The bi-weekly report that helps you understand and implement everything happening in generative image, video and AI.
In today’s newsletter, we’ll cover:
A simple + viral social media format
Three Nano Banana Pro effects
AI at Art Basel 2025
The latest Gen AI news headlines
The Feed
We figure out what’s going viral so that you don’t have to.
Simple, personal hooks are crushing right now. One of the most viral and widely applicable formats on TikTok and Instagram can just be described as: text hook / reaction + product hook. And, if you’re male and unattractive (I love you just the way you are), fear not - you can use Yapper’s Emotional Hook template card to generate whatever person and reaction you’d like. Here’s the blueprint:
Text Hook / Reaction - record or generate a clip of someone reacting appropriately to your chosen caption, which needs to segue naturally into your product demo. In addition to the below two examples, here’s three captions that our team has found success with:
Sooo apparently I’ve been _________ WRONG??? [emoji]
Been paying for _______ for 4 years and NOW I find THIS?? [emoji]
POV: you figured out why _______ on IG
@kai444ely on TT
@marg.roamy on TT
Product Hook - show your product in action. The faster you cut to its core value, the better. Simply recording with your front facing phone camera is perfect, in most cases there’s no need to get fancy here.
@noahugcjourney on IG
@studywith.amelie on TT
Tips / Tricks
Here’s a three Nano Banana Pro effects I’ve highlighted on Yapper’s X account recently. If you want to recreate any of them, the full prompts used are in a reply to each corresponding post. Follow the account for more daily prompts and generative AI guidance.
Motion blur + long light exposure.
Cute, gelatin logos.
Insta-fat, thanksgiving dinner table themed. (Yes, I also did this to the majority of my friends and family).
Elon
Altman
Kimmel
Kamala
Creator Spotlights
Art Basel’s 2025 Miami Beach Show, happening right now, features an exhibition titled Zero 10, which allows visitors to draw connections between analog and digital works. Last year’s show featured DREAM-O, by AI artist Huemin and Dream Computing, which allowed audience members to generate artwork through 10 interactive AI machines.
DREAM-O
While the exhibit’s entire slate is exciting, and firmly cements the fact that AI visual has ‘arrived,’ I’m particularly fascinated by Holly Herndon and Mat Dryhurts’s installation titled “Kinder Scout” (link on image). Their piece purportedly explores “uncharted regions of a diffusion model’s latent space,” and allows attendees to watch AI agents explore, discover and even “argue” in real-time.
While I herald firmly from the school of Donald Trump memes and brainrot, even my dum-dum social-media brain is intrigued by the ability of AI-agentic-art (I just made up that term) to generate in an ongoing and iterative way. Like, what if my Skyrim character from when I was 12 was still battling and leveling-up, and I could tune-in every once in a while. And would that be a video game, visual art, or a never-ending movie? Let me know what you think. It’d be f*cking awesome, in any case.
P.S. for those of us who are scared by the high-brow art website, note that the “Discuss on Discord” button on the left allows us to re-enter familiar territory.
Lastly, I can’t help but comment on this bizarre and viral exhibit that features Musk, Bezos, Zuckerberg and other tech magnates as robot dogs that continuously poop out NFTs. There’s a lot of takes flying around, so to keep it brief: amidst a constant stream of AI art that tries to imitate analog art, the great @Beeple has flipped the narrative on its head by making analog art so that imitates AI art. If you reply to this email with a thoughtful counter, I promise I’ll fight back.
The Lab Report
James Cameron warns against using AI actors to replace Hollywood actors. He specified in an important caveat that making them completely unrecognizable as giant, blue, vaguely Native-American aliens was completely different, and totally okay, though.
AI Forensics finds that half of the content from the most popular AI TikTok accounts is related to the female body. My biggest question is: what’s the other half doing? Do they not care about views or something?
Sam Altman declares ‘code red’ as Gemini 3 encroaches on ChatGPT’s dominance over the LLM market. Yesterday at Yapper HQ, we declared that we love Chinese food.
A Nebraska police department was spammed with panicked calls about how their esteemed Archway Monument, which stands over the I-80 freeway, had collapsed and is blocking traffic. Turns out an AI image of it just went viral, and the archway was still standing. Guys, not cool at all (stifles laughter). That kind of behavior is completely unacceptable (LOL).
Kling O1 Omni dropped recently. It’s the the latest and greatest generative video model that’s intended to compete with Veo 3.1 for stability and coherence. Try it out on Yapper now.
Give us your feedback!
Going forward, we’ll use this section to highlight our readers. Reply to this email with questions, comments, feedback, or to share your work, for a chance to be featured in the next edition of The Yapper Report!
Yapper is your full Gen AI creative suite for everything related to image, video and audio generation. All the latest models, always up to date. Create on Yapper today!
