How to Make AI Videos Like ohneis652: The Design-First Formula
People searching for how to make videos like ohneis652 want AI tutorials that feel designed, not generic.
Explore Ohneis - Andries Ohneisser ProfilePeople searching for how to make videos like ohneis652 want AI tutorials that feel designed, not generic. I analyzed 6 works, and the formula is a problem-hook cold open, a named aesthetic anchor, and a CTA that turns the lesson into a lead funnel.
Methodology: I analyzed 6 of @ohneis652's published works from 2025-09-10 to 2026-03-23 for hook structure, named aesthetics, parameterization, and CTA closure. All tool and prompt references in this guide are inferred from observable signals and reverse-engineered approximations, not confirmed by the creator. Last updated 2026-06-03.
How to Make AI Videos Like ohneis652 Starts With a Specific Outcome
I tracked the opening seconds across the sample, and the hook is never about generic AI excitement. It always announces a specific design promise. The viewer is told what aesthetic outcome is coming before the demo starts. That is why these videos feel like lessons with a thesis instead of clips that just happen to show a tool.
The strongest version of that move is the olive-oil packaging post. It opens with a bold, almost anti-design claim about human imperfection and then immediately turns that claim into a visual system of rough marker textures, layered collage, and product mockups. The hook is not "look what AI can do"; it is "here is the visual problem and here is the style it will resolve into."
The 00:00-00:05 collage build and the 00:37-00:43 workflow screen frame the post as a lesson in deliberate roughness, not a generic branding montage. The final comment-to-get-the-link CTA turns the hook into a lead capture.
Key Insight: A good hook tells you what aesthetic you will get, not what software is running.
Takeaway: Open with the outcome. If the viewer cannot tell what style they are about to learn, the tutorial reads as generic AI speed content.
Bottom Line: Problem-hook cold opens appear in 6/6 analyzed posts. The viewer knows the aesthetic outcome before the workflow begins.
The Named Aesthetic Anchor Is the Lesson
I counted the specific references the creator names, and this is where the tutorial becomes more than a demo. The named aesthetic is the actual lesson. Wes Anderson is not treated like a vibe word. It is treated like a spec: symmetry, pastel colors, grain, deadpan faces, and a miniature-diorama feeling.
That matters because it gives the viewer a design target instead of an abstract mood. The tutorial can then break the target into controllable parts. Centered composition, palette discipline, and film-grain texture become the real subject. The AI tool is only there to execute the anchor.
The 00:00-00:03 flight-attendant frame and the later ComfyUI node screen prove the point: the aesthetic name is the teaching framework, not the caption. The centered framing and pastel palette do the real argument work.
Key Insight: Named anchors make the demo legible. The tutorial is teaching a visual language, not a software shortcut.
Takeaway: Choose a reference that can be broken into parameters. If the aesthetic cannot be named, it is too vague to teach.
Bottom Line: Named aesthetic anchors appear in 6/6 analyzed posts. The lesson is always framed as a style target first.
Print Logic Turns Vibes Into Parameters
I observed a second pattern in the sample: the creator keeps turning airy style language into physical production rules. Risograph is the clearest example. It is not presented as a nostalgic filter. It is treated like a system of ink layers, halftone dots, paper grain, and deliberate misregistration. That is design theory, not decoration.
The important shift is that the tutorial stops saying "make it look cool" and starts saying "control the print behavior." Once the viewer sees that move, the aesthetic is no longer a mood board. It is a repeatable workflow with actual constraints.
The 00:08-00:12 close-up of printed texture and the 00:26-00:34 workflow/animation sequence make the tutorial feel like print theory translated into AI output. The texture, misregistration, and limited ink palette are the point.
Key Insight: Design theory is what makes the tutorial durable. Vibes alone do not survive the demo.
Takeaway: Translate each style into a material rule. Grain, offset, palette limits, and texture all need to become explicit parameters.
Bottom Line: Print/process framing appears in 6/6 analyzed posts. The creator keeps turning style into a controllable system.
Palette Extraction Turns Style Into a System
I mapped the pieces that move from visual reference to reusable brand logic, and this is where the workflow becomes more than a one-off look. The moodboard post shows the move most clearly: take references, extract the palette, and turn the look into something a system can reproduce. That is the bridge between inspiration and repeatability.
The branding-system tutorial pushes that logic further. It does not just show a result; it shows iteration. A lobster becomes a fish while the style stays intact. That is the core of the formula: the subject can change, but the visual system stays stable enough that the audience still recognizes the brand.
The 00:24-00:34 hex-code sequence turns reference images into a color system, not just a collage. The split-screen references and the extracted palette show the viewer how taste becomes input data.
The 00:16-00:24 AI chat-entry section and the 00:38-00:47 lobster-to-fish revision make the iteration loop visible. The style survives the revision because the branding system is the real subject.
Key Insight: Style becomes repeatable when palette and object choice are treated like system variables.
Takeaway: Convert reference images into a small set of controllable inputs. Color, shape language, and object treatment matter more than the first flashy output.
Bottom Line: Parameterized aesthetic systems appear in 6/6 analyzed posts. The creator keeps translating taste into repeatable inputs.
Motion Control and CTA Close the Funnel
I mapped the final beat across the sample, and the ending is not an afterthought. The CTA is part of the product. In this feed, the tutorial is always building toward a specific ask: comment a keyword and receive the workflow or link. That turns the video into a lead magnet rather than a dead-end explainer.
The motion-control piece makes that pattern especially clear because the tutorial does two jobs at once. It proves that animation can be controlled by start and end frames, and then it ends by asking for a keyword comment. The lesson is clear, but the funnel is also clear. That combination is the business model as much as the content model.
The 00:19-00:27 start-frame/end-frame UI and the 00:36-00:43 comment 'LTX' CTA show the two halves of the formula side by side. First the post proves motion can be controlled, then it turns that proof into a response CTA.
Key Insight: The CTA is not a bonus. It is the packaging that turns the tutorial into a funnel.
Takeaway: End with a concrete comment keyword or link capture. If the tutorial has no next step, it leaves attention on the table.
Bottom Line: CTA funnel mechanics appear in 6/6 analyzed posts. The videos end by converting attention into leads.
Where the Formula Is Harder to Verify
A few parts of the workflow can't be confirmed from the videos alone, and any guide that claims otherwise is overselling the analysis:
- The exact tool stack: the finished clips expose camera language and content signals, not the generation stack itself. The visuals are consistent with a workflow combining multiple steps, and the companion G4 analysis will handle that layer [link forthcoming].
- The actual prompt strings: the alici DB text is reverse-engineered from finished output, so it is an approximation of structure, not the creator's original input.
- Production volume per post: the number of discarded generations and retries is not visible in public output.
Acknowledging these gaps is part of the methodology, not a footnote. If you build on this formula, you still hit the same limits.
FAQ
What is the ohneis652 formula?
It is a design-first tutorial system built on a problem-hook cold open, a named aesthetic anchor, a concrete process demo, and a CTA that turns the lesson into a link funnel. The creator teaches taste as a repeatable workflow, not just a software shortcut.
How do you make AI videos like ohneis652?
Start with a specific aesthetic promise, then turn that style into measurable parameters like grain, palette, symmetry, or texture. Show the demo step clearly, and close with a comment-to-link CTA so the tutorial has a next step.
What AI tools does ohneis652 use?
The exact stack is not public. The visible output is consistent with a workflow that combines image generation, UI walkthroughs, and motion framing, but that is an inference rather than a disclosed fact.
Why do these tutorials feel so designed?
Because each one starts with a design target and then breaks that target into concrete visual choices. The videos teach how to think like a designer first and a tool user second.
Why does the CTA matter?
It turns the tutorial into a lead magnet. The viewer does not just leave with an idea; they are told exactly how to request the workflow, which is part of the creator's funnel.