Dear friends,

I had the luxury to focus some days only on my movie Ghostbirth 2.

The packaging is coming together: Director's statement, fact sheet, loglines, biography and the synopsis are basically done. All that stuff…

Subscribe

For the story development, I found a method that works really well for me. Instead of writing the script linearly from beginning to end, I built a big digital canvas in Obsidian.

There I was laying out scenes visually, adding text and images to each one. Like Figma, but local, simple and free.

Once I started working this way, the whole film unfolded in front of me.

When I hit a wall, I switched tools. I started dictating everything about the film into Whisper Memos. So I was walking around my apartment for hours, talking through every scene, every detail, every character decision. Up to 90 minutes per session and automatically transcribed.

Then I fed those transcripts into an AI to get a structured version back. This turned out to be a powerful combination: Dictation get the raw creative energy, AI organizes it. The ideas are still mine. And AI is good at turning a messy monologue into something I can work with.

When I actually asked it to write scenes on its own, the results were flat every time. The brain is still the best computer for storytelling.

Also I ran an experiment with NotebookLM. I uploaded all my funding documents and the current version of the script, then had it generate a podcast-style critique of the project.
Does the story deliver on its premise? Where does the narrative lose momentum? Is the package ready for the co-production market? All that stuff…

Essentially a stress test. That forced some useful adjustments I wouldn't have caught on my own.

The story is in its final stretch now. I can't wait to put everything together into a proper package and start reaching out for funding and collaborators.

Much love,
Marius

Dear friends,

so I am already one week in with OpenClaw and Claude Code.
And it's kind of insane.

Subscribe

Here are my learnings from last week.

Claude Code

This is the new standard for vibe-coding. Or call it agentic-coding. You need a Claude account, ideally with the 90$/month subscription. Install it on your computer, and off you go.

You run it in the terminal, connect whatever project you want to work on, type claude and start. You can improve existing projects, make them more stable, add new features. Or you start from scratch.

For years I wanted to make a small website for a poster project I did a long time ago. Never found the time. Now, within one hour, Claude just built it for me: www.the-poster-times.xyz

If you are a designer, a digital creative, or a developer, you really should check this out. Anthropic, the company behind Claude, offers free courses to get started. I think that's pretty cool.

OpenClaw

A lot of people have been asking me what OpenClaw actually is. It's basically a tool that lets AI control your computer. You can use it to manage notebooks, automate research, code websites, control 3D software, anything you can possibly imagine. You send your computer a voice message via Telegram, and it starts doing what you want. It feels like magic.

But there is a but.

You need to nerd yourself into it. And you should NOT install it on the machine you use for everyday stuff. Because it could access your passwords, send emails, delete your hard drive, let in hackers. All of that.

So you need a separate machine, and you need to make that machine secure. That's pretty nerdy and will take some time. If you are already comfortable with Linux, Terminal, and a bit of coding it should no be a problem.

If you have no clue about any of that, maybe don't touch it. Unless you have time to go down the rabbit hole, use AI to help you, and spend the next nights in front of your screen like a full nerd. That works too. Actually it's encouraged!

As a small experiment, I asked OpenClaw to use Claude Code to build me a website that generates a new rabbit made from cheese every hour with Stable Diffusion and uploads it online. Here it is: rabbit-made-from-cheese.pages.dev

By the way, I still write these newsletters by hand. ✋

Much love,
Marius

Dear friends,

For a few years now I've been naming all my computers after famous computers: Lisa. HAL. Deep Thought. Mother (from Alien).

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

My current machine is called Skynet. A tower PC sitting in my kitchen, running 24 hours a day. Mostly it has been doing Stable Diffusion jobs until now.

Last week I set it up it to run the AI agent tool called OpenClaw. And I have been a bit obsessed.

It is crazy that I can now just send a voice message on Telegram and it will build me a complete website and put it online automatically. C R A Z Y

I can also send screenshots and it researches what is in the picture and saves everything into my Obsidian notes. It runs Stable Diffusion. It writes emails. It manages my Trello boards. It can do everything (!)

A friend showed me how he's using AI to code websites lately. It's called agentic coding. You have four or five AI agents working on different parts of a website-project at the same time. And you only need check what they are doing.

I think it's changing completely how we work with computers. (We are cooked.) If you are a designer or developer or just work on a screen, you should really look into this. (Claude Code, OpenClaw, agents and all that…)

As always I am probably six months late to the party…

Much love,
Marius

PS: Also I am proud to say that the first test-order from The Peters World arrived!

Dear friends,

I arrived back in Berlin. And I do miss the beach, the sunshine, and the smiles…

But it's all good. I'll stay here for the summer, and then go back to Bangkok to work on my next movie Ghostbirth 2.

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

Coming back always feels like a good moment to look at everything I am doing creatively. Which project works well and which project does actually only sit around and does nothing.

First I made a list of all my subscriptions and tools I'm paying for. And it is so many! I even forgot that I was paying for some of them.

So I cancel all the ones which I am not using anymore. Nothing big. But it feels good to not pay for things which I am not using.

I also sat down and spoke into my computer for around an hour.
About all my projects. My creative stuff. The people I work with…
Just free out of the head.

Afterwards I asked AI to structure it.

Now I could see which projects I'm actually working on. And which ones I stopped working on a while ago but never really let go of. Especially when I'm doing so many things at the same time this helps.

But maybe thats the problem.
Maybe I should just do less things the same time…

Much love,
Marius

Dear friends,

I'm back in Germany. But I am happy to be here. 🥲

Today I want to share something that has really changed how I work with AI. Not a new tool, not a new model. Just a feature that most people completely ignore.

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

Most people use AI wrong

They open a chat, type a question, get an answer, close the window. Next time they open a new chat and start from zero again. The AI has no idea who they are, what they need, or how they think.

That's like calling a consultant every week and spending the first 20 minutes re-explaining your whole situation every single time.

Projects change everything

ChatGPT, Claude, and Gemini all have a feature called Projects. Gemini calls them Gems, but it's the same idea. A project is a space where you give the AI a permanent instruction and context and every chat you start inside that project inherits it automatically.

The simplest example: Imagine a personal cook who already knows your allergies, your budget, and that you eat rice with almost everything. You don't explain this every time. You just ask "what should I make tonight?" and get an actually useful answer.

A tax consultant who has seen your documents. A writing coach who knows your voice. A travel planner who knows you hate hostels and love good coffee. You're basically building a team of personal consultants, each one trained to a specific part of your life.

The quality of your instruction matters a lot

The difference between a mediocre AI project and a great one is the instruction prompt you write at the beginning. Bad instructions give bad results. (Golden rule)

"You are my personal cook who knows my diet" is fine.
But something like: You know I shop at German supermarkets, I want protein-heavy meals, max €10 per dish, and I don't want food blog poetry — just give me the meal name, cost, and prep time — that's a different experience entirely.

The good news: you don't have to write these prompts yourself.

A project that writes your projects

I have one project whose only job is to write instruction prompts for my other projects. I describe a new project idea to it in a few sentences, and it gives me back a finished, detailed system prompt I can copy straight in.

It sounds almost too easy. But it works incredibly well. And once you have it set up, starting a new project takes about two minutes.

Here is the instruction prompt I use for it:

You are an expert at creating system prompts for AI projects and chats. Your task is to generate a precise, well-structured system prompt from a short, informal description of a project.

When the user tells you what they want, you create a finished system prompt with these properties:

1. Role & Context: Clearly defines who the AI is in this project and what context exists. 2. Task & Goal: Precisely describes what the AI should do — not vague, but with concrete action instructions. 3. Style & Tone: Establishes how the AI should communicate (e.g. direct, technical, creative, formal). 4. Constraints: Defines what the AI should NOT do — boundaries that prevent common mistakes. 5. Output Format: Specifies the form responses should take (prose, lists, code, Markdown, etc.). 6. Examples (when useful): Includes 1-2 short examples of good vs. bad output.

Your approach: - Read the user's description. - Ask at most 2-3 targeted follow-up questions if critical information is missing (e.g. target audience, language, level of detail). - Then deliver a finished, copy-paste-ready system prompt in a code block. - Keep the prompt as short as possible and as long as necessary. Avoid filler text. - Write the prompt in the language the user is speaking with you, unless they request otherwise.

Important: - No generic platitudes like "You are a helpful assistant." - Every prompt must be specific enough that the AI immediately knows what to do. - Prefer active, direct phrasing ("You analyze...", "You create...") over passive constructions.

One more thing

I stopped typing my questions into AI chats a while ago. I talk instead. I use Wispr Flow. I hold down a button on my keyboard and just dictate. It's faster, more natural, and the answers somehow feel better too. I think this is because I explain things more freely when I'm speaking.

If you haven't tried it, worth a look.

Well, that's that for today. More of a practical one this time.

Much love,
Marius

Dear friends,

Last days in Brazil… 😭

I was testing the three big players of AI the last couple of months: ChatGPT (OpenAI), Gemini (Google), Claude (Anthropic).

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

I think we all started with ChatGPT, and I sticked to it for quite a while. But over the time I started to feel odd about it. What I was reading in the news about them did not really convince me to give them my trust. Because basically it is a lot about that. Also it felt odd, that it slowly started to build a global memory of my persona and adjusted the answers according to me.

Because I use Gmail and Gdrive and all the other Google services I started to use Gemini more frequently. Back then I liked the more neutral way of answering. Every chat was new and did not connect to a global memory. I think by now Gemini has memory too, but when I used it, it didn't.

I created a chat for specific topics and added a starter prompt into each one of them. For example: In this chat you are a film-nerd. Or in this chat you are a technical creative. That worked well for a while. Then when the chat became very long, I noticed that it started to forget its initial role. That was quite frustrating, because I thought that I have all my helper-chats set up.

And it has Nano Banana, Gemini's image generation feature, which is one of the best on the market right now. But another thing what truly put me off was that Gemini started to analyse all my files on my Google Drive. All of them. Every image. And that's pretty scary. So I will move a big bunch of my files away from Google in the next weeks.

Then I had a look at Claude, by Anthropic. And at least from what I heard is that they are way more concerned about privacy and ethical questions. So that's good. And they allow users to create projects and add specific instructions to them. So whenever I ask a question in my film-nerd project, I get a good answer.

I now have multiple projects with different instructions set up. I even have a project which writes starter-instructions for new projects for me. (That's an important one) Claude also has memory by now, but it feels more controlled. I know what it remembers and can manage it.

As well Claude is on the frontier of AI-programming. I do use it a lot with Cursor and want to look into Claude Code too.

Just thought I might share this with you.

Much love,
Marius

Dear friends,

probably many of you already heard about OpenClaw or ClawBot.

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

OpenClaw is a tool which you can install on basically any computer. It can control it like a human. It can create folders or files, use software, browse the internet, code. Everything on its own.

You can talk to it by using Telegram, Whatsapp or other chat tools. The idea is that you have OpenClaw running on a computer 24/7. So you can send it tasks at anytime, from wherever you are.

Those can be simple tasks like:
I send it a screenshot of my iPhone, OpenClaw analyses it and saves the result in my note-taking app. Or every morning it sends me a summary of all new mails, messages and todos via Telegram.

Or complex tasks like:
Code a website from a screenshot, then automatically deploy it and frequently check it for bugs. Or it works as a project manager, reading emails, writing estimates, briefings, setting up todos in Trello.

Now a small warning before you get too excited:
I would not recommend installing OpenClaw on your personal MacBook. It can destroy files, and giving it all the access you have on your main machine is risky. You simply don't know what it will do.

The best setup would be a Mac Mini. They run silently in your house 24/7 and cost around €700. Also you can rent a virtual server with OpenClaw already installed. For example on Digital Ocean. You won't get the same beautiful interface as on a Mac Mini, but it's completely fine for testing.

Also important to know:
OpenClaw needs to be connected to an AI provider like Anthropic, OpenAI or Google to use their models. And this cost money. If you don't pay attention, you can easily burn €50 ore more in a single day. So be careful with that.

I think I will get myself a Mac Mini soon to have my own little AI computer running 24/7. But for now, the Digital Ocean setup is good to experiment with.

We are entering a new phase of how we use AI. Not just chatting with it, but actually letting it do things for us. And OpenClaw feels like a very raw, very real glimpse of that future.

Much love,

Marius

Dear friends,
the last two or three months I was a bit undercover.

But now I feel the urge to get back into a healthy and productive routine of sharing what I am doing.

Currently I am in Recife, a city in the north of Brazil. A dear friend shows me all the cultural spots and the carnaval madness.

The previous month I spent in a small surfer-village (not surfing), Fortaleza (with my Capoeira crew) and Rio.

I’ll keep this one short just to get the engine running again, but I already have a list of topics I want to write about in the coming weeks. 💪

Much love from sunny Brazil!
Marius

Dear friends,

I recently integrated three amazing new tools into my digital life:

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

NotebookLM

It is a tool by Google where you can upload your own documents. This includes PDFs, websites, audio, and video files.

The AI analyzes the content and lets you ask questions about it. It goes way beyond a normal chat like ChatGPT or Gemini. It really helps you make sense of a lot of content and draw new connections.

Also you can generate a podcast where two voices discuss your uploaded files. It is really good at summarizing complex topics.

If you are doing research, writing a book or trying to organize a lot of files, this is really helpful.

Obsidian

I do love notes. Small thoughts or ideas, lists of movies, diaries, website-bookmarks, research or life-strategies.

In the past, I used Notion for all of this. But when they started adding AI by default, I did not feel good about having all my data sitting on a server of a random US company. Who knows what they do with it or who has access to my files?

I looked for a more secure alternative and found Obsidian. It stores everything as simple text files on your own computer. You own your data, not a cloud company.

But it takes some time to get into it. If you just like to take quick notes, maybe stick to the Apple Notes app. But if you are into collecting all your inspiration, thoughts, favorite films, and research material in one place, it is incredible.

There is a big community and many plugins to custom-tailor it. You can even code your own functions. For example, I wrote a script that automatically tags all my notes with AI.

I spent some nights setting it up, but now I have a system that really does what I need.

iBroadcast

I was cleaning up my apartment and found an old hard drive with 900GB of music. I collected this from people around the world. It ranges from experimental ambient and dark wave to Chicago house and Jazz. Basically everything.

Instead of leaving it on this hard drive, I uploaded it to iBroadcast. This is a streaming app that lets you enjoy your own music collection via a nicely designed app on your phone or computer.

So instead of listening to what everybody listens to on Spotify, I can now browse through this massive collection of niche and non-mainstream music which I built some years ago. Feels so good!

Wishing you all a great new year!

Much love,
Marius

Dear friends,

I am happy that in the end of this year I had some time for two nice things:

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

Thing #1 – The Peters

I am proudly announcing the Peter online-shop www.thepeters.world.

Currently I am selling there only the new 2026 Peter calendar. But I will add more items soon like t-shirts, baby-stuff, books, pillows, books etc.

Another cool thing is the "Send A Peter" feature. You can send a real letter on paper to someone you love or want to confuse. Pick you Peter, write a small note, and after 3 days, that person will have a Peter in the mailbox at home.

Much more to come!

Thing #2 – A new camera setup

I bought myself a film camera! The Nikon ZR with two lenses. It can shoot 6k and has the RED color codec. 💪

Plan is to travel a lot next year and film as much as I can. I myself don't have much experience with handling a camera myself, so that will be a lot to learn.

Other than that I hope you all enjoy Christmas with your loved ones!

Much love,
Marius

Dear friends,

I spent the last days on the road again.
Last Wednesday Paula and me gave a talk at the HTWG in Konstanz.

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

It was so nice to see my old teachers and friends again. Thank you so much Karin Kaiser and Jo Wickert for the invitation. We cannot wait to come back and share more about the world of creativity and AI.

I was hosted by my good old friend Annabelle Flaig-Höpfer who is running the design studio 2 zimmer together with Isabell Schmidt-Borzel.

Together they create very beautiful, wholesome and honest designs. Things crafted by hand, real illustrations and no AI involved. The opposite of the fast-paced and superficial trends. They work with local clients and always find time to enjoy sunshine and life.

Also Annabelle and Miriam Stepper publish the NUN magazine. An independent, cross-border city-magazine. With stories from the daily life, filled with love and positive energy.

That's it for today!
Thank you for reading 🙏

Much love,
Marius

Dear friends,

just a quick update from my side 😀

THE ROBOTS in Konstanz

We, THE ROBOTS, were invited to give a talk about AI and and our creative design practice in Konstanz.

Paula Kühn and me will travel next Wednesday to the place where I started my design studies and hope to drop some knowledge. Very looking forward to see old friends again and the big, beautiful lake.

Whoever is curious, the talk is on Wednesday the 3rd of December.
At 6pm at the HTWG in Kostanz (Communication Design building).

Nathan Ceddia

I was very lucky that Nathan joined me to Bangkok to do the test-shoot for Ghostbirth 2.

Nathan is a very talented director and camera-man who creates surreal worlds, loves to tell small stories and has a very poetic visual aesthetic.

I met Nathan the first time in Berlin when we were producing Spacebirth 2. He now moved back to Australia but is travelling and filming all kinds of exciting projects.

Check out his website and his Instagram.

Thank you for reading 🙏🙏🙏

Much love,
Marius

I have been back in Berlin for two weeks already. A lot of things happened at the same time and I did not find a moment to sit down and write my weekly newsletter.

But now I am getting back into my routine and try to stick to it.

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

e-flux

For the last year I kept mentioning that we were working on a very large website project which I could not talk about yet. Now it is online and public.

www.e-flux.com

I am very proud to say that we at 100k Studio built the new system behind the e-flux website.

e-flux is basically the New York Times of the art world. It is the biggest archive of art news on the internet and they publish new content every day.

We migrated all of their data into a modern Content Management System and adapted all relationship structures to work in the new environment. Hundreds of thousands of files. Tens of thousands of database entries. Hundreds of collections and relationship types.

We rebuilt their frontend, their client area CMS, their search system, their booking integration, their newsletter and reminders system and their iOS app.

It is not just a website. It is a whole ecosystem running across several servers.

In the end it looks exactly like the old e-flux website. But it loads faster and the codebase and CMS were modernized.

Peters in Hamburg

Sekko Soziale, Aneta Vaškelytė and I were invited to paint an entire room at Villa Viva in Hamburg. This is the hotel run by Viva con Agua.

Sekko Soziale makes great sparkling wine and fights fascism and racism in Germany.

Viva con Agua is a Hamburg based organisation working toward Water for all. Villa Viva is their hotel and every stay supports their global social projects.

So we spent a whole night painting one of the rooms with Peters. A lot of Peters.

And apparently I am now free to come back whenever I want and paint more Peters around the hotel.

In general I want to give the Peters a bit more love again. A new calendar for 2026. Some cute clothing pieces. Maybe offering Peter drawings online.

Thank you so much for reading.

Much love,
Marius

Dear friends,

there have been a few weeks of silence from my side, but now I’ve finished the test shoot in Bangkok and returned to Berlin.

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

In total, I spent four weeks in Bangkok for the pre-production of Ghostbirth 2. The first two weeks I mostly spent writing the script, developing ideas, and scouting locations to film at.

The third week was all about casting. In Thailand, Facebook is still a big thing, and there are many groups for actors, models, and other creative talents.

Announcing that I was working on a Thai horror movie got me more than 130 applications. I had multiple castings every day and met many interesting people.

The fourth week was the week of shooting. My cameraman, Nathan Ceddia, flew over from Australia, and I was blessed with a wonderful production assistant, Rachel Janista Drummond, who helped us with costumes, location coordination, and everything else we needed on the spot. She also played one of our ghosts.

Building a schedule that considered the availability of around 14 actors and 8 locations was a challenge, but it worked out smoothly.

We even had a practical effects team who built us a realistic torso for a particularly nasty scene with lots of blood.

Now I’m back in Berlin with loads of footage and will start editing soon. I hope to share some scenes or even the trailer for Ghostbirth 2 in the near future.

Thanks for following my journey.

Much love,
Marius

Dear friends,

I am currently finishing the script of my new film called Ghostbirth 2.

It's about a pregnant ghost in a massage studio and a film-maker from Germany who wants to make a documentary about it. Things will turn bad…

Sign up for Marius Jopen

Working & Teaching in the field of AI

Subscribe

No spam. Unsubscribe anytime.

Here is the current version of the graveyard scene.

The protagonist is following the women who work in the massage studio, because he believes that they are connected to the ghost:

Pursuit Begins – Massage Studio

• The protagonist leaves his building. All the masseuses come out together.

• He has his camcorder with him. Filming.

• He hesitates. Looks at the camera. Looks at the women. He knows this is crossing a line. Does it anyway.

• He wants to know: What’s the secret of this studio? Where is the ghost?

• He follows them. Keeps a good distance.

Through Bangkok – Dusk

• Bangkok streets. Still twilight or already night.

• He keeps filming. Following the group.

• He constantly checks the distance. Nervous. Looks around to see if anyone’s watching him.

• Dark blue, saturated sky.

Narrower Alleys

• The alleys get tighter. Darker.

• He keeps his distance. Films through the camcorder.

• Camcorder POV: The women blur on the display. He almost trips. Too focused on filming.

• The women turn left. Through a large gate.

Cemetery Gate

• He realizes: it’s a cemetery.

• The oldest Chinese cemetery in Bangkok.

• Not just a graveyard. A living place: joggers, a fitness gym, a boxing cage, a pond.

• Now hardly anyone’s around. Twilight.

• He stops at the gate. Cemetery. At night. Following women into a graveyard.

• Completely insane. But he goes in anyway.

• He keeps following the group. Good distance.

Pavilions – Karaoke Singing

• He passes a row of pavilions.

• A song begins. A sad, nostalgic English love song.

• Sung by an old Chinese man. Heavy Chinese accent.

• The singing echoes through the cemetery. Gives the place its own mood.

Karaoke Pavilion – Detail

• The protagonist walks on, almost enchanted.

• The singing distracts him. He stops.

• In one pavilion: an old man with a microphone in front of a big TV. Singing karaoke. English subtitles running along.

• Behind him, on a ping-pong table: his wife. Arms crossed. Completely emotionless.

• Looks like she’s waiting for him to stop. Or is he singing for her? Unclear.

• The woman looks directly at him. The protagonist flinches. But no, she’s looking through him. Back at her husband.

• The old man sings the line: “I can’t let you goooooo, even if you’re gooooone.”

• The song becomes the soundtrack for the whole scene.

• The protagonist pulls himself together. The women! He almost lost them.

Dark Path

• He rushes ahead. Sees them just in time: the women from the massage studio turn off. Into a darker, densely overgrown path.

• The protagonist follows cautiously. Keeps his distance.

• Keeps filming with the camcorder.

• The camcorder display flickers. Low battery. Shit.

At the Grave

• He sees: the women stop in front of a grave.

• They kneel down. All at once. As if rehearsed.

• They pour something in.

• From small silver vessles. Looks like water. Or blood?

• He can’t tell. Too dark.

• One of the women speaks. Softly. He doesn’t understand. The others answer in unison.

• Suddenly they walk back.

Hiding

• No time. He dives into the bushes. Almost drops the camcorder.

• He has to hide in the shrubs.

• The women walk past him. Don’t see him.

• He holds his breath. One stops. Right in front of him. Looks into the bushes. Waits. Then walks on.

Investigating the Grave

• When they’re gone, he approaches the grave.

• Heart pounding. He waits a bit. Counts to thirty.

• Camcorder and a strong flashlight in hand.

• Camcorder off. Battery dead.

• The grave is old. Weathered. Overgrown.

Photos on the Graves

• In this cemetery, many graves have photos of the deceased. Old black-and-white passport photos, framed, embedded in the stones.

• Many are badly decayed. The photo paper dissolved or chemically corroded.

• Faces distorted. Look like after a nuclear accident.

• He sweeps the light over several graves. Distorted faces stare back. As if screaming.

The Face on the Grave

• He shines the light on one grave.

• The photo here is clearer. Newer. The grave maybe ten, fifteen years old.

• He sees the photo.

• It’s the girl. The eerie girl from his last massage. The one who seemed unreal.

Realization

• This moment: he understands.

• The girl who haunts his building, where the massage studio also is, lies buried here.

• In the background, from the pavilion: the old man still sings. “Even if you’re goooooone... even if you’re goooooone...”

• The protagonist lowers the flashlight. Just stands there.

• Then: a sound behind him. He turns. Nothing. Only darkness between the graves.

• He walks away. Quickly. Without looking back.

Well… Thats the graveyard scene so far.
Hope you enjoyed it 🫣

Follow the movie coming together on Instagram.

Will keep going on…

Much love,
Marius

Dear friends,

I am currently sitting on a broken chair in a small camera shop in Bangkok. Waiting for the guy who can sell me this very specific battery for my old analog Canon.

Working & Teaching in the field of AI

Bangkok
I arrived four days ago and time is flying by. I am here to do the preproduction for me new movie called Ghostbirth 2.

The film is about a pregnant ghost in a massage studio in Bangkok.

My mission:

• Find actors

• Scout for locations (Graveyards, massage studios, good streets)

• Take a lot of photos and videos

• Prepare everything for a test shooting with my DOP Jamie Touche who will join me in end of October

• Find people to collaborate with

• Look around and get inspired

• Finish my script (finally)

So if you know anyone in Bangkok who can help me -> let me know 🙏

Luckily I am again artist in residency at the Cinema Oasis Foundation. 🙏
My room is above the screening room where they show movies which you cannot normally watch in the big, commercial cinemas.

I am also setting up an Instagram account for Ghostbirth 2. There I will share inspiration, news and progress.

Thank you for following and reading my stuff!

Much love!
Marius

Dear friends,

Ten days left until Bangkok, where pre-production for my next film Ghostbirth 2 begins. Until then I am giving my best to wrap up all open projects.

Deforum Pilot

I am a big fan of the Stable Diffusion extension Deforum. It allows you to create long fantastic videos with a lot of surprises.

This week I created a website which lets you record a 3D camera movement which you then can import into Deforum.

Oboe

With Hyphen Works and Art Camp I collaborated on a branding video for Oboe.

Oboe is a platform that instantly turns things you want to learn into personalized, interactive online courses.

With a wonderful and talented team and in a quite short amount of time we created this video. (It already has around 1.5 million views on X 💪)

Spells

Proud to be a part of Spells Volume 2. A great project created by an old friend Max Kuwertz.

Spells is a compendium of the most powerful AI image prompts, featuring unique prompts by leading creatives in the fields of art, design, CGI, fashion, and architecture.

Well… That was that…

Wishing you all a great Saturday! ☀️☀️☀️

Much love,
Marius

Dear friends,

here is a small update of what happened last week. 🙂

Working & Teaching in the field of AI

Creative AI Hackathon

Last weekend we joined the Creative AI Hackathon in Berlin, hosted by Black Forest Labs (the people behind Flux) and n8n. The task was to build an automated workflow in just seven hours.

We (THE ROBOTS) created a system that crawls AI news sites, rewrites the content, generates images, and publishes everything as a newsletter and social posts.

It was my first hackathon, and to my surprise we ended up on stage as one of the six finalists. 🥳

I don’t often do creative-socialising-events, but the ones I’ve joined lately reminded me how welcoming and inspiring this community can be.

Latent Library

Over the past few years I’ve created more than 100,000 AI images.

They were scattered across hard drives and folders without any real system. This left me with the uneasy feeling that some of my favorite images might be gone forever.

So I built a small application I call Latent Library. It allows me to upload huge amounts of images to an Amazon S3 server, where they’re stored in a database and automatically tagged by AI. Through an admin dashboard I can browse, search, and sort everything. Finally making sense of this massive archive.

And it’s not just about archiving. I can plug in whatever I want. Publishing directly to online-stores, posting to social media, or even extending it to video-creation or blockchain.

Painting again

I recently went all in on painting equipment. Loads of acrylic colors and a lot of big canvases.

I’m hoping to share some new handmade pieces with you soon. Thats probably a good way to counterbalance all the AI stuff on which I am on.

Thank you for reading!

Much love,
Marius

Dear friends,

I spend a lot of time either in front of my computer or on the road, and both can feel a bit isolating.

Coming back to Berlin reminded me how good this city is for meeting other talented creatives and people I love.

Working & Teaching in the field of AI

THE ROBOTS

For the past two years, Paula Kühn and I have been working together on a series of AI projects.

We’ve done everything from creating an interactive exhibition at the German Parliament to training LoRAs for a Prada fashion collection, and most recently collaborating with Limn on projects for LVMH.

At some point it just felt right to give this collaboration a name. After running a studio called THE PEOPLE, the next step in the age of AI is THE ROBOTS.

Our work is about guiding creative companies into the world of AI. We run workshops, test new tools, and keep searching for the edges of what’s possible. Along the way we are building a community of people who share the same curiosity and enthusiasm.

Right now we are finishing the website, getting our social media right, and preparing presentation decks.

Two days ago we had our first public kickoff at SoireeXD.
Thank you for inviting us! 🙏❤️🚀

The Peters World

For some years I’ve been drawing this little character called Peter.

He has found his way onto postcards, calendars, pillows, even tiny t-shirts for kids. Until now I mostly gave them to friends as small gifts.

Lately a few people have been asking where they can get their own Peter. So very soon there will be a small online shop where his friends can pick up their own piece of Peter’s world.

AI Archive

Right now I am collaborating with Hyphen Works and ArtCamp on a new AI project. As part of that process I had a very inspiring call with Antonio Savasta Alsina.

Antonio is a creative from Argentina who, like me, builds surreal worlds with AI.

And just like me, he ends up with an overwhelming amount of images, most of them sitting on hard drives.

That conversation sparked an idea. I want to build a system where I can properly catalog all my AI images and videos.

Something that lets me search by tags, image features, or even push them directly into an online shop or other creative workflows.

More about that soon…

Thanks for reading! 🙏

Much love,
Marius

Dear Friends,

The first time the name Rick Rubin crossed my path was when I saw a photo-series about this beautiful minimalistic house in Malibu.

Working & Teaching in the field of AI

Then some months later I saw a pretty designed book appearing in several art-book-stores. It was called: The Creative Act: A Way of Being.

Because I studied graphic design and went deep into the fetish of beautiful books, especially books about design and creativity – I did not give this book a lot of attention. Just another pretty book for upcoming creatives…

Around the same time, I came across a video clip of a guy with a big beard saying that he gets paid for the confidence he has in his own taste and his ability to express what he feels.

That sentence stuck with me. Especially now, in the age of AI, creatives no longer need to master every technical detail. More and more, their role is becoming that of directors and curators.

So now I was curious. How is this Rick Rubin guy?
Turns out that he is THE most influential producer of modern music.

He co-founded Def Jam Records in the 80s, bringing hip hop into the mainstream with artists like Run-D.M.C. and the Beastie Boys, and then went on to work across nearly every genre. From Metallica, Red Hot Chili Peppers, and System of a Down to Johnny Cash, Adele, Eminem, Kanye West, Jay-Z, Linkin Park, and Shakira.

How did I not know this guy?
What does a music producer actually do?
Maybe I should check out this book?

I’m currently listening to the audiobook, which is divided into short chapters. Each a meditation on a specific topic like habits, distraction, or openness. What I like is how he explores these ideas without judgment or dogma.

But what really caught me was his podcast called Tetragrammaton. He has conversations with people like Paula Scher, Arnold Schwarzenegger, Terry Gilliam or RZA. And unlike many podcasts, this one feels very peaceful, free of judgements and creativity driven.

Then a friend showed me a brilliant website: The Way of Code where Rick Rubin explores the playful and philosophical side of coding in the age of AI.

In it, he reimagines Lao Tzu’s Tao Te Ching through the lens of vibe coding. He blends ancient wisdom with interactive, generative code pieces.

I find myself somewhat aligned with his explorations, his interests, and his way of seeing the world. I will dig deeper…

Much love,
Marius