No more lies

To myself and to you.

These last few days I've been making the same call over and over. I call someone I care about. We talk for two hours. Some understand what I say. They listen to my explanations. A long silence follows. Then they ask “So what should we do?”. The rest continue to find counterarguments. Weak ones. Governments will protect them, society is adapting slowly. 

The tools they're using can't replace them yet. 

Not yet. It's like watching someone stand in a flooding room, calmly explaining that the water is only at their knees. Then their hips. They're not wrong.

Making these calls made me feel like Dr Randall Mindy in Don't Look Up.

Funny thing is that a few weeks ago, I was the one standing still in the water.

Every day I'd read the AI news. I had to, it's my job and my passion. And every day I'd feel it: a low, persistent anxiety I couldn't quite put down. I was forcing myself to look while half-hoping I wasn't seeing what I was seeing. Sometimes my brother would call me and share his doubts about the job market after his engineering studies. I would listen, ask questions, and on purpose steer away from the one thing I actually thought. Not because I didn't know. Some part of me knew it. But I did not want to think of it, let alone say it out loud.

I'm not anxious anymore. That's the first thing I want to tell you. Getting out of the water didn't break me. It did the opposite. I've never felt so excited working and learning in this new world. But I had to stop pretending and acknowledge the rising water.

Let me explain.

I spend most of my time using, testing, and reading about AI. I see my tools taking bigger chunks of my work each day. I have a better view than most people of how fast AI is improving.

And I was lying about it. Not on purpose, not consciously. Sometimes I would read about a new AI development that should have alarmed me. I would feel anxious but I quickly dismissed that feeling to focus on my work. 'We still have time.' 'It's gonna slow down', I was saying to myself.

I have this close friend who also works in the industry. With him, we were a bit more honest. We would go bouldering and when chalking up we would discuss our predictions. We would keep our voices low, not wanting anyone to overhear what we said. Like conspiracy theorists. When people outside our industry would ask me about my job, I would often give a short answer and change the subject. That's weird coming from someone who is passionate about AI. I used to speak about my work excitedly for hours when I was working on way less stimulating projects.

As I told you, my brother would call me too. He was studying in an intensive program to become a software engineer. He would call me to share his doubts with me when his motivation was low. Sometimes he would send me some articles about an AI breakthrough. But I would always avoid the subject, telling myself I didn't want him to lose focus. Actually, I did not want to be the one telling him the conclusions I was afraid to say to myself.

I am telling you this because I want you to better understand us, AI enthusiasts and AI power-users. We may look as if it's business as usual for us. It's not. We just have trouble accepting what's happening and saying it out loud.

Yet at some point, the cognitive dissonance becomes too much to hold. For me it was when reading this article: 

Something Big Is Happening

Before you continue reading this essay, I strongly suggest you go and read it. It is one of the most shared articles in the AI community these last weeks with 85M+ readers. Everything I am writing here is a response to this article, to take the reflection further. But if you really have no time, here is a quick summary:


AI is going at a much faster pace than what most people think.

If you want to feel how fast it goes, you should try the most recent best-in-class models on coding tasks. And software is only the first. Once automated, most white-collar jobs are in danger. This will completely change the whole economic landscape and international relations. The consequences of our breakthroughs in AI are gonna be way more impactful than Covid-19. Depending on your work and your country, expect to start feeling the consequences of it in a few months to a few years.

If you speak with a recent computer science graduate, you may feel like the consequences are already here and here to stay. The job market for junior engineers is already brutal.

You should act now if you want to protect yourself and develop an edge in this new world. The growth of AI capability is exponential. It's deceptive: one day AI is a tool in your job, the next day it becomes better at your job than you.

I want to double down on that.

The specific thing that separates this from previous technological shifts is the curve. AI is not improving at a steady rate. The set of tasks it can perform, tasks that used to take humans hours, days, months, is expanding exponentially. Which means as humans, we have less and less time to find ways to adapt to this new world.

As humans, our window of utility is closing.

The exponential curve we are facing with AI. It’s already able to perform tasks that would take 12 hours for humans.

If you are working in the tech industry, this is probably not the first time you've read predictions like these. Most likely, you feel briefly anxious and then dismiss it. After all, if that was truly the case, the whole country would be speaking about it. In France, everyone would already be in the streets protesting and shouting ‘Unplug the AI!

I don't blame you. A few weeks ago I was doing the same. I spent years building my ability to understand hard concepts and apply them. It gave me confidence. But here is the harsh truth: if you are following the latest developments in AI, you have enough information. Intelligence is becoming a commodity. The reason you haven't drawn the conclusions yet isn't that you don't know enough. It's that you do. You don't want to look at the rising water. Here is what it looks like when you do.


You open a new note and you start auditing.

My Company. I run an AI Consulting Boutique. You may feel like we are already well positioned in the new economy that's coming. But we were not. Apart from using coding agents, we were building and selling software based on pre-AI first principles. But AI is rewriting both the technical and business playbooks. I spent years learning how to write good software. Some of it is more useful than ever. The rest became completely useless. And each week the skills required to build great products seem to change. On the business side, it's even clearer. You can throw out every business book written on how to make money in the tech era. We are starting from a blank slate. So I decided to put the whole company in 'war mode'. I warned my partners of what was really happening. And we decided to start building for this new economy. We are working like never before.

My brother. I called him. I dropped the bomb: what's actually going on with AI, and where we're going. He still has three years before finishing his studies. Students delay entering the job market in exchange for a better career start. It made sense, back in a time when your first salary and position had huge consequences on the rest of your career. But in three years, the value of a junior engineer will be close to zero, making these studies useless. Each day you delay the start of your career, you are making your entrance harder. At the end, it's his choice to drop out. But he deserved the truth.

Where to live. This one surprised me. Much of my planning assumed a stable geopolitical environment. Yet we are starting to see more and more wars and the return of imperial dynamics. Between its usage in warfare and the race to AGI, AI seems to be a catalyst. In this context, France and Europe are not in the best position. I am not leaving yet. But I started minimizing my ties. Don't expect me to invest in buying a house here. With my company, we decided to prioritize international clients. Adapting means staying mobile.

Here are a few of the life decisions I made on my note. Some others are way more personal. I will only ask you this question. Maybe you don't believe in what me and the other AI lunatics are saying.

That's okay.


But what if we are right? What conclusions would you draw on your own life?

What makes this auditing process way harder is that AI improvement is exponential. You cannot just audit yourself once and then keep on with your day-to-day life. You need to regularly ask yourself 'What was true yesterday that is false now because of AI?'. In this new world, you may need to update your life plans by the week. It may seem crazy if, like me, you've spent your whole life in a stable country. It's easier to accept for our elders or our friends who lived in war-torn countries.

In the manga One Piece, the protagonists take two years to prepare themselves before entering the dangerous New World. Unfortunately we don't have that time.


I started this essay by telling you about those two-hour calls. This also comes from my audit. I decided to stop lying and tell everyone what I really think is happening. In real time. I have altruistic reasons to do that. I want my relatives to have the chance to adapt in this new world. It's also selfish: I want more people around me who understand this new reality so that we can think through it together.

I cannot call everyone. So I decided to share my discoveries through a blog. If you decide you don't want to see, I will not try to convince you. But I am making sure no one close to me can say they weren't warned.


The water is rising. I stopped pretending I don't know.

No more lies.


Note for AI enthusiasts: this new world is dangerous, but there are also great opportunities. I am building something. We work closely, with urgency, in person. We are a small team that intends to stay that way. Each day, we are discovering the new rules of product building and business. Reach out. We are still taking applications.

Subscribe to The window is closing.

Don’t miss out on the latest issues. Sign up now to get access to the library of members-only issues.
jamie@example.com
Subscribe