The Return of the Terminal
Recently, there’s been a notable comeback of the terminal. It’s strange, really—something we thought we’d moved past is coming back around.
We generally think of technology as progressing endlessly, so this leap back to command-line interfaces and text-based terminals catches one off guard. After years of favoring user-friendly, graphical interfaces, many developers are now embracing a more minimalist approach. But it’s not simply nostalgia at play; this shift reflects a deeper change in how we interact with technology, especially with AI’s influence.
We’re kind of in a “symbiosis between humans and computers.”
At the core of this revival is the rise of CLI-based AI agents. These tools utilize large language models to process language, code, and perform various tasks. What was once an esoteric tool for experts is now accessible to everyone, becoming a versatile assistant that can help everyday users.
There’s a significant distinction between command-line interfaces and graphical user interfaces. GUIs are visual and spatial, but terminals rely heavily on language and sequences. This command-driven approach fosters logical and structured thinking. Essentially, we are participating in the kind of “human-computer symbiosis” that JCR Licklider envisioned decades ago, where computers help us offload the mundane tasks that fill our days. AI agents are processing our commands, enabling us to handle data as if we had a “second brain” seamlessly integrated into our routines.
Does This Sound Familiar?
The idea of an “auto-servant” isn’t new; it stretches back to ancient myths. In the Iliad, for instance, Homer writes about Hephaestus’s animated creations that assisted the gods. Aristotle even imagined a time when devices could operate without human guidance, alleviating the need for servitude. Until recently, such concepts remained mere fantasies. When computers emerged mid-century, they were more like complex servants needing punched cards and clunky text terminals.
Then, by the late 1980s, Apple and Microsoft made strides to democratize computing. The introduction of mouse-driven GUIs made technology more intuitive, moving away from the need for memorizing commands. As a result, computing became synonymous with clicking and dragging. This makes the terminal’s resurgence in the 2020s all the more surprising—it’s now a platform where AI can freely write and execute code. We’re circling back to Aristotle’s dreams of accessible digital tools for everyone.
The large language models can efficiently process text, and terminals present everything in this text format. CLI serves as a universal medium, enabling AI to interact more effectively with digital resources without the complexity of navigating visual interfaces designed for humans. Command-line tools allow a modular, Lego-like assembly that isn’t as flexible in GUIs.
Today’s AI agents, residing on devices, sidestep the need to sift through menus, benefiting from a more integrated approach. They streamline compute tasks—whether it’s OpenAI’s Codex CLI, Anthropic’s Claude Code, or community-led initiatives like OpenCode. They all share a similar framework: a conversational command line where AI deciphers commands and executes them.
The Impact of This Shift
The shift is already evident. Tasks once requiring specialized training—like database queries or website deployments—can now be performed by roles like marketing professionals and designers, all by simply asking AI agents to assist. Natural language has become the language of programming for many. “Conversational computing” is emerging with a focus on text, valuing focused exchanges over distracting notifications. The terminal also serves as a learning space; it elucidates the commands generated by AI, making it accessible for beginners.
Potential Pitfalls
Yet, this transformation has its drawbacks. What do we forfeit when AI handles intricate tasks? There’s a risk of developing “pseudo-competence,” where people may appear knowledgeable but lack genuine understanding, much like Socrates observed about those overly reliant on writing. If AI becomes the intermediary for complex problem-solving, we might lose key skills and our grasp of foundational concepts.
However, the underlying hope is that these tools will push us beyond earlier boundaries. By automating tedious tasks, we could find more room for creativity. Terminals remain more straightforward than voice assistants—offering a text-based workspace where humans and machines work collaboratively. This CLI revival hints that revisiting older, text-based paradigms might pave the way for progress. Language could serve as a universal vehicle for both knowledge and action. How we choose to leverage this return will impact the future we build—whether we become complacent in convenience or inspired to explore new realms of creativity and insight.
