jhon, this is gold

Inspired Zona Paisley Pattern Vintage Floral Background Stock Il Tauqeer Jutt New Hand Drawn Floral Buds Combination Stock Illus Seamless Digital Repeat Pattern Abstract Flowers S

shutterstock 2471... shutterstock 2471569853 copy hosted at ImgBB Seamless Digital ... Seamless Digital Repeat Pattern Abstract Flowers S Vicomte Font by B... Vicomte Font by Black[Foundry] New Hand Drawn Fl... New Hand Drawn Floral Buds Combination Stock Illus 9df04a09ce hosted... 9df04a09ce hosted at ImgBB shutterstock 2367... shutterstock 2367814685 hosted at ImgBB

See more

Was this email useful? Yes No

Based on your activity

Help Center  ·  Privacy Policy   ·   Terms & Conditions  ·  Unsubscribe from emails like this   Pinterest. Inc  ·  651 Brannan Street   ·   San Francisco  ·  CA, 94107

Read more......

🍓 Stanford Retests AI’s Famous Flaws

Made-up sources nearly vanish from top chatbots, Apple’s tabletop robot waits on Siri, a haiku sent from orbit, and more...  ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌

November 03, 2026   |   Read online   Hey, Jhon! Welcome back to the #1 AI newsletter in the world! Here’s what we have today: Stanford retests AI’s famous flaws. Apple’s tabletop robot takes shape. An AI agent delivers a mouse’s brain treatment. Google’s satellite sends a haiku back.       AI Myths Stanford Retests AI’s Famous Flaws Source: Stanford University A couple of years ago, somebody bringing up how AI couldn’t count the R’s in “strawberry” got a laugh. People still repeat it like it’s true. Stanford took 50 of the most-cited findings about what AI gets wrong and ran every test again. This time they used 20 models released over the past two years. Almost half of those flaws are gone now, like the strawberry one. Losing track of a fact buried in the middle of a long doc is gone as well. The rest have mostly faded. Made-up sources were the big one for a while. When ChatGPT first took off, more than half the references an early model handed you were fake. In Stanford’s retest, the newest models from OpenAI and Anthropic made up close to none. Some models still make them up. Flaws also didn’t stick around as long as you’d think. Half of the ones with a clear end date disappeared within 15 months. A few went the other way. AI still leans toward agreeing with you. Tell it what you like and it now praises your work even more to match. You can dig through all 50 on their site. Pick a flaw and see how each model did over time. Or pick a model and see which flaws it has left. Our take: Next time someone tells you AI can’t do something, ask when they last checked. A lot of those claims were tested on models that are two years old. Agreeing with you is the one flaw that has stuck around. When you want an honest read on your own idea, make sure to ask your AI to argue against it first.       Together with Attio A CRM That Keeps Working After You Log Off A CRM tends to sit still until you open it and type something in. The lead no one called back and the customer who’s slowly going quiet stay right where they are until someone happens to notice. Attio is a CRM where AI agents keep working on your customers around the clock. It brings in your emails and calls on its own and keeps each customer’s full history in one place. You describe a job once in your own words and their agents run it on every record. One could flag which new sign-ups are worth a call. Open any run to see what the agent thought and every tool it called. Pause it the second something looks off. It works inside ChatGPT and Claude as well. You can ask about any customer from the chat you already use.       Smart Home Apple’s First Robot Is Waiting on Siri Source: MacRumors Apple has a robot in their labs that’s been mostly built for around two years. It’s a screen on a motorized arm that sits on your kitchen counter. The plan was for it to act like another person in the room. It’d jump into a conversation with something useful or talk a trip plan through with you, the way a friend at the table might. Reportedly, the hardware has been ready for a while and the software isn’t. Apple has trimmed the first version down to FaceTime calls and home controls like switching on the lights. The part that makes it feel like a person is now waiting on Siri. The HomeView Apple already announced is the version without the arm. None of the robot is announced yet. Apple is aiming for 2027 and it could slip to 2028. Our take: This robot could end up showing just how far Siri has come. If the first version ships as a FaceTime screen that turns its head, Apple is still playing catch-up. If it can hold up its end of a family conversation, it’s a new kind of device for your home.       Coming in Hot AI Tools of the Day 🌐 WebsitePublisher AI lets you describe the site or app you want to ChatGPT or Claude and publishes it on a live URL, with hosting and payments already handled. 🎯 Alsona spots companies showing signs they’re ready to buy, like a new hire or a tool switch, and has AI agents write LinkedIn and email messages around what tipped them off. 🕺 BeatMV turns any song into a beat-synced music video with every shot planned for you, or makes a photo of your cat or dog dance to the track. 🔎 Programmable builds an agent from one sentence about who you sell to, then finds new matching companies every week and drafts outreach you approve before anything goes out. 🤖 Local Code runs a private AI agent on your Mac that writes code and updates your spreadsheets and slides without an API key or sign-in, free and open source on GitHub.       Lab AI Mouse Gets Its Brain Treatment From an AI Source: Gabriel Ocana-Santero When doctors treat epilepsy, two of the hardest calls are what to use and exactly where in the brain it needs to go. A small biotech startup let an AI agent make both calls for a mouse with seizures. They gave it 19 gene therapies to choose from and a motorized rig with a camera. The agent went with a therapy that turns up the brain’s own calming signal. It’s the one that tells busy nerve cells to settle down. It picked the spot where this kind of seizure starts and read three studies to work out where to aim. A vet still did the opening part of the surgery and approved every step that couldn’t be undone. They could cut power to the rig at any moment. Then the AI guided the needle in on its own and checked the camera 32 times as it went. Claude didn’t jump at the task right away either. The biggest version said no the moment it figured out this was a biology experiment. Another one backed out right before the real surgery. It worried about getting cut off halfway and decided that was worse than not starting. It took a smaller Claude to see it through (the mouse is doing well). Our take: The refusals stand out more than the surgery. An AI that won’t start what it can’t be sure of finishing is the kind you’d want in the room someday.       Interesting AI Glasses That Type What You Mouth Source: Subvocal Talking to your AI out loud works fine at home. On a quiet train or in an open office it gets awkward fast. Subvocal is a pair of glasses that lets you skip the talking part. You mouth the words without making a sound and they show up as text on your computer. There’s no camera involved. Sensors in the frame send radio signals through your face and read how your mouth and tongue move. With some practice it can even pick up words with your lips closed. You do a short setup first so it learns how you speak. After that it types into the apps you already use, ChatGPT and Claude included. You can talk a lot faster than you type, and that’s the speed it’s built for.       Space Google’s Satellite Sends a Haiku From Orbit Source: Google A while back we covered the fridge-size satellite Google sent up to test AI chips in orbit. It’s part of Project Suncatcher, their long-shot idea for running AI on sunlight in space. Now the satellite has answered its first prompt. Google beamed up a request for a haiku about being in space as it passed over a ground station. Google’s open AI model, Gemma, wrote one right there on the chip and sent it back down. Blue marble below Data streams in silent grace Starlight feeds my core Google calls it the first AI response ever from a data-center chip in orbit. It’s their version of a “Hello World,” the tiny first program people run to prove a new setup works. The satellite has since written limericks and reported on how its own systems are holding up. Our take: A poem looks like a party trick. The proof is what matters. A regular data-center chip ran a full AI model above the atmosphere and sent back a usable answer. That wasn’t a sure thing before launch.       Prompt of the Day Stick the Landing Important Note: You’ll need to click the button to get the complete prompt. View and copy today’s full prompt This prompt seats a panel of five Olympic judges to score one routine you run every day, like getting out the door or winding down for bed. You leave with your score out of 10, the element costing you the most points, and one fix set up before tomorrow’s run.       That’s all for Tuesday! Before you go, let an Attio agent work through your customer list overnight and flag the next lead worth a call. You'll see every step it took when you check in the next morning. 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......