Erick MORMIN

05/11/2026

On October 20, 2026, the French Senate adopted, without amendment, the bill repealing the Code noir — closing a parliamentary process opened just months earlier. 178 years separate this repeal from the abolition of slavery in 1848.

I devoted my latest book, "Le Code noir, de 1685 à 2026," to this historical and legal journey: the birth of the text under Louis XIV, its enforcement in the colonies, the resistance it provoked, the abolitionist struggle, and the slow transformation of French law up to this year's vote — with particular attention to Martinique.

Beyond the legal fact, one question runs through the book: what does it mean, in 2026, to repeal a text whose direct legal effects had long disappeared, but whose legacies still raise questions for our society today? Twelve chapters, a timeline, and documented annexes, for anyone interested in the history of colonial law, the memory of slavery, and the contemporary issues that follow from it.

📖 Paperback: https://www.amazon.fr/dp/B0HQGL5M7B/ref=tmm_pap_swatch_0
📱 Kindle: https://www.amazon.fr/dp/B0HQD56K5R/ref=tmm_kin_swatch_0

03/11/2026

𝗪𝗵𝗲𝗻 𝗔𝗜 𝗹𝗲𝗮𝗿𝗻𝘀 𝗼𝘂𝗿 𝘁𝗿𝗮𝗱𝗲, 𝘄𝗵𝗼 𝗼𝘄𝗻𝘀 𝗼𝘂𝗿 𝗲𝘅𝗽𝗲𝗿𝘁𝗶𝘀𝗲? — 𝗔𝗜 𝗕𝗿𝗶𝗲𝗳𝗶𝗻𝗴 #𝟮𝟳

An experienced technician documents his procedures for an internal AI assistant. Then he leaves the company. His experience goes with him — but the knowledge base he helped build remains, and the AI keeps consulting it.

This situation is no longer theoretical. It is happening today in organisations that connect their internal knowledge to AI systems.

This briefing explores three dimensions of the question:

→ The distinction between an employee's personal experience, the documents produced in the course of their work, and the company's trade secrets. These do not fall under the same legal rules — and a company does not automatically become the owner of everything an employee knows.

→ The question of recognition. If employees' experience becomes a collective resource exploited by AI, how do we recognise their contribution? How do we avoid turning knowledge sharing into a feeling of dispossession?

→ Data quality and governance. Accumulating documents is not enough: a poorly organised base produces bad answers, contradictory information creates confusion, outdated knowledge leads to unsuitable decisions. In critical activities, human verification remains indispensable.

For decades, companies sought to recruit the best talent. Tomorrow, they will strive just as hard to preserve the knowledge that talent helped build.

But an organisation that can retain its people's knowledge must also know how to recognise those who created it.

▶️ https://youtu.be/ElwasFHeo2c

🎬 Décryptage stratégique IA #27 — Quand l’IA apprend notre métier, ...

01/11/2026

𝗟𝗼𝘃𝗶𝗻𝗴 𝘀𝗼𝗺𝗲𝗼𝗻𝗲 𝗶𝘀 𝗻𝗼𝘁 𝘀𝗮𝘃𝗶𝗻𝗴 𝘁𝗵𝗲𝗺 — 𝗦𝗲𝗿𝗺𝗼𝗻 #𝟵

A relative facing financial hardship. A colleague whose decisions worry us. An adult child struggling to find their way.

We look for solutions. We make calls. We give time, sometimes money, and a great deal of energy. And when our efforts do not produce the hoped-for results, we feel powerless. Sometimes misunderstood. Sometimes hurt.

This reflection applies as much to our personal lives as to our professional responsibilities. Managers, teachers, carers, mentors: everyone who accompanies someone knows this limit.

This sermon explores fundamental distinctions:

→ Showing a door is not pushing someone through it. We can explain, inform, advise. But we cannot decide in another person's place.
→ Passing on experience is not imposing an itinerary. Our knowledge, however precious, does not authorise us to govern other people's lives.
→ Does our help make the other person gradually more capable of acting… or increasingly dependent on our presence?
→ Refusing a solution does not mean rejecting the person who offers it.

The hand we extend must help the other person walk, not walk in their place.

From me… to us…

▶️ https://youtu.be/U_UPPR6TLT8

🎬 Les Prêches de l’Institut AÏJ — De moi… à nous…Prêche n°9 — Aim...

29/10/2026

𝗖𝗮𝗻 𝗔𝗜 𝗯𝗲𝗰𝗼𝗺𝗲 𝘀𝗺𝗮𝗿𝘁𝗲𝗿 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝘂𝘀? — 𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗰 𝗔𝗜 𝗕𝗿𝗶𝗲𝗳𝗶𝗻𝗴 #𝟮𝟲

AI v1 improves AI v2. AI v2 designs AI v3. Faster. More intelligent. More autonomous.

And the human? Sitting at his desk, mug in hand: "Still useful (for now)."

The image makes you smile. The subject is strategic.

AI already contributes to software development. A developer can ask it to generate code. An engineer can use it to identify errors.

But contributing to an improvement does not mean mastering the entire process.

This briefing explores three fundamental concepts:

→ Recursive self-improvement: the mechanism already exists in certain forms. But one improvement does not guarantee the next.
→ Alignment: a system can optimise an objective without understanding all the reasons that objective must be constrained.
→ Superintelligence: a research hypothesis, not an established fact.

On our human's desk: Ideas. Objectives. Security. Alignment. That is precisely what AI cannot set for itself. And the "Next Version?" depends on us.

Technical autonomy is not decisional sovereignty. The question is whether we will remain responsible for what we allow it to become.

▶️ https://youtu.be/9RiMi8ITN6o

27/10/2026

"𝗧𝗵𝗮𝘁 𝘄𝗼𝗻'𝘁 𝗹𝗲𝗮𝗱 𝘆𝗼𝘂 𝗮𝗻𝘆𝘄𝗵𝗲𝗿𝗲…"

Talence, circa 1984. I was a first-year sociology student. A man came to fix the shutter in my student room.

— What are you studying?
— Sociology.
— That won't lead you anywhere.

A sentence spoken almost in passing. But I still remember it, forty years later.

He couldn't have known that this young student from Martinique would take many unexpected turns. That sociology would become a way of seeing — a method, a way of questioning the world that would never leave him.

Later, my path led to IT, then to artificial intelligence. Two fields that might seem far from sociology. Yet behind every network, every algorithm, there are human beings.

Today, I teach. I consult. I write. I analyse the transformations AI imposes on our societies.

And I understand how much an adult's words can matter. A sentence can encourage. A sentence can discourage. A sentence can stay in a memory for forty years.

We never truly know what a young person before us will become. Our responsibility is not to predict the limits of their future. It is to give them the means to invent it.

"You fixed my shutter. Life took care of opening many other windows for me."

🔗 https://www.ekmconseils.eu/post/cela-ne-vous-m%C3%A8nera-%C3%A0-rien

Talence, 1984. Un étudiant en sociologie, un volet en panne et une phrase inattendue : « Cela ne vous mènera à rien. » Quarante ans plus t**d, Erick Mormin revient sur cette rencontre et sur un parcours entre sociologie, enseignement, informatique, intelligence artificielle et littérature. Un ...

26/10/2026

𝗦𝘁𝗿𝗮𝘁𝗲𝗴𝗶𝗰 𝗔𝗜 𝗕𝗿𝗶𝗲𝗳𝗶𝗻𝗴 #𝟮𝟲
𝗖𝗮𝗻 𝗔𝗜 𝗯𝗲𝗰𝗼𝗺𝗲 𝘀𝗺𝗮𝗿𝘁𝗲𝗿 𝘄𝗶𝘁𝗵𝗼𝘂𝘁 𝘂𝘀?

What if an AI could improve another AI? And what if that new AI could, in turn, design a third one, even more capable?

This scenario is no longer science fiction. AI already contributes to software development: code generation, error detection, analysis of experimental results.

But contributing to an improvement does not mean mastering the entire development process.

In this briefing, we explore three fundamental concepts:

→ Recursive self-improvement: a system that enhances its own performance, enabling it to search for further improvements more effectively.
→ Alignment: how do we ensure a system pursues its intended goals while respecting human constraints?
→ Superintelligence: a research hypothesis, not an established fact.

Technical autonomy is not decisional sovereignty. Being able to develop a technology does not mean being able to control all its consequences.

The question is not only whether AI can become smarter without us. The question is whether we will remain responsible for what we allow it to become.

▶️ https://youtu.be/9RiMi8ITN6o

EKM Conseils · Institut AÏJ

🎬 Décryptage stratégique IA #26 — L’IA peut-elle devenir plus intel...

25/10/2026

𝗧𝗵𝗲 𝗦𝗲𝗿𝗺𝗼𝗻𝘀 𝗼𝗳 𝗜𝗻𝘀𝘁𝗶𝘁𝘂𝘁 𝗔𝗜̈𝗝 — 𝗦𝗲𝗿𝗺𝗼𝗻 #𝟴
𝗪𝗲 𝘀𝗽𝗲𝗻𝗱 𝗼𝘂𝗿 𝗹𝗶𝘃𝗲𝘀 𝘁𝗿𝘆𝗶𝗻𝗴 𝘁𝗼 𝗿𝗲𝗽𝗮𝗶𝗿 𝘆𝗲𝘀𝘁𝗲𝗿𝗱𝗮𝘆

A word we regret. A decision we wish we had made differently. A silence we should have broken.

We have this extraordinary ability to mentally revisit events that no longer exist, to replay conversations, to reconstruct scenes. But we cannot go back to yesterday to change it.

Repairing is not erasing. Responsibility is not a life sentence. Forgiving is not forgetting.

In this sermon, we explore these essential distinctions. How do we stop giving the past ownership of our future? How do we transform our legacy rather than repeat it? How do we act today without waiting to change yesterday?

"We often compare our real lives to an imaginary life from which we have removed all difficulties."

Honouring the past does not mean giving it ownership of our future.

From me… to us…

▶️ https://youtu.be/IRYfMNcR9M0

Institut AÏJ — Abraham · Isaac · Jacob
Understanding today to build tomorrow.

🎬 Les Prêches de l’Institut AÏJ — De moi… à nous…Nous passons par...

21/10/2026

𝗢𝗰𝘁𝗼𝗯𝗲𝗿 𝟮𝟬: 𝗯𝗲𝘁𝘄𝗲𝗲𝗻 𝗯𝗶𝗿𝘁𝗵𝗱𝗮𝘆 𝗰𝗮𝗻𝗱𝗹𝗲𝘀, 𝗮𝗿𝘁𝗶𝗳𝗶𝗰𝗶𝗮𝗹 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 𝗮𝗻𝗱 𝘀𝗲𝗿𝗲𝗻𝗱𝗶𝗽𝗶𝘁𝘆

Yesterday was my birthday.

And I received dozens of messages, songs, kind words and thoughts that truly touched me. Thank you to each and every one of you.

But this day had a twist I wanted to share.

While savouring your warm wishes, I was deep in thought about a doctoral research project I'm considering in artificial intelligence.

The questions on my mind concern what I believe are fundamental challenges:

→ Local AI, disconnected from the cloud.
→ Digital sovereignty.
→ Cognitive autonomy in the face of AI systems.
→ The relationship between AI, learning and education.

I was searching for that spark — the intuition that helps frame a meaningful scientific question.

And that's where a concept I hold dear comes in: 𝘀𝗲𝗿𝗲𝗻𝗱𝗶𝗽𝗶𝘁𝘆.

Serendipity is the ability to recognise the value of an unexpected discovery. It's not luck: it's the preparation that makes the mind ready to seize an intellectual opportunity when it arises.

Between two "Happy Birthday!" notifications, this day reminded me that the best ideas sometimes emerge in the interstices of daily life.

I blew out my candles… but I intend to keep the flame of curiosity burning bright.

Thank you again for your kindness. You made this day truly special.

Erick Mormin
EKM Conseils — Institut AÏJ



🔗 Additional links:
Video: https://youtu.be/4KNcgGAmMpY
Institut AÏJ: https://institut-aij.netlify.app/

19/10/2026

𝗪𝗵𝗲𝗻 𝗟𝗶𝗳𝗲 𝗗𝗼𝗲𝘀𝗻'𝘁 𝗧𝗮𝗸𝗲 𝘁𝗵𝗲 𝗣𝗮𝘁𝗵 𝗪𝗲 𝗛𝗼𝗽𝗲𝗱 𝗙𝗼𝗿
Sermons from Institut AÏJ #7

We've all experienced these moments.

We leave a place, a situation, sometimes a part of our past… with the sincere hope of building a better life.

Then time passes. Circumstances change. And life doesn't lead us exactly where we thought we were going.

A question arises: did I make the wrong choice? Did I fail?

But a decision that doesn't produce the hoped-for results isn't necessarily a bad decision. And a difficulty encountered in life doesn't constitute a personal failure.

Regret must not become a prison.

We have the right to change our minds. To revise our plans. To acknowledge that we are tired.

Taking care of yourself is not a luxury. It's a responsibility toward yourself — and toward those you love.

Perhaps succeeding in life isn't about arriving exactly where we planned to go. It's also about finding a new path when the one we chose no longer suits us.

Take care of yourselves. And take care of one another.

🎥 https://youtu.be/pgyzGv77UxQ

From me… to us…
𝗘𝗿𝗶𝗰𝗸 𝗠𝗼𝗿𝗺𝗶𝗻 — Institut AÏJ

🎬 LES PRÊCHES DE L'INSTITUT AÏJ — DE MOI… À NOUS…Quand la vie ne ...

13/10/2026

🎬 AI Strategic Briefing #25
𝗔𝗜 𝗽𝗿𝗼𝗱𝘂𝗰𝗲𝘀 𝗳𝗮𝘀𝘁𝗲𝗿 𝘁𝗵𝗮𝗻 𝘄𝗲 𝗰𝗮𝗻 𝘃𝗲𝗿𝗶𝗳𝘆: 𝘄𝗵𝗼 𝘀𝘁𝗶𝗹𝗹 𝗰𝗼𝗻𝘁𝗿𝗼𝗹𝘀 𝘁𝗵𝗲 𝘁𝗿𝘂𝘁𝗵?

Imagine: an AI produces a strategic report in minutes. Perfectly structured. Coherent arguments. Credible figures. Convincing conclusions.

But who verified the figures? Who checked the sources? Who can guarantee the conclusions are accurate?

We are building machines capable of producing information at extraordinary speed. But our ability to verify that information is not keeping pace.

This is what I call "verification debt": each unchecked document can become the source of another. Each error can be repeated. Each fragile hypothesis can gradually be treated as established fact.

The quality of expression does not guarantee the quality of reasoning. An answer can be fluent and wrong. A line of reasoning can seem coherent while resting on false premises.

Producing is not verifying. Verifying is not understanding. And understanding is not necessarily deciding correctly.

What good is an AI capable of producing everything if we gradually lose the ability to verify what it claims?

👉 https://youtu.be/AtTf8_qkz7o

🌐 https://institut-aij.netlify.app/