Language > Tool
When programming moved from procedural to object-oriented, developers had to learn a new conceptual framework. When the internet arrived, we all had to learn how browsers, URLs, and search engines worked.
These shifts happened within specific domains. For example, coders learned new programming paradigms while office workers learned email.
By now, we can see how AI cuts across every domain. Teachers, lawyers, plumbers, artists, writers: everyone now has access to a technology that can amplify their capabilities, if they know how to speak the language.
Thomas Kuhn wrote the book on paradigms and in it he discuss incommensurability, which is about people who operate in different paradigms who can’t understand each other because they lack a shared language.
This is where we are in 2026. I don’t think it’s about a decision to vibe-code or not. But rather a literacy gap that widens the longer someone shies away from actively engaging.
And just like learning any language, the best way to do it is immersion.
You Should Know Some Basics
The command line exists. You should know what it is and how to copy-paste into it.
GitHub matters. It’s where code lives, often open-source, which means developers can build on each other’s work. Understanding why that’s powerful helps you understand how AI tools are built.
APIs connect things. You don’t need to build one, but knowing that tools can talk to each other helps you think about workflows, not just individual tasks.
These are table stakes.
Now For The Fun Stuff
Talk
the language of loveAI. Prompting is not poking. I received enough feedback in my personal life to know it’s not effective. Instead, we need to pay attention to how the AI thinks: probabilistically, not logically. Hmmm, back to my personal life apparently … the AI doesn’t know things - not anything really, but it predicts patterns better than Rainman. That changes how you ask questions.Compose workflows. One AI call rarely solves a problem. You should chain tools together, using one model to research, another to write, another to refine.
Navigate ethical questions. If an AI wrote it, who owns it? If an AI suggested it, do you trust it? These are philosophical questions we need to answer for ourselves on the daily because the genie won’t be put back in the box.
Front vs. Back
The old distinction between “front-end” and “back-end” developers is blurring. You don’t need to be either anymore to build functional tools. But mastery still requires depth.
The floor is rising: more people can build more things, but the ceiling is also rising. Specialists will always matter.
The key difference is that generalists with AI fluency can own entire value chains by themselves.
-Pulp Conversations
The Choice
Everyone who learns AI will face a decision.
Will you use AI to gain advantage at someone else’s expense?
Or you will you use AI to amplify your ability to help others?
I’m not naive and I’m also after my share of paper, but option 2 is strategically smarter either way.
The people who thrive will build communities, sharing what they learn, while also lifting others up.
Why?
Because AI is a network effect technology. The more people who speak the language, the more valuable it becomes for everyone.
Plumbers and other trades
Their entire ecosystem of professions is predicated on apprenticeships. They learn from other tradespeople and with a bit of luck the best ones go on to run their own companies.
This is what Paulo Freire meant when he said literacy is political. AI is about power, access, and who gets to shape the future.
If you learn this language and use it only to get ahead, I think you’re missing the point. The real opportunity is collective flourishing.
How to Start
The best way to learn a language is to spend time in the host country. So:
I use AI daily. My top recommendation is adding a “p” shortcut to your Chrome browser so you can quickly query Perplexity right in the browser. If you are already on Comet, this doesn’t apply to you.
Build a glossary. Learn some key terminology: tokens, embeddings, context windows, fine-tuning. Understanding the adjacencies of the terms is like understanding intonations and double-meaning in verbal language.
Find your community. Find one or two AI teachers. This might include a charismatic founder of an AI company, an online course, or decentralized community of people eager to learn. Follow people who are building in public. And don’t worry about playing with every new tool (there are too many). Instead focus on people who teach in a way that makes sense to you. Then learn the tools and also the concepts behind them. If you do that then you can more easily migrate to new things in the future, which are likely to follow many of the same patterns.
(Optional)
I engage in ethical thought experiments on a weekly basis. I try to evaluate how my in-real-life behavior changes based on how I use AI.
