The Laziest Way To Make Money With AI
How one AI agent built a product, five content formats, and a deployed waitlist in a single afternoon for $121 — and returned 6x.
June 1stA 26-minute scene-by-scene breakdown of how one creator built a cinematic AI short film with no camera, no crew, using character sheets, reference anchoring, and voice cloning.
The missing step in AI filmmaking is starting with emotional stakes and a three-sentence story before touching any generation tool, then using character sheets and reference anchoring to maintain cinematic consistency across every shot.
AI short filmmaking fails at the prompt stage because most creators start there. The workflow reverses this: begin with a feeling and a three-sentence story synopsis, build character and location reference sheets, then use Cinema Studio multi-shot with an anchor technique where you describe the image first and drop the reference second. Voice cloning via Higgsfield Audio applies to the full audio track in one pass, preserving the original performance emotion while swapping the voice. The result is a complete narrative film made without a camera or crew.
Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.
Create a free account →
Host intro, short film played cold before tutorial: locker room, patrol car, birthday surprise.

Core rule: do not start with a prompt, start with emotional stakes. 3-sentence story structure introduced.

Soul ID creation, character sheet generation in Nano Banana Pro, location reference generation.

Multi-shot setup with character and prop sheet references, handheld camera moves, POV from inside locker, anchor prompting technique demonstrated.

Dashcam aesthetic, Claude used to write prompts from plain-language descriptions, b-roll batch generation, radio interruption for scene transition.

Third character Selena added, dark hallway tension shot, candlelit reveal: the emotional payoff of the film.

Trimming, mirroring shots, discovering mid-edit that shots were missing and regenerating them on the fly.

Higgsfield Audio voice cloning, preset voice for Dave, final film playback with correct voices, comment giveaway CTA.
The reason most AI films look like demo reels is that their creators started with a prompt rather than a feeling, and the rest of the workflow follows from that first wrong move.
“Most people start with a prompt. Thats the wrong move. Start with a feeling.”
“Iteration is the skill.”
“Nothing is locked until you lock it.”
See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.
Everyone talks about AI filmmaking. Adil at Higgsfield actually ships it: a complete short film starring his own face, his own voice, and a story built from emotional stakes rather than a prompt. This is what the workflow actually looks like.
Before generating anything, write three sentences answering these three prompts. This becomes your entire production brief.
Four-step process to lock visual identity before a single video frame is generated.
Describe the image you want first, then drop the character reference tag. The model anchors to your description before pulling from the reference.
Take a screenshot of the last frame of a generated clip and use it as the start frame for the next generation.
“Drop your concept in the comments: 3 sentences. The character, the location, and the moment. I will pick 3 people with the most compelling ideas and give them a free Ultimate plan.”
Giveaway prompt mirrors the 3-sentence brainstorming framework taught in the video.
00:00
00:23
00:50
01:10
01:27
01:40
02:01
02:25
02:42
03:13
03:23
03:40
04:12
04:27
04:39
05:04
05:27
05:47
06:03
06:27
06:47
07:06
07:24
07:38
08:06
08:26
08:54
08:56
09:25
09:44
09:57
10:21
10:45
11:07
11:21
11:49
12:04
12:19
12:50
12:57
13:23
13:43
14:04
14:26
14:43
15:08
15:17
15:37
15:55
16:22
16:35
17:04
17:13
17:38
17:54
18:12
18:38
18:54
19:19
19:45
19:55
20:15
20:40
21:00
21:26
21:35
22:00
22:11
22:35
23:02
23:19
23:38
24:00
24:13
24:29
24:49
25:26
25:39
25:57
26:18How one AI agent built a product, five content formats, and a deployed waitlist in a single afternoon for $121 — and returned 6x.
June 1stA 17-minute experiment that turns Claude plus a video-generation MCP into $1,280 in one day — from zero portfolio to five paid clients.
June 12thA 19-minute deep-dive into the updated AI cinema pipeline — three changes that killed the plastic look for good.
May 27thHow one follow-up prompt turns a Claude chat into a fully interactive Seedance 2.0 cinematography generator — and how to export it as a standalone HTML app.
June 15thA 23-minute walkthrough of every capability Google's new thinking video model unlocks — from reference-based editing to character consistency to a live demo of the Google Flow toolset.
May 20thFive mistakes that turn AI video generators into expensive slot machines — and the structured prompting systems that fix each one.
May 27th