In partnership with
Welcome Automaters 👋
So guess what? Dr. Google got malpractice’d
We all knew it was coming. When the big AI firms started playing "Doctor" with medical queries, it was only a matter of time before things got messy. Well, y’all, the bill just came due.
We just found that Google quietly nuked its AI Overviews from medical searches after a massive investigation by The Guardian caught it serving up dangerously misleading info. The main offender? Liver function tests. The AI was spitting out "normal range" numbers while completely ignoring minor details like age, sex, ethnicity, or nationality.
In the medical world, those aren't just details; they are the whole picture. Without that context, a seriously sick person could look at a "normal" AI summary and think they are totally fine when they actually need an ER visit. Not great, Bob!
Now, Google didn't just tweak the results; they scorched the earth. AI Overviews for queries like "what is the normal range for liver blood tests" have been completely removed. But here is the part that will make your brain melt: Google’s own team of clinicians reviewed the flagged queries and claimed that "in many instances, the information was not inaccurate."
If you ask me, that’s some world-class corporate gymnastics right there. "Not inaccurate" is a far cry from "safe for a human being to bet their life on."
But hey, If you think this setback is slowing anyone down, think again. The timing here is pure comedy:
OpenAI just dropped ChatGPT Health last Wednesday. They are partnering with b.well so users can securely connect their actual medical records and Apple Health data to the bot.
Anthropic (not wanting to be left out of the party) released its own healthcare features for Claude just days later.
The Pitch: Both companies are promising that their tools are for "understanding" and "navigation," not diagnosis.
The industry is learning the hard way that healthcare isn't just another dataset to optimize. It is a high-stakes minefield where "hallucinations" aren't just funny quirks; they are potential lawsuits. One health expert warned that Google is just "nit-picking" single search results instead of tackling the systemic lack of nuance in AI medical summaries.
For now, maybe don't fire your actual doctor for a chatbot just yet.
Here's what we have for you today
⚕️Anthropic Launches Claude for Healthcare as AI’s Medical Push Accelerates
If you thought the AI arms race was moving fast, buckle up cuz while we were all distracted by the Google medical fail, Anthropic decided to crash OpenAI’s healthcare party.
Just days after OpenAI dropped ChatGPT Health, Anthropic rolled up to the J.P. Morgan Healthcare Conference and said, "Our turn." On Sunday, they launched Claude for Healthcare. Coincidence? Please. This is a full-on sprint for the $4 trillion healthcare market, and Anthropic is playing for keeps.
So what’s in the Bag?
This isn't just another chatbot. This is a massive, full-stack play for the entire medical system. Here is the breakdown:
Your personal health concierge (beta): If you’re a Claude Pro or Max user in the U.S., you can now link your Apple Health and Android Health Connect data. And thanks to a partnership with HealthEx, Claude can even pull in your actual medical records. The pitch: Claude becomes your health sidekick, summarizing your history, translating scary lab results, and helping you show up to doctor visits prepared instead of confused.
The "Paperwork" Killer: Anthropic is going after the absolute worst part of medicine: the bureaucracy. Claude now taps into the Medicare & Medicaid Services Coverage Database, ICD-10 codes, and the National Provider Identifier Registry.
Prior authorizations, but faster: You know those nightmarish insurance requests that take weeks? Claude can cross-reference clinical guidelines and patient records to speed up reviews that currently take hours of human labor.
Drug Trials: In a demo, Anthropic showed Claude reducing the time to draft a Phase II clinical trial protocol from "many days" to about one hour. Pharma execs definitely sat up straighter at that one.
Now… for the "not so great" part. Anthropic says Claude Opus 4.5 hits 92.3% accuracy on medical math.
Sounds impressive until you realize that means 1 out of every 13 calculations is wrong. In a hospital, that is a lawsuit (or worse) waiting to happen.
And history isn’t exactly calming us down. The FDA’s own tool Elsa (powered by Claude) has been caught hallucinating fake studies. UnitedHealthcare is facing lawsuits over AI-driven claim denials. Google’s Med-PaLM has caught some heat. IBM Watson Health was supposed to cure cancer and instead got sold for parts.
So yeah, the tech is powerful, but "powerful" doesn't automatically mean "ready for life-or-death decisions."
Anthropic knows this, which is why their head of life sciences, Eric Kauderer-Abrams keeps stressing that humans stay firmly “in the loop.” Their policy in fact requires a qualified professional to review everything. In other words: maybe don’t let a chatbot calculate your heart meds on its own just yet.
The Bottom Line
We’re officially past the "experimental" stage. OpenAI launched last week. Anthropic launched yesterday. Google and Microsoft are already deep in the trenches.
So yes, the era of "Vibe Medicine" is here. Whether that excites you or mildly terrifies you probably depends on how much you trust a chatbot with your medical history. Both companies promise your data won’t be stored or used for training, but yeah, we’ve heard that song before.
So stay curious. Stay cautious. And for the love of all things sterile and holy, keep your actual doctor on speed dial.
Hiring in 8 countries shouldn't require 8 different processes
This guide from Deel breaks down how to build one global hiring system. You’ll learn about assessment frameworks that scale, how to do headcount planning across regions, and even intake processes that work everywhere. As HR pros know, hiring in one country is hard enough. o let this free global hiring guide give you the tools you need to avoid global hiring chaos.
🧱 Around The AI Block
❌ Indonesia and Malaysia block Grok over non-consensual, sexualized deepfakes.
🦾 Google announces a new protocol to facilitate commerce using AI agents.
🤦♀️ X could face UK ban over deepfakes, minister says.
📝 Meta signs nuclear energy deals to power Prometheus AI supercluster.
🏎️ Motional puts AI at center of robotaxi reboot as it targets 2026 for driverless service.
🔎 Meta and Harvard researchers introduce the Confucius Code Agent (CCA): A software engineering agent that can operate at large-scale codebases.
🛠️ Trending Tools
For the E-com Hustlers: Product Link To Video turns any product URL (Amazon, Shopify, you name it) into a high-converting video ad. Just paste the link, and it automatically grabs your images and info to build a scroll-stopping promo.
For the Visual Editors: GPT Image 1.5 – OpenAI just upgraded image generation inside ChatGPT. It’s 4x faster and way better at keeping faces consistent, making it a proper "creative studio" for iterative design.
For the Efficiency Obsessed: Gemini 3 Flash is Google’s newest model built for pure speed. It uses 30% fewer tokens while delivering PhD-level reasoning, making it the perfect (and cheaper) engine for coding agents and complex workflows.
For the Automation Pros: Refly.ai lets you build complex, no-code visual workflows with an AI co-pilot. You can connect over 3,000 apps and even monetize your creations by earning credits when other people use your automation templates.
For the Presenters: AiPPT.com brings "Death by PowerPoint." to an end. You drop in a PDF, document, or URL, and this tool generates a full, professional slide deck with layouts and visuals ready to go in seconds.
Hope these tools help you reclaim some of your time today!
🤖 AI Workout Of The Day: How To Use AI to Create Better Product Images for Ecommerce
Let’s be real for a second. If you’re running an Etsy shop or a side hustle, professional photographers are expensive. Like, "there goes my entire Q1 profit" expensive.
Well, good news: you don't exactly need them anymore.
AI product photography is currently eating the creative world. In 2025, 76% of small businesses using these tools reported cost savings of over 80%. This means they were getting professional-grade shots without the professional price tag.
Here's what makes this so exciting: one Amazon seller reported sales jumping from 3 bottles of perfume per day to 50 bottles after switching to AI-enhanced product photos, folks, that's 16 times more sales! Not bad for a few clicks, right?
Well, lets break down the top tools that literally anyone can use:
The Budget Heroes
Photoroom ($7.50/mo): The GOAT for mobile sellers. It is super user-friendly and handles background removal like a pro.
Gemini / Nano Banana: You can snag a few free shots in Gemini, but if you want the high-fidelity "wow" factor, Nano Banana is where the magic happens.
ChatGPT: Yep, the chatbot you use for emails can now make product photos. It is great for a quick 2-3 shot trial before you hit usage limits.
The Premium Powerhouses
Pebblely ($15/mo): This is for the "I have too much inventory" crowd. It handles bulk processing for 25+ products and has 40+ themes ready to go.
How to "Vibe Code" Your Photos (Step-by-Step)
The Base Shot: Take a clear photo with your phone. No fancy lights are required; just make sure it is clean and in focus.
The Upload: Drop it into your tool of choice. Most will auto-detect your product and delete the messy background instantly.
The Magic: You can either pick a template or describe your dream scene. Pro-tip: try "minimalist marble slab with soft morning sunlight."
The Polish: Let the AI handle the shadows and reflections. This is what separates "bad Photoshop" from "professional studio."
The Tweaks: Guide the AI tool until it removes objects or resizes your product images to perfection.
The Ship: Download and go. Stick it on Amazon, Instagram, or your website and watch the clicks roll in.
💡 Pro Tips for 16x More Sales
Keep Prompts Simple: Don't overthink it. Focus on lighting and setting, then let the AI do the heavy lifting.
Batch It Up: Some of these tools can churn out a lot of images per day. If you have 25 products, do them all at once for a consistent brand "vibe."
Don't Be Lazy: AI can get "hallucination-happy" with shadows sometimes. Always do a quick quality check before you post.
🎁 Special Bonus: The "Zero Dollar" ChatGPT Hack
Here's a super simple method using ChatGPT that costs nothing:
Upload your photo to ChatGPT and use this exact prompt: "As a creative product photography expert, replace the background for [Your Product] while maintaining a professional style. Use a 3:4 aspect ratio."
It is free, it is fast, and it works surprisingly well.
So what are you planning to "vibe-shoot" first? Reply and let us know, or send us your best AI-generated product shot!
⚡ Prompt to try for generating Product Images for Sneakers, Apparel, Accessories & Cohesive Catalog Styles:
"Create a realistic image showing a person wearing these sneakers in an urban street setting with natural lighting"
"Generate a lifestyle photo of someone wearing this shirt in a coffee shop, warm afternoon lighting, casual atmosphere"
"Show this product being worn by a person in a minimalist modern setting, soft natural light, focus on the product"
"Place this product in a styled home office desk setup, morning sunlight through window, professional atmosphere"
"Show this product in a cozy autumn setting with warm tones and fall leaves in the background"
Create a realistic lifestyle photo showing this product being used in a natural setting with soft natural lighting"Is this your AI Workout of the Week (WoW)? Cast your vote!
That's all we've got for you today.
Did you like today's content? We'd love to hear from you! Please share your thoughts on our content below👇
What'd you think of today's email?
Your feedback means a lot to us and helps improve the quality of our newsletter.
