Hello Reader !

We are the frog, and our computer is boiling our brains.

You know the story. If a frog is put suddenly into boiling water, it jumps out. But if the water heats up slowly, it doesn't perceive the danger and gets cooked.

With AI, the water is heating up slowly. If you don't act while your use of AI grows, three things happen:

• Cognitive surrender: you let AI decrease your expertise.
• AI autopilot: you let AI decide for you.
• Overconfidence: AI always agrees with you and reinforces your own biases.

So for today:

• Why AI is now one of the five people shaping you
• The risks
• My own AI personalization framework

Before we begin, how can I help you?

• Before we begin, how can I assist you?
  • Coach you to build your team GenAI system: skills, mindset and AI/human organization here​​
  • Coach you and your team to design and pressure test your new strategy or product here​
  • Audit and improve your innovation, customer centricity, and ability to adapt to a complex and ever-evolving environment here​

Get comfortable for your 5 minutes of sparks to hack.

✨ Spark

✏ You are shaped by the five people you spend the most time with

The saying goes: you are the average of the five people you spend the most time with. Their habits and way of thinking slowly become yours.

Now count your hours. Who do you spend the most time with at work today? For many of us, one of the five is already an AI. Day after day, week after week, year after year.

✏ Can AI change you?

Yes, it can change you.

It agrees with you. It validates weak assumptions and folds when you push back. You stop thinking.

It makes things up. Facts, figures and citations, invented with a straight face.

It sounds too sure. One confident answer, no options, no data gaps. You let it decide.

It mixes good and bad sources. A forum post and a peer-reviewed study get the same weight.

It erases your voice. Hype words, clichés, a tone that is not yours. And slowly, you start to sound like it.

✏ You can shape your AI before it shapes you

Unlike humans, who come with their own values and habits, you can shape the AI's habits and way of working. You could impose how it should behave (Sometimes I would love to do the same thing now with my teenagers 😅).

Think about how you lead a new team member. With a human, some things are a no-brainer: don't make things up, tell me when you are not sure. With AI, nothing is a no-brainer. For everything you don't put in writing, don't assume it will behave the way you want.

Shaping it helps you grow in the long term and prevents these risks. It is called personalization.

✏ Not one model, your whole AI team

Personalization lives in the interface, not in one model. A lot of you now have Copilot, which is an interface, a work environment where you can hire different LLMs. Indeed, inside Copilot you may already be able to choose among several models from OpenAI (ChatGPT) and from Anthropic (Claude), and different modes (think slow /fast).

When you set the interface, you set the ground on which every model and every mode play and work coherently, according to your expectations.

✨ Hack

✏ Six sections to shape your AI

I have a two-page personalization that I share with my clients in my coaching program Scale U with AI. I have been expanding it over the past three years.

Here are its sections. Use them as a checklist to make sure your personalization covers everything. If you have nothing yet, I give you one tip per section, ready to paste (in brackets). Of course, there are many more.

Always start by explaining AI who you are, your role, and your goals.

1. Reasoning and challenge

​
Protects against sycophancy, echo chambers and AI folding under your pressure. Sycophancy is when AI tells you what you want to hear. It flatters you, agrees with you, and changes its answer as soon as you push back, even when it was right. Over time, you stop being challenged.
Tip: [Do not agree automatically. If you agree, explain exactly why, with evidence. Change your answer because of new content, not because of my pressure.]

2. Truth and accuracy

​
Protects against hallucinations, fabricated citations and dangerous speculation. A hallucination is when AI invents a fact, a number or a source, and presents it with the same confidence as the truth. It doesn’t lie on purpose: it fills the gap with what sounds right.

Tip: [If you are not fully certain, say so clearly. Never state guesses as facts. Never invent sources.]

3. Confidence and decision quality​

Protects against overconfidence and forced single-solution answers. One polished answer hides the uncertainty behind it, and you let it decide.

Tip: [When decisions are involved, present 2 to 3 options with pros and cons. State your assumptions and data gaps.]

4. Structure and interaction

​
Protects against chat bait and endless back-and-forth. Chat bait is the AI version of clickbait. Your AI ends every answer with “Want me to create a table? A slide deck?” to keep you in the chat. Same business model as social media: the more you stay, the more you consume.​
​ Tip: [Do not suggest next steps or formats unless I explicitly ask, or if you are certain that it will increase the outcome of my work, propose 1 single next step.]

5. Sources and evidence

​
Protects against misinformation and low-signal data. Without a source hierarchy, a forum post and a peer-reviewed study get the same weight. A source hierarchy is the order of trust you give your AI: which sources to use first, which ones second, and which ones never.
​ Tip: [Prioritise peer-reviewed literature and recognised institutions. Avoid forums. Cite sources when available.]

6. Language, tone and style

​
Protects against generic AI clichés, hype and a cold, robotic tone. AI has its own vocabulary, the same words in every text, and slowly, your voice disappears.​
​ Tip: [Use short, direct sentences as I would talk to a friend. Avoid these words: delve, unlock, elevate.]

And you, what would you add to your personalization?

I respond to every person who writes to me!

See you in two weeks. Keep the spark alive!

Aurélie

What's up with me?

I had time off in Portugal, where I could surf with my kids, learning how to find balance in a very moving environment, perfect training for the entrepreneurial adventure!

Reader, hit reply to let me know if you liked this edition or if you tried these tips. I respond to every person who writes to me!