Terminal Fear: The 15 Minutes That Could Change Your Workflow
Many of my colleagues and students freeze the moment I open a terminal. The command line looks hard because it's unfamiliar. AI assistants work in text, so avoiding the terminal now cuts you off from the AI tools that run there.
A story has stuck with me for over thirty years.
During my Master’s at the University of Canterbury in New Zealand, my supervisor’s colleague visited from the United States on sabbatical. He was a well-known researcher in the human factors field and was incredibly generous with his time, helping me through the final stages of my thesis. One afternoon we were talking about new technologies in research, and he told me a story about writing his own PhD.
At the time, theses were written on typewriters. During the revision process, when his supervisory committee recommended changes, sometimes significant, sometimes minor, the only option was to retype entire chapters. The whole process, first draft through revisions to submission, took about six months, much of it pure mechanical retyping.
Partway through his write-up, his department acquired several word processors (the old dedicated machines, not software). Someone suggested he try one. He declined. He didn’t have time to learn a “complicated new technology” in the middle of his thesis.
A few days after he finally submitted, he wandered back into the department with some time on his hands and sat down at one of the word processors. Within fifteen to thirty minutes, the air turned blue. He was, let’s say, loudly and colourfully expressing his frustration, having realised that this “complicated technology” would have taken minutes to learn and saved him weeks, if not months, of work. He could have produced a better thesis in less time, if he’d just spent that same fifteen minutes at the start instead of at the end.
I think about this story every time I see someone recoil from a terminal. Humans have been interacting with computers through terminals since the early days of modern computing. The AI-powered command-line tools and coding assistants that run in them are new.
The Flinch
It happens constantly. I open a terminal, or an IDE like VS Code with a terminal in it, and start working, and the reaction from colleagues and students is immediate: That looks too complicated. That’s for programmers.
The terminal looks hard to them because it’s unfamiliar. They’ve spent years, sometimes decades, interacting with computers exclusively through graphical interfaces: clicking buttons, dragging windows, navigating menus. A blinking cursor on a dark background feels like being dropped into a foreign country without a phrasebook.
The terminal is just text. You type a command and you get a result, with no hidden complexity behind it. In many ways, it’s simpler than a GUI. There are no settings buried five clicks deep and no mysterious icons.
AI Assistants Work in Text
In 2026, terminal fear is costly, because AI assistants are text-native.
Large language models (ChatGPT, Claude, Gemini, Copilot) are fundamentally text-in, text-out systems. A command line works the same way.
Chatbots have nice web interfaces. Coding assistants are getting better at operating GUIs, but they’re still pretty terrible at it. When an AI assistant works through a terminal or a code editor, it’s fast and precise, and you can see exactly what it’s doing. You can copy, modify, and rerun its commands.
If you avoid the terminal, you’re cutting yourself off from the most direct way to interact with AI tools. You’re limited to whatever buttons and menus someone else designed for you, instead of being able to tell the computer, through an AI assistant, exactly what you want done.
It Takes Fifteen Minutes
The terminal inside VS Code is a good place to start. It opens as a panel at the bottom of your editor, so you don’t have to find a separate application or face an intimidating full-screen black window. Press Ctrl+` (Control plus the backtick key) on Windows, Mac or Linux to open it, and press the same keys to close it when you’re done.
The commands below are safe to experiment with. Moving between folders, listing what’s in them and running a script won’t damage your system, and the terminal only does what you tell it to. The habit to build is reading what an AI assistant proposes before you approve it, because these assistants can move or delete files when asked. By default they show you each command and wait for your OK before changing anything.
To get started, you need four commands:
cd: move to a folderls(Mac) ordir(Windows): see what’s in a foldermkdir: create a new folderpython script.py: run a Python script
You can learn them in five minutes.
Next, install a CLI tool like Claude Code, Gemini CLI, or OpenAI Codex. These are AI assistants that live in your terminal. Once installed, you can ask them to search the web and summarise what they find, write a Python script to plot your data, draft an email, restructure a folder of files, or explain a concept, all from a single text prompt. You describe what you want in plain English, and the assistant does it.
The chatbot web interfaces and the GUI plugins are wrappers around this same text exchange, usually with less control and less visibility into what’s happening.
The Word Processor Lesson
The colleague from my Master’s years avoided the word processor because it looked hard, and he assumed the learning curve would cost him time he didn’t have. The opposite was true. The lesson he learnt the hard way, and passed on to me: never be afraid to at least try something new. The terminal looks as complicated to many of my colleagues and students as the word processor looked to him.
You don’t need to become a command-line expert. You just need to get past the flinch. Open a terminal, install an AI assistant, and ask it a question.
It takes fifteen minutes.
I’ve written about how 95% of my work now happens in VS Code, a single environment where writing, coding, data analysis, and AI assistants all live side by side. I’ve also written about ditching PowerPoint entirely in favour of coding lecture slides in HTML with AI assistants. My students, with zero coding experience, are building presentation decks in 20-30 minutes using the HTML Slides Starter Kit.
Michael Richardson Professor, School of Psychological Sciences Faculty of Medicine, Health and Human Sciences Macquarie University
AI Disclosure: I wrote this article with help from AI tools, including Claude. The ideas and the anecdote described here are my own. The AI helped me draft and edit the text. I use AI throughout my research and teaching, and I encourage others to do the same. Using AI well is a skill to be learned openly.