The problem

Every new chat with an AI can feel like meeting a stranger. You explain your project, your preferences, your tone of voice — and the moment you close the tab, it's gone. Next time, you start from zero again.

<aside> 🚫

That's how most people assume Claude works too — a blank slate every single time, even on the free plan.

</aside>

The solution

Claude can actually remember context across separate chats — things you've told it before, preferences you've set, ongoing projects — without you re-explaining everything each time. You stay in control: you can see what's remembered and remove anything you don't want kept.

This is different from a single long conversation. It's context that carries between chats, days apart.

How it looks on a tablet

1. Open Claude and just talk normally — mention a preference, a project, a recurring detail
2. In a later, brand-new chat, refer back to it without repeating yourself
3. Check your settings any time to see what's been remembered, and edit or clear it

<aside> 💡

Example: Tell Claude once that you're building a Bluesky content brand in a specific tone. Weeks later, open a fresh chat and ask for a post — Claude already knows the tone, no re-briefing needed.

</aside>

The trick

Treat memory like a running notes file, not magic. If Claude gets something wrong or outdated, just say so — you can correct or delete what it remembers at any time.

The takeaway

Memory across chats isn't locked behind Pro. It's part of the free plan — so the start-from-zero-every-time assumption is simply outdated.

🔗 Next in the series: Artifacts — how Claude builds working mini-apps, not just text