Claude’s new memory feature solves my biggest AI frustration — here are 5 prompts to try first
[post_content]
Disclaimer: This article has been automatically aggregated from
No matter how much I use ChatGPT, Claude or Gemini, there are still times when I feel like I’m introducing myself all over again. Each chatbot will often throw something out that makes me question, “Do you even know me?” It’s frustrating to constantly explain and then re-explain how I like to work, what I’m trying to accomplish or some other piece of context that I’m fairly certain I’ve mentioned before.
Unlike static custom instructions that you have to write manually, Anthropic’s latest update gives the chatbot the opportunity to learn about you as you go.
Claude’s memory now works across regular Chat and Claude Cowork, meaning something you tell Claude in a regular chat can follow you into a Cowork project. Claude can also update what it remembers while you’re still chatting, rather than waiting until a conversation ends.
Latest Videos FromTom’s Guide
Manage settings to organize memories
You can actually manage what Claude knows about you and remembers. Head to Settings > Memory and Claude organizes those memories by topic, where you can edit or delete individual entries.
But better memory is only useful if Claude actually puts it to work. These five prompts are designed to find out what it remembers, whether that information can save you from repeating yourself and when it might be time to make Claude forget.
1. Find out what Claude already knows about you
Prompt: Based only on your saved memory about me, tell me the 5 most useful things you know about how I work, what I prefer and what matters to me. Don’t use information from this conversation unless it’s already in memory.
This prompt is useful because before adding anything new, you can find out what Claude has already picked up.
This prompt is slightly different from asking a generic “What do you remember about me?” You’re asking Claude to separate the things that could actually help it answer you from the random facts it happens to remember.
Getting more specific helps Claude understand things such as the way you prefer replies, that you regularly work on certain kinds of projects or even that you prefer to tackle longer tasks in the morning could actually change the answer Claude gives you.
It’s also an easy way to spot anything Claude has gotten wrong. If you find something inaccurate, you can head into Memory settings and edit or delete it before it influences future conversations.
2. Ask Claude what you’re tired of repeating
Prompt: Look at what you remember about me and identify 3 pieces of context I probably shouldn’t need to explain to you again. For each one, show me an example of how you could use that information in a future conversation without me repeating it.
A memory feature shouldn’t just be a digital notebook filled with facts about you. The real benefit comes when the AI knows when to use them.
If Claude knows how you like emails to sound, for example, you shouldn’t have to explain your preferred tone every time you ask it to draft one. If it knows the tools you regularly use, it could take those into account when suggesting how to tackle a task.
3. Put its memory to work on something new
Prompt: I’m starting a new project. Before I give you any other instructions, use what you already remember about my preferences and the way I work to suggest how we should approach it. Separate what you’re basing on memory from anything you’re assuming.
This goes beyond asking Claude to recite what it knows about you. Now you’re finding out whether knowing you actually makes its answer better.
AI assistants are very good at filling in the blanks, which can make it difficult to tell what’s coming from memory and what’s simply an assumption. Asking Claude to separate the two lets you see whether its new memory is actually influencing the answer.
Once Claude establishes that baseline, give it the actual project, such as planning a vacation, organizing a home project or starting something new at work, from there, you’ll probably find the answers are much more useful.
4. Let Claude tell you what it should learn
Prompt: Based on the kinds of things I ask you to help with, what are 5 things about my preferences, routines or goals that would make you more useful to me if you remembered them? Ask me one question at a time to fill in the gaps.
Claude can learn as you go along, so now, instead of trying to guess everything it needs to know about you, let Claude identify what’s missing.
It might know what kind of work you do but not how you like answers formatted. Maybe you regularly ask it for recipes but it doesn’t know what foods you dislike. Or perhaps you use Claude for travel planning but it has no idea whether you’d rather stay at a luxury hotel or save that money for something else.
Since Claude can now update its memory while you’re chatting, some of the answers you give it can become information it remembers for future conversations.
There is an important privacy consideration here. Anthropic says Claude won’t save certain sensitive information by default, including details related to health, religious beliefs and political views. You can choose whether to allow sensitive topics in memory through Claude’s settings, so it’s worth checking what you’re comfortable with before deliberately teaching it more about yourself.
5. Give Claude’s memory an audit
Prompt: Audit what you currently remember about me. Put each memory into one of four categories: useful, outdated, unclear or unnecessary. Explain briefly why, and tell me which memories you think I should review or delete.
The more an AI remembers about you, the more important it becomes to occasionally check whether it should still remember all of it. Chances are, you really don’t need it to remember everything.
For example, that project you finished three months ago might no longer matter. In fact, by giving the memories an audit, you may even find something that Claude could be remembering wrong or useful in the first place.
That’s why I like this prompt so much. Instead of scrolling through every memory and deciding what stays, you’re asking Claude to do the first pass for you. You’re still in charge of the final call, but dive into Claude’s Memory settings and review individual topics, then edit or delete anything you no longer want it using.
The takeaway
The more you use AI, the more you’ll realize how important Memory is to stay productive. Now that Claude can learn as you chat, the AI assistant is considerably less annoying to use.
Give it a try by using these prompts or talking through a project with Claude in Chat and then move into Cowork without starting the explanation all over again. Just remember that Cowork still requires a paid Claude plan, while the broader memory feature is available across Free, Pro and Max accounts. Bottom line: You might be surprised by what it already remembers.
Follow Tom’s Guide on Google News and add us as a preferred source to get our up-to-date news, analysis, and reviews in your feeds. Subscribe to Tom’s Guide on YouTube and follow us on TikTok.
More from Tom’s Guide
for informational purposes only. We do not claim ownership, accuracy, or liability for the content provided. All rights belong to the original publisher.