I grew up with computers in the classroom. Computer class was an ordinary part of my education: we learned to type, use applications, navigate the internet, and play video games. By middle school, many of us carried computers in our pockets. We did not think much of it. Computers and smartphones were simply part of how we learned, communicated, and created. We were computer-native.
But computer adoption across society was not uniform. Some workers and businesses embraced computers, changing the nature of their work. Then a computer-native generation entered the workforce, making computer fluency the baseline. Those who did not embrace this new technology were left behind.
AI is creating the same dynamic. It is becoming core to how students learn, communicate, and create far faster than computers did for my age group. Meanwhile, adoption across the workforce remains uneven: some workers and firms have fully leaned in, while others have barely started. As AI-native students enter the workforce, AI fluency will become the baseline. Those without this fluency will see their prospects narrow, just as workers without computer skills did a generation ago.
The way to become AI-native is to use the technology yourself and use it often. You cannot develop that intuition purely secondhand.
Access is not a barrier. AI tools are inexpensive and available to anyone with a phone or computer. But becoming AI-native requires going beyond chatting with AI and committing time to experimentation. Build an agent to sort your email inbox.1 Use its coding capabilities to create an application.2 Put it in a loop on a task where it can test its own work and improve.3
Watch what other people are trying, then do it yourself. X.com4 is full of builders documenting their experiments, and conferences are popping up around the world.
Try using AI for everything you do. It will fall short, sometimes badly and embarrassingly. Its intelligence is jagged.5 The more you use it, the better you will understand where it works and where it fails. That skill set is what makes you AI-native.
An example of a Grok Bot keeping an inbox sorted, including organizing receipts and billing emails.
Andrej Karpathy’s “Vibe coding MenuGen”, on building a working application with AI.
Andrej Karpathy’s “AutoResearch”, on putting an AI agent in an experimental loop that tests its work and keeps improving.
My GenAI list on X, a collection of builders and researchers sharing their work in public.
Colin Fraser’s visualization of jagged intelligence: AI’s uneven ability to perform astonishingly well on some tasks while failing at others.