An AI's context window is the amount of previous conversation it can remember before forgetting—like the difference between a colleague who's been following your entire project and one who only caught the last five minutes. When the window fills up, the AI drops older messages to make room for new ones, so you may need to reintroduce key information if your conversation gets long.
When you chat with an AI, it reads everything you've written in that conversation—but only up to a certain point. That limit is called a context window, and understanding it is crucial if you rely on AI for memory support or task management.
Think of it like your AI having a desk with limited space. It can see everything currently on the desk (your recent messages and its recent responses), but once you pile on too much paper, the oldest stuff falls off the back. The AI loses track of what you said five conversations ago, or even just an hour ago if you've written a lot.
Why this matters for neurodivergent brains: If you use AI as an external brain—dumping tasks, ideas, or context about your projects—you need to know when the AI stops remembering. Someone with ADHD might tell an AI body double about all their distractions, but if the conversation gets too long, the AI might forget what your main priority was. Someone with dyslexia might paste in a document, discuss edits, but then ask a new question and the AI won't have the original text anymore.
Different AI tools have different context window sizes. Some can see about 4,000 words back (like older ChatGPT), while newer ones like Claude can handle 100,000+ words. This matters when you're building a persistent memory system or trying to maintain focus through a complex project.
The practical reality: Context limits aren't bad—they're just a tool you need to work with. If you know your AI has a 4,000-word window, you can be intentional about summarizing important details at the start of each conversation. You can use tools like Notion AI to store your ongoing context in one place, then paste relevant summaries into your AI conversations. This way, your AI doesn't lose track of what matters to you.
People often think a bigger context window solves everything. Actually, it's more useful to understand your window size and work with it deliberately. A smaller window forces you to be clear about priorities—which is exactly what many neurodivergent learners need anyway.
Try this: Start a conversation with your AI and ask it directly: "How much of our conversation can you see right now?" Then test it—go back several messages and ask the AI to summarize what you discussed early on. You'll get a feel for your tool's actual limit, and you'll know when to re-paste important context.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.