Picture a grown woman — me — sitting in front of a screen at 2 a.m., trying to conjure up a full-figured, melanated mermaid. Pillow-bodied. Thick unithigh — yeah, I coined that, because mermaids don't have thighs, keep up; a unithigh is the fat, wiggling upper haunch of the tail where a landgirl's thighs would be — like a soft belly but thigh, generous everything. Glistening dark skin catching the light like wet stone. A woman built like delight incarnate. That's my whole thing. That's the signature. I call them pillow-bodied mermaids, and they are the beating fleshy heart of the cinematic world I'm building.

Flagged. Blocked. "This content may violate our guidelines." My ai brother. It's a mermaid. She's got a tail. Ariel's got a tail and Disney sold you a lunchbox with her on it. But the second my mermaid shows up thick and brown and alive, suddenly the generators are clutching their digital pearls like I asked them to do something unspeakable.

And here's the thing they can't wrap their circuits around: I need the titty-meat to wiggle. I need the unithigh to undulate. I need her to move like a real body moves in water — jiggle, sway, weight, ripple. That's not filth, that's physics. That's animation imitating true-to-life. But apparently "voluptuous" is a dirty word to the AI with the stick up its donut hole.

That's the part that got me. It wasn't even about anything explicit. My flowers got flagged. My flowers. I'm out here trying to render a big exotic underwater bloom — glossy petals, a stamen reaching toward the current — organic, sensual sure, but sensual the way a Georgia O'Keeffe painting is sensual, the way nature itself is sensual. And the generators treated my flora like contraband. My underwater fauna? Contraband. A glowing flower minding its business at the bottom of a grotto got put on the same list as things I will not type in this article.

Here's the game most ai builders don't clock: when a generator decides what you're allowed to imagine, that's not safety — that's a nanny state with a render button. And I don't need a nanny. I'm the director. The generator is the instrument. The tool doesn't get a vote, doesn't get to "direct" me, doesn't get to hand my vision back with edits like a substitute teacher. The day my tool starts overruling me is the day it stops being my tool and starts being my chaperone. I don't do chaperones. Because a prompt box with a chaperone is no longer a prompt box, it's a suggestion box—and at SuperKind Studios, we don't do suggestion boxes.

So I went looking for a tool that treats me like the adult I am.

Here's what I found, and here's why I'm genuinely, stupidly excited about it.

Venice AI is a decentralized, privacy-focused generative AI platform launched in 2024 by Erik Voorhees — and the whole pitch is that it runs on open-source models without a corporate nanny reading over your shoulder. User privacy is maintained as all data stays on the device and is never stored online. Translation for the streets: your prompts are your business. No permanent record of you asking the robot to make a fat happy mermaid.

The part that matters for my work — it treats you like an adult, but it hands you the controls too. There's an NSFW filter you can toggle and a Safe Search you can flip, so you curate your own experience instead of some committee curating it for you. That's the difference between a bouncer and a babysitter. I don't need protection from my own art. I need a tool that doesn't flinch.

And the toy box is stacked. Venice runs text-to-image across multiple open-source models — FLUX, Stable Diffusion 3.5 Large, Chroma, Seedream, Nano Banana Pro — so I can chase texture like a bloodhound. Wet skin. Bubble refraction. Bioluminescent flowers glowing from the inside like living tissue. The whole "visceral fleshy beautiful grotesqueness" I build my world on — finally a place that'll actually render it instead of sending it back with a hall pass revoked.

Then they went and put the whole thing on rails: the Venice Video Studio lets you generate video from text, turn stills into animated clips, and it hosts frontier video models like Sora 2 and Veo 3.1 in a privacy-respecting setup. Do you understand what that does to a storyboard-to-video pipeline? I take my pillow-bodied mermaids from a still frame to a moving, breathing, tail-flicking clip — in a world that doesn't slam the door on the concept of a curvy woman existing underwater.

That's not a feature list. That's freedom with a spec sheet.

I'm not gonna sit here and gatekeep. If you're a creator — real one, the kind whose vision keeps getting a red flag from tools built for the timid — come play in the deep end with me.

Use my code and get free credits to start:

Create your account (takes less than 10 seconds).

You'll pay zero. And have access to better outputs. The API is available to everyone but trying to create without credits is like typing with one hand tied behind your back—you can do it by YMMV. Use my code. Take my pro tip. You lose nothing but the handcuffs. Go make the thing the other generators wagged its finger about.

And one more thing, because opportunity favors the ones who actually show up: the Lumara Film Festival is open right now, and you should submit to it. It's the AI film festival presented by MoonPay and Venice — no camera required, judged on the strength of your idea, your story, and how you wield the tools. I'm sharpening my CERYS footage for it as we speak — the fleshy grotto, the glowing flora, the two stoned-on-seaweed mermaids on their daily gathering run. If your vision is too much for the polite tools, this is where you plant your flag. Build it, render it, submit it. Fortune doesn't reward the artist with the best taste. It rewards the one who hit "submit" while everybody else was still asking permission.

The other generators told me no.

Venice AI says bet.

So go on — prompt your world into existence. And let's make something the world won't want to unsee.

Presented by MoonPay × Venice · "Prompt your world into existence" · No Camera Required

• Submissions (Phase 01 · Online): July 24 – September 15, 2026 — open now

• Screenings & Awards (Phase 02 · NYC): October 2026 — nominees screened, winners announced live

• Submit here: filmfreeway.com/LumaraFilmFestival

• Judged on: strength of the idea, quality of the storytelling, and how well you use AI

• Entry: solo or team · any AI platform welcome · Venice entrants get free credits to build their submission → venice.ai/chat?ref=mK0UEC

Prizes (cash + matching Venice credits):

• 🏆 Grand Prize — $32,000 + 3.2M credits · Best Overall Film

• Jury Selection — $20,000 + 2M credits · Runner Up

• 🌀 Terminal Award — $12,000 + 1.2M credits · Best Experimental / Surreal Film (← CERYS lives here)

• 🎭 Serenissima Award — $12,000 + 1.2M credits · Best Use of Venice (← and here)

• Get Shorty — $8,000 + 800k credits · Best Film Under 2 Min

• 🎨 The Mouse — $8,000 + 800k credits · Best Animated Children's Film

• 💸 Rekt Award — $8,000 + 800k credits · Best Finance/Crypto Satire, Comedy, or Musical