Making F1 Babies with AI: A 1.5M View Viral Experiment
- Tiki Tim
- Jul 15
- 3 min read
Here at Tiki Media, a lot of what I do is grounded in working with brands and clients — but sometimes, the best ideas come from trying to make your kids laugh. That’s exactly how the F1 Babies video came about: a playful AI-generated concept that unexpectedly took off, racking up over 1.5 million views across social media and landing squarely in the sweet spot between sports fandom, AI video production, and internet humour.
The Spark: F1 Meets Toddler TikTok
My daughter is mad about Formula 1, and I’d started noticing a wave of babyfied celebrity videos doing the rounds on social platforms. But oddly, I hadn’t seen anything in the F1 space. That was all the excuse I needed. I found a trending audio clip on TikTok that featured a high-energy, meme-ready back-and-forth, and decided to bring some of the F1 grid’s biggest personalities to toddler life using AI.
Tools of the Trade: Sora + Kling
For the visuals, I needed a tool that could generate photo-realistic faces that also clearly referenced real-life F1 drivers. MidJourney wasn’t the right fit — it’s not designed to understand celebrity likenesses — so I turned to Sora for image generation. Its ability to reference well-known faces gave me a solid foundation for instantly recognisable baby versions of the drivers.
Once I had the stills — carefully crafted with attention to character likeness, angle, and background — I moved into animation. Kling was my tool of choice here, letting me bring the images to life and apply lip sync to match the original TikTok soundtrack. No extra sound design was needed — the humour and pace came directly from the audio itself.
Getting the Visuals Right
Generating the images wasn’t always straightforward. I had to re-roll and refine prompts quite a few times. Each driver needed a different approach — one generic prompt wouldn’t work across the board. Describing each character individually and adjusting the visual tone took a bit of trial and error, but the result was worth it: they looked just enough like their real-life counterparts to be instantly recognisable as their toddler selves.
Going Viral: 1.3M Views and Counting
I published the video across TikTok, Instagram, and YouTube — and the response was very different on each platform. Instagram really took to it, where the video now sits at over 1.3 million views and continues to draw likes and follows daily, even a month later. TikTok topped out around 326,000 views, still strong but not as viral. YouTube… let’s just say it didn’t quite find the algorithm.
Has It Changed How I Work?
In short: no. I’m lucky enough to have friends who’ve had wild viral success with content, and I know it rarely changes the day-to-day reality of creative work. It’s exciting, of course — and lovely to see people enjoying something you made — but I don’t think it’s healthy or sustainable to chase virality. I’m still focused on creating work I find interesting, that my kids enjoy, and that my clients find useful for their marketing and sales goals.
What I Learned from the Process
The biggest surprise? Just how quickly the video took off — and how long it’s lasted. It still drives interaction a month later, with regular comments, likes, and new followers. That kind of tail is rare and has been a great reminder of how timing, tone, and relatability matter just as much as tools.
If you’re experimenting with AI video content, especially using tools like Sora and Kling, here’s one tip: expect to re-roll more than you think. You’ll need flexibility — what you imagine won’t always appear first time — but with persistence, you can create something fun, relevant, and highly shareable.

Comments