Why Executives Should Use Claude Code

Sam Gaddis interviewed by Claude Code

---

Let's start with the obvious question: what even is Claude Code?

Claude Code is what's called a CLI tool — command line interface. These are programs that run in your terminal. On Mac, that's Terminal. On Windows, it's PowerShell or Command Prompt. If you ever used the old DOS interfaces, same idea. Text in, text out, things happen on your machine.

The critical difference from browser-based AI is scope. When you use Claude in the web interface, or ChatGPT, you're in a sandbox. You drop a file in, it stays in the browser. Claude Code breaks out of that box. It can see files on your computer. It can take actions. Launch applications, modify Word documents, read Excel spreadsheets, check how much space is on your hard drive, examine how many devices are on your wireless network. And — this is the important part — it can use other CLI tools.

How do you actually install it?

It's surprisingly easy. On Mac or Linux, you open Terminal and run: curl -fsSL https://claude.ai/install.sh | bash. On Windows, open PowerShell and run: irm https://claude.ai/install.ps1 | iex. That's it. Then you authenticate with your Anthropic account — either a Claude Max subscription or API credits — and you're ready to go. Type claude in your terminal and start talking to it.

I've seen that Claude Code is now available in the browser too. Why bother with the terminal version?

They recently added Claude Code to the web interface, which is great for accessibility. But here's the thing — every reason I like Claude Code is a reason the web version falls short. The whole point is that it can touch your local file system, use other CLI tools, interact with your actual machine. The web version is sandboxed. It's Claude Code without the code part, if that makes sense. You're back in the browser, back in the box. If you want the real capability — launching scripts, modifying local files, piping output from one tool to another — you need the terminal version.

What does that mean in practice?

Say you want to set up a hosting environment in AWS. There's an AWS CLI. You can tell Claude Code, in plain English, what you want — or ask it questions to figure out what you want — and it will guide you through the entire process. An executive could theoretically spin up cloud infrastructure, something that was previously restricted to IT. I'm not actually encouraging executives to do that, but it's important to understand the capability ceiling has moved.

You keep emphasizing that this isn't just for coding. Walk me through how you actually use it.

One of my favorite workflows is creating single-page web applications — just HTML files with animations, layouts, visual depth. I use those for proposals, strategy documents, sharing concepts with clients. It's not a PDF sitting in someone's inbox. It's an interactive artifact they can click through.

But honestly, you can use Claude Code the same way you'd use ChatGPT or Claude in the browser. Ask any question. Get a response. The difference is that when it creates something, that thing can live anywhere on your machine. You just tell it where you want it.

You've talked about using voice memos as a starting point. Explain that.

This is how I write most of my articles. You open Voice Memos on your phone, talk for a few minutes — extemporaneously, without worrying about structure or coherence — and then grab the transcript. On iOS it's built in now. Drop that transcript into Claude Code, and now you have a seed. Raw material to work with.

So you could describe your 2026 strategy while you're out for a walk. Stream of consciousness. Bring it back to Claude Code and have it create either a strategy document or a strategy website. Whatever format serves the idea.

What else can it do that people might not realize?

Create local databases. Convert files between formats. Generate charts. Wrangle Excel spreadsheets — and I mean wrangle, like cleaning up messy data, not just reading it. Run analysis on your inventory. Connect to external systems using APIs. If you have even a modicum of nerdiness, this is incredibly fun. Build a family calendaring system. Create a website for your kid's soccer team. Ask questions when you don't know how to do something, and it will guide you.

What's the learning curve like?

The hardest part isn't the tool itself — it's recognizing how much of your daily work it can apply to. You have to build the habit of asking: how can I use Claude Code for this? The key skill is actually pretty mundane. You need to be able to copy a file path so you can tell Claude Code where to look. On Mac, you right-click a file, hold Option, and you'll see 'Copy as Pathname.' That's it. That's the unlock.

What about plugins and customization?

It goes deep. There are plugins for Claude Code, slash commands that let you automate repetitive workflows. I have a slash command for generating article images — it uses my standard format, I describe what I want, and it creates the image and drops it in the right folder for this magazine. One command, done. But you don't have to go that far to get value.

What's the cost?

The max package is $200 a month. If you can't figure out how to create $200 a month of value from this, you're not trying very hard. And — not that this should be the reason — but you will command a lot more respect from your dev team and technical people if you're using tools like Claude Code instead of just cranking out stuff in ChatGPT.

You coined the phrase 'vibe strategy.' What do you mean by that?

Everyone's heard of vibe coding — using AI to write code based on vibes rather than specifications. I'm saying the same approach applies to strategy work. You don't need to be a developer to benefit from these tools. You can vibe your way through creating presentations, analyzing data, building prototypes of ideas. The underlying capability is the same. The applications are whatever you need them to be.

So: vibe strategy instead of vibe coding. Use Claude Code to think out loud, create artifacts, and iterate on ideas. The tool is ready. The question is whether you're ready to use it.