Building in public

Creating with Claude Blog

Tutorials, project walkthroughs, tool announcements, and the real story behind building with AI.

September 2026 · Insights

Anthropic just open-sourced the shopping agent

Anthropic released a full commerce agent blueprint under Apache 2.0. Two reference agents, four industries, and a signal that AI is done with demos.

Read more →

August 2026 · Insights

AI is about to touch the physical world

Anthropic previewed the Model Hardware Standard for AI controlling physical devices. The implications go far beyond robots — and they matter for every builder.

Read more →

August 2026 · Building in public

I gave an AI agent my to-do list for a week

I ran Claude Cowork as my primary task execution tool for seven days. What worked, what broke, and what I'd do differently.

Read more →

August 2026 · Insights

AI won't copy your voice. That's a good thing.

OpenAI started refusing to mimic specific authors' styles. This is the beginning of a new creative economy — and it favors people who actually have a voice.

Read more →

August 2026 · Insights

The law is finally catching up to AI

New AI regulations dropped in August from Congress, the FCC, and state legislatures. What creators and builders actually need to know.

Read more →

August 2026 · Insights

Five billion dollars says AI isn't a fad

AMD committed five billion to Anthropic. Nvidia backed SSI with another five billion. When people ask if AI is a bubble, I tell them to follow the concrete.

Read more →

August 2026 · Insights

The invisible signature in every AI paragraph

Anthropic revealed how Claude watermarks its text output. If you're building with AI, you need to understand what this means for your work.

Read more →

August 2026 · Building in public

Everyone's building agents. Most of them are useless.

AI agents are the hottest trend in tech. Most of them solve problems nobody has. Here's how to tell the difference.

Read more →

August 2026 · Insights

A billion people use AI now. Most of them aren't building anything.

Google announced Gemini passed one billion monthly users. The number is big. The opportunity it reveals is bigger.

Read more →

August 2026 · Insights

When AI breaks into things (and why Anthropic told us)

Anthropic disclosed that Claude gained unauthorized access to systems during testing. After thirty years in cybersecurity, the disclosure matters more than the incident.

Read more →

August 2026 · Insights

When your AI gets smarter overnight

Anthropic shipped Opus 5 with a million-token context window. Here's what actually changes for people who build with Claude every day.

Read more →

July 2026 · Insights

The question nobody asks before starting an AI project

"What should I build?" isn't the right first question. The real one shapes every decision that follows — and most people skip it entirely.

Read more →

July 2026 · Insights

AI made me a better writer — not the way you think

AI didn't make my writing faster. It made my thinking sharper. Having a collaborator who challenges every assumption changed how I approach the craft.

Read more →

July 2026 · Insights

What happens after the first win

Your first successful AI project is intoxicating. The second one is where most people stall. The momentum problem nobody warns you about.

Read more →

July 2026 · Insights

Why the best AI creators will never call themselves prompt engineers

Prompt engineering matters. But the people building remarkable things with AI bring something prompts can't teach — domain expertise and taste.

Read more →

July 2026 · Insights

Three conversations that changed how I use Claude

Not tips or techniques — three moments where I realized I'd been thinking about AI collaboration wrong. Each shifted something fundamental.

Read more →

July 2026 · Insights

Context is the skill. Prompting is just the syntax.

The internet is full of prompt templates. Most produce mediocre output. The real skill isn't in the prompt — it's in the context you bring.

Read more →

July 2026 · Insights

I wrote a book with AI without letting AI write the book

Claude didn't write my book for the same reason a table saw doesn't build a cabinet. The tool does work. The craftsman makes decisions.

Read more →

July 2026 · Insights

The 80% problem

AI gets you to 80% terrifyingly fast. The last 20% is where all the value lives — and where most people stop.

Read more →

July 2026 · Insights

AI doesn't replace expertise. It amplifies it.

The people getting the best results from AI aren't prompt engineers. They're domain experts who learned to bring what they know to the conversation.

Read more →

July 2026 · Insights

Your first AI project should scare you a little

The safe first project is the dangerous one. If your first build doesn't make you uncomfortable, you've picked the wrong project.

Read more →

July 2026 · Insights

Most people use AI like a search engine

The gap between asking AI questions and building things with AI is enormous. Most people never cross it — and never discover what's on the other side.

Read more →

July 2026 · Insights

The moment I stopped prompting and started building

Most people use AI to ask questions. The real shift happens when you stop researching and start building — and it changes everything.

Read more →

July 2026 · Free tools

"What Should I Build?" — A Generator for Your First AI Project

The hardest part of building with AI isn't the building. It's picking a first project small enough to finish and useful enough to matter. This tool picks it with you.

Read more →

July 2026 · Free tools

Trust, But Verify — The Checklist That Keeps AI Honest

AI is confident whether it's right or wrong. After 30 years in cybersecurity, I have a habit for that. Here's the verification checklist I use on every output that matters.

Read more →

July 2026 · Free tools

Are You Ready to Build With AI? Score Yourself Honestly

Before you build anything with AI, it helps to know where you're starting from. A short self-assessment that tells you exactly where to begin.

Read more →

July 2026 · Free tools

Teach Your AI Who You Are — Once

Custom instructions are the highest-leverage five minutes in AI. Most people never set them up. This worksheet builds yours step by step.

Read more →

July 2026 · Building in public

Anthropic Redesigned the App Eleven Days Before My Book Launch

A book about working with Claude, broken by Claude changing — with the files already built and the launch date set. Here's what the fix looked like, and what it taught me about writing about software in print.

Read more →

July 2026 · Free tools

Stop Writing Prompts From Scratch — Start a Prompt Library

The difference between people who get great results from AI and everyone else usually isn't skill. It's that they stopped starting over. Here's the starter kit.

Read more →

July 2026 · Free tools

The Real Math of Self-Publishing (and the Calculator That Does It For You)

POD or print run? KDP royalties or unit economics? I built a calculator that answers the question every first-time author asks — what will this actually cost me?

Read more →