Okay, so, what is it?
Right, you know how chatbots… chat? As in, write to you, talk to you, compose your college papers? Claude Code, which comes from the AI company Anthropic, is an AI tool that can actually do things with your computer. Actually, many of the things you can do with your computer. (Well, not you, if you’re the target audience for this FAQ, but someone who is an expert programmer who never sleeps, never says no, and works at an impressive speed.)
Do… like what things?
Honestly, it’d be easier to list the things it can’t do with a computer. But an incomplete rundown of what users have accomplished with Claude Code would include: a Spotify Wrapped program but for text messages; personalized daily briefs that pull in emails, newsletters, and more; a Pokémon card management system; a personal DNA analyzer; and a “cyberpunk” Tetris game. You will need at least a $20 a month Claude Pro account — no freebies for you.
…Cool? But it has “code” in the name — do I have to know something about programming?
No worries! Yes, Claude Code is designed to work in what’s known as a “command-line interface,” or the part of your computer where instead of clicking on icons or writing normal sentences, you type commands with a programming language into a terminal, aka the black screen where nerds are entering their code.
Whoa, whoa, whoa! Do I look like 1995 Angelina Jolie in the movie Hackers? I don’t know what any of that means.
It’s okay — neither do I!
It’s true that experienced programmers can get the most out of Claude Code (though they’re also the ones who are undergoing the deepest existential crises). But the learning curve for using Claude Code is descending faster than a Six Flags roller coaster, and you can increasingly interact with Claude Code more or less as you would with a chatbot if you want — with plain English and relatively few commands. Be warned that it’s clunkier than using it in the terminal, but honestly I wouldn’t trust either of us with that.
Bottom line, the process works like this:
- You tell it what you want (fix a bug or make a new feature).
-
It looks through the project’s codebase — all the files that make the program you’re working on run, including the actual code plus the configuration and test files around it — to understand what’s happening.
- It edits the relevant files.
- It can run tests/commands to see if it broke anything.
- It iterates.
In the best-case scenario, it closes the loop, mostly on its own: plan → change → check → fix. That’s why people who build software for a living are acting like they’ve been freed from a thousand tiny paper cuts.
In the worst case, it can, uh, delete your files, leak your secrets, or burn through your rate limits. Which would be bad.
But I would like to keep my files. Ideally all of them. In their current state of existence.
Smart person. Claude Code is agentic-ish, meaning it can carry out tasks with little to no supervision, and as any manager knows, the benefits of an agent (“it can act autonomously!”) are also the drawbacks of an agent (“oh no, it just acted autonomously!”).
So if you start messing around with it, be sure to be very, very explicit in your directions — like, “Do not delete anything. I really mean this.” (Fortunately, by default Claude Code still taps you on the shoulder before doing anything irreversible.) It’s sort of like parenting a 5-year-old with superpowers.
Also, keep backups of anything important. But obviously you already do that.
Uh, sure… moving on, I understand why this is such a big deal for programmers. But does it really matter for the rest of us?
Sure does! As Future Perfect contributing editor Dylan Matthews wrote last year — borrowing a phrase from AI writer/investor Leopold Aschenbrenner — the scary endgame is “drop-in remote workers.”
Put simply, if you are a remote worker, it likely means you execute most of your tasks on a computer. Like I’m doing right now. And while I may not think of myself as manipulating computer code in my work, under the hood, that’s exactly what’s happening with every letter I press in this document.
Large language models — especially complex reasoning ones like Claude’s Opus 4.5, the preferred model for super-charged Claude Code work — are already very good at thinking, analyzing, and writing, and they’re only likely to improve.
Claude Code is what happens when you take a language model and give it tool access — file editing, searching, running commands — inside your codebase, with guardrails you can loosen (or, regrettably, remove). In other words, if you’re a remote worker, Claude Code could conceivably “drop in” and do some, most, or maybe all of that work. If chatbots could really just advise, models like Claude Code can actually do.
And Anthropic is already trying to port that same “Claude with hands” feeling out of the programmer cave and into the rest of your digital life. That’s the idea behind the just-released Claude Cowork: Instead of pointing Claude at a codebase, you point it at a normal-person folder — your notes, docs, spreadsheets, PDFs, screenshots, the junk drawer of modern work — and it can read, organize, extract, and draft inside that space to produce real deliverables, not just suggestions.
If Claude Code is a drop-in remote worker for software teams, Cowork is the version that can drop into the work most remote workers actually do: turning messy inputs into usable outputs, faster than you can say “sorry, circling back.”
Uh oh.
Yes, now perhaps you understand why Anthropic CEO Dario Amodei warned that we could be “sleepwalking into a white-collar bloodbath,” with AI wiping out huge numbers of entry-level jobs fast.
Have you tried Claude Cowork?
Nope — Cowork currently requires at least a $100 a month Max account, and obviously I have to save up for the post-work apocalypse.
Wait, aren’t you supposed to be the Good News guy?
Indeed I am! (Sign up for the newsletter here.) And if you squint, you can argue that what we’re likely to see is less replacing human jobs than rearranging them, turning workers into managers of teams of future AI agents, responsible for setting goals, checking outputs, and making judgement calls. So I guess in this more optimistic future, we’ll all be Office Space’s Bill Lumbergh, directing our army of AI agents to fill out infinite TPS reports.
O brave new world, that has such agents in it!
Yeah, I think the one thing we can count on is that it’s going to get weird. I mean, weirder.
But in the meantime, unless you’re planning on going around sabotaging data centers — please don’t — you really can meaningfully improve your work and even your life if you begin to play around with these tools. The first time you actually create something that works is a pretty powerful feeling. Like I imagine how Mickey felt halfway through The Sorcerer's Apprentice.
Do… you know how that ended?
And they all lived happily ever after.
// end of input
(Disclosure: Future Perfect is funded in part by the BEMC Foundation, whose major funder was also an early investor in Anthropic; they don’t have any editorial input into our content.)
