The Big Story

A Reddit user posted a full set of 30 ready-made prompts designed to make ChatGPT generate images that draw directly from everything the chatbot has already learned about you through past conversations. The idea is simple: instead of writing a long description of yourself from scratch, you tell the chatbot to pull details from its memory of your chats, including hobbies, personality quirks, recurring topics, pets, work, and inside jokes. Think of it like giving an artist a thick notebook filled with everything you've shared over months and asking them to create one picture that captures the real you rather than a generic version. The prompts produce results such as an illustrated cutaway view called "Inside My Brain" with separate rooms for your work, hobbies, and running jokes, or a video-game character select screen that shows different sides of your personality as playable characters. Other examples include turning your life into a board game with spaces and rules based on your actual habits, or creating a personal coat of arms filled with symbols that only someone who knows you would recognize. This approach stands out because most AI image tools default to broad stereotypes that could fit anyone, while these prompts aim for results that feel specific and personal. For students and creative teens, the prompts open a new way to explore self-expression, create shareable social media content, or brainstorm visual ideas for school projects about identity and memory. You might notice the AI pulling in small details you had forgotten mentioning, which can feel both surprising and revealing about how much context these tools retain across conversations. The prompts work best when you have already used ChatGPT for a while and have Memory enabled so the chatbot has a richer set of details to reference. To get started, go to chatgpt.com on your phone or laptop, copy one of the 30 prompts from the original post, paste it in, and add the instruction to use everything it already knows about you from past conversations and memory. If you want the person in the image to resemble you, upload a reference photo first and include a note asking the tool to preserve your recognizable facial features. Source: reddit.com

Explain Like I'm 14

You know how when you text the same friend every day, they start remembering little things without you having to remind them every time — like your favorite snack or that you hate early mornings? AI chatbots do something similar, but instead of a brain they keep a running record of everything you've typed in your conversations. Each time you send a message, the system stores the key details in a kind of ongoing notebook it can look back at later. When you ask it to make something personal, it flips through that notebook and pulls out the details that match what you're requesting right now. The more you chat over weeks or months, the thicker that notebook gets, which is why the images feel more accurate after you've used the tool for a while. It's not magic or mind-reading — it's just the chatbot treating your past messages like a helpful reference sheet it keeps handy. Once you see it that way, the "wow, it actually knows me" feeling makes a lot more sense.

Cool Stuff & Try This

Watch out for "shadow AI" at work or school A recent warning points out the risks of using personal versions of tools like ChatGPT for tasks involving company or school data without official approval. When you type sensitive information into a personal account, that data can be stored or processed by the AI company in ways your employer or school did not agree to. Students with part-time jobs or internships should check their organization's rules first before pasting any internal documents, customer details, or private project notes into a free chatbot. The concern is often called "shadow AI" because people use the tools quietly without realizing they might be creating security or privacy problems. You can avoid the issue by keeping work and personal chats completely separate and asking your teacher or boss what AI tools are approved for school or job tasks. If you're unsure, start by using the chatbot only for general brainstorming that doesn't involve any real names, numbers, or private files. Source: digitaltoday.co.kr

Quick Bits

Court blocks Pentagon attempt to blacklist Anthropic A judge ruled that the Pentagon's earlier decision to blacklist the AI company Anthropic was illegal and based on weak reasons. The case shows how governments and AI labs are still figuring out their relationship. It's a reminder that rules around AI are still being written in real time. Source: engadget.com

💬 Reply to this email — Patrick reads every one.

Share: X · LinkedIn · WhatsApp

Forwarded this email? Subscribe here — it's free.