📈 TRENDING
The star of every reel below is a character who doesn't exist. A herd of goats watching the match like lifelong fans. A grey alien who moves into a stranger's life and takes it over. A rock duo howling into vintage mics. None of them are real — but the personality is, and that's the part worth stealing.
A goat that watches football like a die-hard fan pulled 8.6M views.
An alien who moves in and takes over hit 79K likes.
A cinematic duet that never happened drew 846K views.
Here's the character move inside each one — and why the personality does the heavy lifting.
Goats at the football night
Video: Watch on Instagram
@realisticaivid drops you into a rowdy outdoor football night, then lets you notice the crowd: half of it is goats, locked on the screen like the die-hards beside them. The gag lands a beat before your brain does, and the closing shot — a goat draped like a ghost — buys the replay.
📈 8,608,592 views — 4.8× the account's average (@realisticaivid)
Why It Works:
Bury the punchline in plain sight. The goats read as background until the viewer suddenly clocks them, which forces the second watch.
Shoot it like real phone footage — night, handheld, a glowing screen — so the absurd part feels caught, not staged.
End on an escalation, not your opening joke. The ghost-goat gives the loop somewhere new to go.
An alien moves in
Video: Watch on Instagram
@ota.visual runs one premise all the way down: a grey alien lands to study humans and just becomes one. It crashes a party between two women, soaks in a candlelit bath, thumbs a phone, hitches a scooter ride. Every human ritual, played by the last face you'd expect — and the deadpan is the whole bit.
📈 1,107,188 views — 5.6× the account's average (@ota.visual)
Why It Works:
Hand a non-human character human chores. The gap between "alien" and "bubble bath" is the joke, and it refreshes every scene.
Commit to the found-footage grain. The lo-fi look sells "this was really filmed" harder than a clean render would.
Run one plain throughline — "became the main character" — so a wordless montage still reads as a story.
An AI band that never existed
Video: Watch on Instagram
@bob.dominator stages a full music video for a song and a band that don't exist. A blonde vocalist and a weathered frontman trade lines into vintage mics under moody light, shot like a label-funded session. The performances sell so hard you forget there was never a room, a mic, or a singer in it.
📈 846,211 views — 17.7× the account's average (@bob.dominator)
Why It Works:
Direct the performance, not just the face — breath, squint, lean-in. Micro-acting is what separates an AI singer from an AI portrait.
Borrow a format viewers trust. The moody studio-session look slots the clip into a genre they already accept.
Cut on the vocal phrasing so the edit feels scored to the track — even though the track and the band are both generated.
Notice the thread: not one of these creators can act, sing, or herd a goat. They just gave a made-up character enough personality to carry a whole post — and that's the skill now. If your reels look incredible but feel like nobody's home, you don't need a better model. You need a CHARACTER worth watching. Want the workflow we use to build ones people actually follow?
✨SPONSORED
Porkbun is the domain registrar trusted by folks who want low prices without the nonsense. Get most domains at cost, free features like WHOIS privacy & SSL certificates, plus real 24/7 human support. Get $1 Off Your Domain
🤖 NEWS & UPDATES
HeyGen pushed its AI Clipping into the API, so a single call now turns an hour of raw footage into scored, social-ready vertical clips with captions. It scans the whole video for the strongest moments, reframes them upright, and hands back drafts you can send straight to a scheduler. You can even point it at a folder and let every long upload generate its own highlights. Worth wiring up if you're sitting on stream or podcast footage you never get around to clipping.
Alibaba unveiled Qwen3.8-Max at the World AI Conference, a 2.4-trillion-parameter multimodal model it says trails only Fable 5 among the systems it tested. It's already live on Alibaba's Token Plan, with open weights promised to follow. The benchmarks are self-reported for now, so treat that ranking as a claim rather than a verdict. Even so, it's a near-frontier model you can prompt today for scripts, research, and image-heavy work without a frontier-lab bill.
Moonshot released Kimi K3, a 2.8-trillion-parameter open model with a million-token context window and native image input. In blind testing it ranked first in Arena's Frontend Code evaluation, ahead of Fable 5, and landed near the top on agentic benchmarks. The API is live now at $3 per million input tokens, and the full open weights are dated July 27. Handy if you want a cheap, near-frontier brain for writing, research, or agent workflows you can eventually self-host.
Pinokio's cocktailpeanut spotlighted OpenEval, a universal agent and evaluation dashboard from RasputinKaiser that runs entirely on your own machine. It scores how your agents and models actually perform across your sessions, so you can see which setup earns its place instead of guessing. It installs through Pinokio alongside the rest of the local-AI stack. Handy if you're juggling several AI tools and want real numbers on which one is pulling its weight.
AgentsView is a new local-first desktop and web app for browsing, searching, and analyzing your past AI coding sessions. It breaks down where your agents' time and spend actually land across every project, model, and tool, turning a month of scattered runs into something you can read at a glance. Everything stays on your machine. Worth a look if you're building with AI agents and have no real sense of what they're costing you.
🤫 THE DAILY SECRET
Posting every day isn't the same as selling.
Daymond John once said a line I keep coming back to: sales cure all. He built FUBU from a $40 budget into a multi-billion-dollar brand, and it comes down to one idea — you can make up your own opinions about your business, but you can't make up your own facts. The only fact your business has is a sale. So go make noise, and go make sales, and people will find you.
So what did you actually sell this week? Not what did you post — what did someone pay you for? If the honest answer is nothing, that's not a content problem. You've been making noise for months and calling it a business. Noise is the easy half. It feels like work — the posts go out, the numbers twitch — but it only gets you seen, not paid. The sale is the half you keep flinching from. A post can't reject you. An offer can.
Posting matters — you do have to make the noise, because nobody buys from someone they've never heard of. But noise without an offer behind it is just a hobby with an audience.
You call it "building an audience" — but really you're putting off the day someone could say no.
You keep "getting ready to launch" — tweaking the price, the funnel, the landing page, so you never have to actually ask anyone to buy.
You count likes as proof — but a hundred people clapping for a free post tells you nothing about whether one person will pay.
Put a price on one thing and offer it to one real person today. Not when it's perfect — today.
Ask yourself
“What could I put a price on right now — and what's really stopping me from offering it to one person today instead of polishing it for another month?”
Here's the thing. You can turn all that content into real income — IF you have the offer and the system that turns a viewer into a buyer. If you're ready to finally get paid for what you make, 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.