The AI Memos

Two memos, written for myself and published on LinkedIn. The first in April 2025, the second ten months later. They grew into a book.

Here they are, unchanged, with everything that looks wrong, naive, or surprisingly right in hindsight.

Memo #1 · April 2025

Personal AI Memo

Reflecting on AI's impact and future.

AI. ๐Ÿค–
I'm excited. I'm fascinated. I'm full of respect. And yes โ€” I'm afraid.
Afraid of the speed. Afraid we're not asking the right questions. Afraid we start believing everything AI tells us โ€” without questioning it. Afraid we lose track of what's true and what's not. Afraid of what happens when we build something more intelligent than ourselves. Afraid of what AGI might mean โ€” not in theory, but in reality.
A reality where intelligence is no longer human-only.

A New Era

I believe we're entering a new era. Technology won't just be a tool โ€” it will become deeply embedded in daily life. Not something we use, but something we live with.

We will talk to machines. And they will talk back โ€” with memory, emotion, and context. We will walk into rooms where robots are not the exception, but the norm. Not in labs or factories โ€” but in homes, stores, schools, streets...

This isn't science fiction anymore. It's the beginning of a new human-machine relationship. And the gap between possibility and reality is closing โ€” fast.

In the future, the Singularity could become a reality โ€” with human and machine merging into a hybrid intelligence.

In fact, what most still don't get: Reed's Law (Metcalfe's Law squared) is driving exponential network effects. This isn't just another product cycle. The pace isn't just fast โ€” it's accelerating exponentially. We don't know where this will lead.

The potential is massive:

Risks

But so are the risks:

Work will change

AI will make much of human labor more efficient โ€” but it will also make parts of it obsolete. This is both an opportunity and a disruption. And this is the shift that's happening in the job market โ€” some jobs will disappear, while others will be redefined.

Still โ€” I'm not afraid of the future of work. We've seen this before. Like the industrial revolution in the 19th century, or the rise of digitalization decades later โ€” entire professions will change.

Not overnight. But inevitably. Some jobs will fade. Others will emerge. All of them will evolve.

This isn't something to fear. It's a shift.
And the best response is not resistance โ€” but readiness.

Upgrade Yourself

Everyone should be using AI to boost themselves.
Not just engineers or data scientists. Everyone. Employees. Students. Creatives. Kids.

Not using it means falling behind โ€” it's that simple. AI will become your personal assistant โ€” smarter, faster, and always on.

This is not the time to fear replacement. It's the time to upgrade yourself. Learn how to work with AI โ€” and you will outperform those who don't.

But also:

Because we're not at the end of the road. We're at the very beginning.
And what we do now โ€” as individuals, companies, societies โ€” will define how this future unfolds.

Future of AI

That's why it matters who controls this technology.
AI should not be in the hands of just a few corporations. It should be decentralized โ€” giving more power back to the people.
Only then can we ensure that AI aligns with our shared interests โ€” not just commercial goals.

To achieve this, we must embed human values at the core of how we build and deploy AI. Responsibility and ethical action are crucial for the success of these technologies.

If we ensure AI is shaped by these values, we can create systems that truly benefit us, rather than undermine us.
The future is ours to shape โ€” let's make sure we do it wisely.

Martin Krenn, April 2025
(Memo drafted by Martin, polished by AI. Future teamwork in action.)

Original on LinkedIn, April 2025
Memo #2 · February 2026

Personal AI Memo #2

Reflecting on AI's impact and future. Ten months later.

AI. ๐Ÿค–
Ten months ago I wrote that I was afraid. Not of the technology. Of the speed.
That fear has not gone away. It has changed shape.

But something else happened. I am building. I am shipping. Ideas that sat in my backlog for months are suddenly moving. And this is just the beginning. The next months will be another leap.

I am still afraid. But now I am also acting. Both at the same time. And I think that is the honest answer most people are not giving.

What Changed ๐Ÿš€

Every founder carries the product in their head. Every detail, every edge case. That used to be a bottleneck. Now it is a superpower.

I shipped my first pull request to our webshop. I build internal tools. And I am just getting started. I did not learn to code. The tools changed what one person can do.

From Operator to Orchestrator ๐Ÿงค

As a goalkeeper, you do not save every ball yourself. You organize your defense. You read the game before the shot comes. You direct your back line. That is what working with AI feels like now.

You are not the operator anymore. You are the one orchestrating agents. Telling them what to do, when, and how. Like a goalkeeper directing his wall before a free kick. Except the wall is made of software.

My Own Infrastructure ๐Ÿฆ‘

I built my own AI assistant that runs my home. Not in someone else's cloud. On my own hardware, under my own control. The API calls go out, but the data, the logic, the decisions stay with me.

The technology that could give power back to individuals is being absorbed by a handful of companies. One policy change and everything built on their API is at risk. That is why open models matter. That is why local infrastructure matters. That is why I run my own setup.

The Adoption Gap ๐Ÿ“ˆ

At KEEPERsport, we are laying the guardrails right now. Setting up the systems, the workflows, the security. If we get this right, we will ship significantly faster in the coming months. While other companies are still debating whether to allow ChatGPT.

A small, sharp, AI native team can outperform a traditional department of fifty. I know because I am living it.

The Uncomfortable Math ๐Ÿ”ข

If five people can do what fifty did before, a lot of roles will change. Some will disappear. Not in ten years. Soon. Most companies already know this. Very few are talking about it openly.

Every repetitive task is a candidate for replacement. It is already happening. The question is not if. It is how fast.

Software Will Die ๐Ÿ’€

But it is not just jobs. Entire categories of software will die. When you can talk to an AI and get things done through a chat window or even your voice, you need far less UI. Apps that are just a layer of buttons on top of a database have no future.

We do not know yet what this means for e-commerce. Nobody does. And anyone who tells you they do is selling something.

Content Is Noise ๐Ÿ“ข

We are drowning in AI generated content. Most of it is garbage. When everyone can produce ten blog posts in an hour, none of them matter.

What matters is the real thing. The thought behind it. The experience. The risk of saying something honest.

Security Worries Me Most ๐Ÿ”’

When you give an AI eyes to read your emails, that is one thing. When you give it hands to access your systems, your bank, your infrastructure, that is something else entirely.

Most small and mid size companies are not ready for this. They plug in an API key and think they are done. They are not.

The Big Picture ๐ŸŒ

The big picture has not changed since my first memo. This is still exponential. Reed's Law still applies. The network effects are compounding. I still do not know where this will take our industry. Nobody does.

But I know this.
The people who open up to this technology, who learn it, who embrace it, they will multiply themselves. The people who refuse, who say this will not affect them, who wait for it to blow over, they will be left behind.

You do not need to be technical. You need to be willing.
Start now. ๐Ÿงค

Martin Krenn, February 2026
(Memo drafted by Martin, polished by AI. Future teamwork in action.)

Original on LinkedIn, February 2026
Memo #3 · August 2026

There is a third memo. Written in German for the first time, and for the first time not for the internet but for a book. That's why it's not here yet.