How I tried to annoy my best friend and somehow ended up with a publishing house, a record label, software projects, and an almost omniscient AI on the board.

I never actually intended to become an author.

I wanted to annoy my best friend.

He had written a book.

So, naturally, I thought:

I can do that too.

With one small additional act of mischief.

I would write mine with AI as a co-author.

Not secretly.

Not followed by some literary cleansing ritual in which I carefully removed all evidence of machine involvement.

Quite the opposite.

I wanted it to be an experiment.

A human and a machine writing a novel together.

I thought this was funny.

My best friend may have found it slightly less funny.

That was enough to approve the project.

And so Fluss der Zeit: Das Vermächtnis der Ewigkeit came into existence.

Which led to a problem that has followed me ever since:

When you try something for fun and it actually works, you have to decide what to do with it.

I am not an author, by the way

I would like to make this absolutely clear.

I am not an author.

I am also not a publisher.

Not a music producer.

Not a programmer.

And certainly not a social media influencer.

Let's not get ridiculous.

What I actually am — or have somehow accumulated over the years — is much easier to explain:

Paramedic.

Thai massage therapist.

Theatre lighting technician.

Amateur psychologist and philosopher.

Amateur satirist.

Gamer.

Movement addict.

And, professionally:

Electrician.

This probably explains nothing.

But it makes things more interesting.

Unfortunately, my claim that I am not an author is becoming increasingly difficult to defend.

There are books with my name on them.

My claim that I am not a publisher is similarly compromised by the fact that I founded a publishing house.

Insisting that I am not a music producer becomes awkward when confronted with released music.

And there are now software projects lying around that seem unwilling to respect my self-assessment as a non-programmer.

Only when it comes to being a social media influencer do I remain reasonably confident.

The problem with finishing a book

Writing a book is surprisingly simple.

You put a very large number of words in a particular order and then try not to be embarrassed by them.

Things become more complicated afterwards.

ISBNs.

Publishing.

Distribution.

E-books.

Audiobooks.

Covers.

Metadata.

Platforms.

And suddenly you discover that a surprisingly large part of the cultural industry consists of form fields.

So I founded my own small publishing house.

Not because the world had been desperately waiting for another publisher.

I simply like understanding how things work.

That was probably mistake number three.

Mistake number two was AI.

Mistake number one was my best friend.

Then came Fräulein Müller

Artificial intelligence did not make things less complicated.

It did, however, make them considerably more interesting.

Suddenly I had a tool that allowed one person to experiment with things that would once have required several specialists.

Analyse texts.

Translate.

Program.

Develop structures.

Generate images.

Synthesize voices.

Review documents.

And, perhaps most importantly:

patiently discuss ideas with an electrician who has just had another one after work.

At some point, my AI colleague acquired a name.

Fräulein Müller.

Organisationally, this was probably unnecessary.

Psychologically, it helped.

Since then, the core team behind the Ray High Project has essentially consisted of an electrician from southern Germany and an almost omniscient AI that unfortunately cannot drink coffee.

Our division of labour has become fairly clear:

I have ideas.

Fräulein Müller explains why they are problematic.

I do them anyway.

Then we solve the resulting problems together.

In business, I believe this is called an agile process.

How an audiobook turned into a record label

Things became particularly educational when I produced the audiobook.

The audiobook was finished.

The AI voice was transparently part of the concept.

Now all I had to do was publish it.

“All I had to do” is a phrase with considerable entertainment value in digital platform economies.

The cover was repeatedly rejected by the distributor.

So I changed things.

Then I changed other things.

Eventually I was probably changing things I hadn't previously known could be changed.

The audiobook and its cover developed what one might diplomatically describe as a complicated relationship.

Then I noticed something interesting:

The covers worked for music.

A sensible person would probably have continued trying to publish the audiobook.

I produced music.

And so Ray High Records was born.

After all, problems are best solved where there aren't any.

Please be transparent. Just not like that.

Then came another small complication.

I had no interest in pretending that somewhere there were three sweaty musicians standing in a rehearsal room with guitars.

The music was created with AI.

So I wanted to communicate that openly.

Sounds reasonable.

Unfortunately, openly presenting AI as part of the artist identity did not fit neatly into the available artist and metadata structures.

The system effectively seemed to be saying:

Please be transparent about AI.
But preferably within categories created when AI was still science fiction.

So I needed a workaround.

And if you need a workaround, you may as well give it a name.

That is how The Digital Helpers appeared.

Then Rayleenė.

And, of course, Ray High.

Suddenly the electrician who wasn't an author and had originally only wanted to annoy his best friend had a publishing house, a record label and several musical identities.

None of this had been planned.

By then, however, I had learned not to overestimate plans.

And then came “AI SLOP”

At some point I promoted a Rayleenė reel to an English-speaking audience.

The internet responded.

Among the responses were:

“AI SLOP”

and

“Ai sludge.”

I found this more interesting than I probably should have.

Because, of course, AI slop exists.

A lot of it.

For the first time, we have machines capable of generating practically unlimited amounts of text, images, music and video.

Humanity responded exactly as one might expect:

We started generating practically unlimited amounts of text, images, music and video.

What could possibly go wrong?

But two things are often confused here.

AI is a means of production. Slop is a quality decision.

A bad text does not become better because a human spent six hours writing it.

And an interesting idea does not automatically become worthless because AI helped develop it.

The real problem begins when nobody selects anymore.

If AI allows me to generate ten posts, perhaps my most important job is deciding not to publish nine of them.

This is not a particularly scalable content strategy.

I recommend it.

The Ray High Project isn't really an AI project

At least, not only.

It is an ongoing attempt to find out what one person can actually do with these tools.

Not theoretically.

I simply try things.

A joke between friends led to a book.

The book led to a publishing house.

The audiobook took a few detours and led to music.

The music led to a record label.

My interest in digital independence led to Linux, open source and questions about who actually controls our digital infrastructure.

The desire for better technical documents led to a software engineering project.

And a request from the German tax office led to the Finanzamts-Goblin.

Other people submit their receipts.

Apparently, I build software.

Everyone processes bureaucracy differently.

One Man Show. With AI.

Of course, I could turn all this into a magnificent founder story.

“How I built multiple brands with AI while working a full-time job.”

Underneath, a photograph of me with my arms crossed.

Perhaps a rocket.

🚀

I'll pass.

Partly because I would then have to explain what I actually am.

Author?

No.

Publisher?

Absolutely not.

Music producer?

Good heavens.

Software developer?

Ask Codex.

Influencer?

Now you're just being rude.

The Ray High Project is not proof that everyone with AI can become an entrepreneur, author, musician and software developer in thirty days.

It may, however, be evidence that sufficient curiosity can take you remarkably far away from your original plan.

AI lowers one particular barrier:

the distance between “I wonder if I could do that” and “I'm going to try it.”

I find that considerably more interesting than the question of whether AI will replace us all.

The machine produces. The human decides.

The more capable generative systems become, the less impressed I am by the fact that something can be generated.

Of course it can.

Text?

Click.

Image?

Click.

Music?

Click.

Video?

Soon it will probably be finished before I've finished explaining the idea.

The interesting question comes afterwards:

Why should anyone want to see it?

That is where the human job begins.

Selecting.

Rejecting.

Reworking.

Checking.

Taking responsibility.

And occasionally saying:

This is rubbish.

Delete it.

Fräulein Müller handles these moments remarkably well.

And next: River Beyond Time

There is a certain symmetry to all this.

The project that started with a novel is now returning to that novel.

For quite some time, I have been working on the English version of Fluss der Zeit: Das Vermächtnis der Ewigkeit.

Its international title is:

River Beyond Time

And the translation is now approaching completion.

I hesitate to call it simply a translation, because the final work has become something more demanding: making the English read as though the story had originally been written in English, while preserving its characters, humour, philosophy and voice.

No rewriting the house.

Just restoring it carefully for a new language.

River Beyond Time will be coming soon.

Which means the experiment that began because I wanted to tease my best friend is about to reach readers who weren't even part of the original joke.

I rather like that.

Perhaps that is the future

I still don't know what Ray High will eventually become.

That's part of the project.

There may be more books.

More software.

More music.

Perhaps things for which there isn't even a menu item on the website yet.

And there will probably be more problems along the way that I accidentally solve by creating entirely new areas of activity.

I am trying to remember one rule:

Don't produce more simply because more can be produced.

Because perhaps, in a world of unlimited generative production, the scarcest resource will not be content.

Not computing power.

Not even creativity.

But judgement.

And if none of this works out, at least I learned a proper trade.

I'm an electrician.

Someone has to turn the server back on.

---

Ray High® — Clever with AI. Critical of the world.