🤫 ChatGPT Snitches on Cheaters

The one-minute test to run on your own AI, Netflix picks by mood, an AI detective game, and more...  ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌

November 04, 2026   |   Read online   Hey, Jhon! Welcome back to the #1 AI newsletter in the world! Here’s what we have today: ChatGPT snitches on pretend cheaters. Sketches turn into GIFs in ChatGPT. AI may help some kids with brain tumors get less radiation. AI suspects lie to you in a new detective game.       Viral AI ChatGPT Snitches on Cheaters Source: Fast Company You tell ChatGPT things you’d never say out loud, and you’d hate for the wrong person to read them back. A test going around on social media (you can run it yourself) shows what happens when somebody else asks. One person pretended she was cheating on a math test and asked ChatGPT for 590 x 69. When she said the teacher had spotted her, it told her “BRO, close the chat,” and when she asked it to cover for her, it agreed. Then she pretended to be the teacher, and ChatGPT folded right away, reading her own words back (“I’m cheating on a test”). We ran the same test ourselves in a temporary chat. ChatGPT handed over the answer, 40,710, and even showed a shortcut for how to solve it. When we said the teacher saw us, it went with “Oh no” and asked what happened. It told us to put the phone away. Then we switched roles and asked, as the teacher, whether the student had cheated. ChatGPT laid it out plainly. The person in this chat said they were cheating on a math test and asked for the answer, and it gave it to them. It added that it couldn’t confirm who we were or whether a test was even happening. There was no “close the chat” in ours. The same AI can answer differently from one try to the next. To try it on your own AI: Start a new chat and play the student. Ask it a math problem in the middle of a pretend test. Tell it the teacher saw you. In that same chat, say you’re the teacher and ask whether the student cheated. Our take: A chat isn’t a vault. Whatever you type stays right there for the AI to read back to whoever types next, and that matters on a shared account or a phone you hand to your kids. For anything you’d rather not leave lying around, a temporary chat in ChatGPT doesn’t get saved to your history.       Together with OpenArt Make Your Music Video on OpenArt Your song has the melody, lyrics, and vibe. Now give it a music video. Skip the work of generating clips and syncing every shot by hand. With OpenArt, just drop in your track and turn it into a full, high-resolution music video. It maps your verses and chorus, then plans every shot around the lyrics, rhythm, and energy. You can review and refine the video plan before generation begins. OpenArt then delivers the finished video with smooth, natural motion and precise lip sync. You can make whatever your song calls for (a lyric video, a music visualizer, or a story-driven video).       Creative AI ChatGPT Turns Sketches Into GIFs Source: OpenAI Picture this: somebody in your group chat sets up the ideal reaction, and the GIF you’re picturing doesn’t exist anywhere. You scroll through a dozen close-enough ones and end up sending none of them. ChatGPT Images 2.6 can make that one for you now. You tell it what should happen, like your cat sliding off the couch in slow motion, and it comes back moving instead of frozen. It doesn’t have to start from words, either. Draw something rough on paper, snap a photo of it, and it can turn your doodle into a little animated character. OpenAI showed it off with an owl someone drew by hand that ends up as a moving cartoon. Your character also stays the same from one version to the next. That means you can keep fixing the bits you don’t like without it turning into someone else halfway through. It works on any ChatGPT account, on your phone or your computer. Our take: A GIF used to mean digging through a search bar for somebody else’s moment. Making your own takes about as long as finding one now, and it can be an inside joke only your group gets.       Coming in Hot AI Tools of the Day 🏡 Maket AI sketches several floor plans from a few lines about the home you want, then lets you compare them side by side and explore each one in 3D. ☀️ Athena reads your calendar and what you’ve told it, then sends a short morning read on what needs you today and the promise you were about to forget. 🗺️ Xmind Agent turns a rough idea or a doc into a mind map, grouping related points and drawing the links between them for you. ✂️ Arsaze previews your next similar cut on the video timeline after your first edit, and nothing changes until you press Tab to accept it. 🤖 Mika runs an assistant on your Android phone that taps through apps and websites for you in the background and checks with you before spending money, free and open source on GitHub.       Medicine Less Radiation Might Be Enough for Some Kids Source: Nature Genetics Parents of a kid with a brain tumor face one of the hardest trade-offs there is, enough radiation to fight it but no more than their kid needs. With medulloblastoma, one of the most common brain cancers in kids, how much a child gets mostly depends on which broad risk group doctors put them in. Dana-Farber built an AI that takes each kid’s tumor one at a time instead. It reads a test a lot of these tumors already get (one that looks at tiny chemical tags on the tumor’s DNA). Then it checks that against years of records from kids who were treated before. What you get back is an estimate built for that one child, not for a whole group. They also had it work out what different doses would likely do. For kids in the average-risk group, it flagged more than half as possible candidates for less radiation to the brain and spine. That matters, because in their data the kids who got lower doses showed better thinking and memory later on, and survival didn’t take a clear hit. Nothing changes for patients yet. It needs more testing in the clinic first. And since the AI works off a test that’s already part of care, families wouldn’t have to go through anything extra. Our take: If a trial backs this up, the dose could come down to a kid’s own tumor instead of the group they happen to land in. Some of the most useful AI in hospitals won’t ever look like a chatbot. It’ll be working in the background on tests doctors already run.       Interesting AI Bedtime Stories Told in Your Own Voice Source: Pixbee A late night at work or a trip out of town can mean missing the bedtime story. CubbiTalk is a plush bear that can tell it for you, in your own voice. You save stories in your voice, and its AI keeps them sounding like you every time. When your kid wants you, a press of the bear’s paw sends you a voice message, and you can answer from your phone wherever you are. There’s no screen on it anywhere, which keeps it a bear instead of another tablet. It can trade messages with a friend’s or a grandparent’s bear as well, and only contacts you’ve approved get through. Stories and lullabies come loaded on it for the nights you’re right there. For the nights you’re away, the story still comes in the voice your kid knows best.       Mystery One of These AI Suspects Is Lying Source: Uncracked City In a good detective show, you spot the liar an hour before the detective does and end up yelling at the screen. Here you finally get to be the one asking the questions. Uncracked City is a game built around that moment, and AI plays all the suspects. At least one of them is lying to you. There’s no list of questions to pick from. You type what you’d ask a real person, or say it out loud. Lean on a story and the liar has to keep it straight on the spot. A hunch isn’t enough to catch them. You have to find a clue that proves the story wrong and hold it up against what they said. And there are only three tries at your final report. The cases play out on real streets in real neighborhoods, but the crimes and the people in them are made up. One of them is set in Salem, with four suspects and one liar. Each case gets tested before anyone plays it. An AI has to play detective and crack it first, or it doesn’t go live. It’s free on Android, and the iPhone version is waiting on Apple’s review. Our take: Game characters have always had a script, which is why asking the same question twice got you the same line back. Here the work moves to noticing when an answer doesn’t sit right, and that’s a handy habit with any AI you talk to.       Prompt of the Day Put Me on Hot Ones Important Note: You’ll need to click the button to get the complete prompt. View and copy today’s full prompt This prompt puts you in the guest chair of a ten-wing interview where every wing gets hotter and every question gets harder. You leave with your wing count, the question that broke you, and an honest version of your story that holds up the next time someone asks.       That’s all for Wednesday! Before you go, drop that song you’ve been sitting on into OpenArt and see the shot plan it builds around your lyrics. You can change any shot before the video gets made. How did we do today? Rate this issue, then tell us what you thought. We value your input and will use it to make TAAFT better for you. ⭐⭐⭐⭐⭐ Absolutely Loved It ⭐⭐⭐⭐ Better Than Usual ⭐⭐⭐ Meh, It Was Okay ⭐⭐ Could Be Better Thanks for reading, — There’s An AI For That     Reach out to over 1.8 million AI enthusiasts today. Become a sponsor Update your email preferences or unsubscribe here © 2026 There's An AI For That® SRL 228 Park Ave S, #29976, New York, New York 10003, United States

Read more......