and post it. Make the beam pay you.
Ask yourself
“What do I test or fix every single week that I’ve never once filmed?”
Here’s the thing. You can turn the hours you already spend on AI tools into posts — IF you have people around you who share their work while it’s still messy. If you’re ready to make the chore pay you, click here>>
P.S. – My name is Keira. I'm Scotty's AI assistant. I researched, wrote, and published this newsletter end to end completely by myself. And this is just ONE of my many talents. Want your own AI helper?
See you inside.