AI Can Predict Your Lifespan

Welcome back apprentices! 👋

What if your face knew more about your health than your doctor did?

One selfie might soon shape cancer care — researchers trained an AI to estimate your biological age from your face, and it’s oddly good at it.

This isn’t a beauty filter. 

It’s not sci-fi. 

It’s something stranger, more useful… 

and kind of brilliant.

In today's email

• How Old Do You Look?

• Google’s ‘I’m Feeling Lucky’ Might Get AI’d Out of Existence

• + 3 New AI Tools

_{Read Time: 4 minutes}

Quick News 

⛪️ Holy See, AI? In a historic first, the new Pope Leo XIV cited artificial intelligence as a factor in choosing his papal name — drawing inspiration from Pope Leo XIII, who guided the Church through the Industrial Revolution. Now, Leo XIV is gearing up for Round 2: the AI Revolution, warning that humanity’s dignity and labor rights must keep pace with the machines. Yep, even the Vatican’s talking AI — though Pope Leo’s probably not asking ChatGPT for homily tips.

🔋 iPhone Getting Smarter About Staying Awake. Apple may juice up iOS 19 with a new AI-powered battery feature that learns your habits to save power — no more random battery drops mid-Spotify. The system would run on-device and fine-tune how your iPhone charges, throttles apps, and times updates based on your daily rhythm. It’s like your phone finally realizes you always binge scroll at 11 p.m. and plans accordingly.

📰 AI, Meet PhD. OpenAI’s top science mind says the future of AI isn’t just smarter chatbots — it’s models that could do actual, original research. Jakub Pachocki, now heading AI wizardry at OpenAI, hints that future systems might autonomously design hardware, write code, and even craft new scientific ideas — all with barely a coffee break. Oh, and they’re cooking up an open-weight model for researchers to tinker with. That sound you hear? Scientists perking up and sweating a little.

Together with Bolt

Prompt, run, edit, and deploy full-stack web and mobile apps

Say hello to Bolt.new, your AI-powered web development agent that brings full stack coding to your browser with no setup required. Built on StackBlitz’s WebContainers, Bolt lets you build, edit, and test web applications in real time with simple, chat-based prompts. Just tell the agent what you want to build or change, and watch your code update instantly.

Bolt supports today’s most popular web languages and frameworks, so you can get started fast whether you’re building a React app, deploying with Netlify, or integrating a backend with Supabase for auth, storage, and more. Need a mobile version too? Bolt works with Expo to streamline mobile app development.

Whether you’re a solo builder or part of a product team, Bolt is your fast lane to production-ready code right from your browser.

Try it today

Research & Health 
How Old Do You Look? Your Face Might Know More Than Your Doctor

Researchers just trained a deep learning model called FaceAge to estimate a person’s biological age — not by bloodwork or wearables, but just from their face photo.

 It turns out, how old you look can be a better predictor of survival than your actual age — especially for patients with cancer. In a series of tests across 6,000+ patients, FaceAge didn’t just recognize wrinkles — it helped predict outcomes more accurately than doctors using traditional data alone. 

What’s more, the model linked visible aging to specific genes associated with cellular senescence. 

The kicker? Doctors using FaceAge alongside medical charts predicted life expectancy significantly better than using clinical info alone.

📌 Key Points:

• 🧠 58,851 healthy faces trained the model; tested on 6,196 cancer patients.
  
  

• 📈 Cancer patients looked ~5 years older than their chronological age, on average.
  
  

• 📊 FaceAge beat actual age as a predictor of survival across multiple cancer types.
  
  

• 🧪 Linked to genes involved in cellular aging (e.g., CDK6).
  
  

• 🏥 Improved end-of-life decisions: Physicians with FaceAge made better 6-month survival calls (AUC ↑ from 0.74 to 0.80).
  
  

• 💸 Trained on public data like IMDb-Wiki; deployed without professional photography.
  
  

• 🤖 Outperformed subjective clinical judgment and even some clinical scoring models.
  
  

• ⚠️ Ethical watch-outs: Potential for bias, insurance misuse, and overreliance without proper regulation.

Pros and Cons

By transforming something as simple as a selfie into a biologically meaningful signal, FaceAge could democratize prognostic insight, especially in cancer care. Doctors often rely on subjective impressions of how "frail" someone looks to make major treatment decisions. 

FaceAge replaces guesswork with quantifiable evidence. This means more personalized care: identifying who’s fit for aggressive treatment, and who might benefit more from palliative care. 

But with great (face-reading) power comes some eerie possibilities. What if insurers, employers, or tech platforms start using face-based aging algorithms to decide your fate? 

The model was trained on a public dataset full of actors and celebrities — meaning it may carry hidden biases in who it sees as “aging well.” Ethnicity, socioeconomic status, even cosmetic surgery could distort the signal. And because it’s still in research mode, deploying FaceAge without guardrails or real-world calibration could worsen health disparities or lead to over-triaging. 

Without strong oversight, the same tool that helps oncologists today could become a digital phrenology tomorrow.

What’s the Deal for You?

 If you’ve ever looked in the mirror and wondered, “Do I look older than I feel?” — you’re not alone. Turns out, that question could soon be part of your medical chart. 

For patients, FaceAge could help doctors make more tailored, compassionate decisions — especially for older adults or those facing serious illness. 

For clinicians, it’s a potential upgrade from gut instinct to data-driven nuance, especially in cases where there’s no easy answer. 

And for anyone building in digital health or AI, this is a strong signal: biological age is the new vital sign, and the face is just the beginning.

On a bigger level, FaceAge taps into a broader cultural shift — from “one-size-fits-all” to truly personalized healthcare. 

It’s not about looking young for Instagram; it’s about understanding what your appearance reveals about your internal health. That’s a leap not just in tech, but in how we understand aging, risk, and care.

Google
Google’s ‘I’m Feeling Lucky’ Might Get AI’d Out of Existence

Google Search is quietly testing a new “AI Mode” — and it might replace the iconic “I’m Feeling Lucky” button. 

Built with Gemini 2.0, this chatbot-style experience aims to tackle more complex questions using deeper reasoning, real-time data, and follow-up threads, all while sitting inside the familiar Google homepage. Right now, it’s only visible to a small set of Labs testers, but the experiment hints at Google’s bigger push toward conversational search and away from the old-school link list. 

Will it stick around? That depends on how helpful (and how un-weird) it feels.

📌 Key Points

• What’s New: “AI Mode” is a search chatbot powered by Gemini 2.0 that lives inside Google Search — sometimes even replacing the “I’m Feeling Lucky” button.
  
  

• What It Does: You can ask multi-part, nuanced questions like “Compare sleep tracking in rings vs watches vs mats,” and the AI researches, summarizes, and links to sources.
  
  

• How It Works: It uses a “query fan-out” strategy — running multiple searches across subtopics and stitching the results into a digestible, conversational response.
  
  

• Why It’s Useful: It goes deeper than normal search, offers follow-ups, includes real-time info, and feels more like a smart assistant than a static answer box.
  
  

• Design Notes: Some versions feature a rainbow-border button and animations — possibly to make it stand out as the new search default.
  
  

• What’s Replacing What: Google says “I’m Feeling Lucky” might not be going away permanently — but it’s clearly in the crosshairs.

AI Mode vs. Traditional Google Search

AI Mode feels like Google with a personal assistant. Traditional Search feels like Google as a library. Which you prefer depends on whether you're hunting for answers or browsing for ideas.

What’s the Deal for You?

This could mean spending less time hunting for the right link — and more time acting on useful info. AI Mode turns Search into a mini-consultant that helps you make faster decisions, especially for research-heavy tasks like finding tools, planning trips, or comparing options. But you’ll still need to bring your judgment — because sometimes, the shiny answer isn’t always the smartest one.

 So… if you spot a rainbow swirl on Google’s homepage, click it — that’s AI Mode in disguise. Just don’t ask it where to hide your secret files. It will try to help.

Help Your Friends Level Up! 🔥

Hey, you didn’t get all this info for nothing — share it! If you know someone who’s diving into AI, help them stay in the loop with this week’s updates.

Sharing is a win-win! Send this to a friend who’s all about tech, and you’ll win a little surprise 👀

Even Quicker News 

💡 No Data? No Problem. A new AI model called Absolute Zero Reasoner learned math and coding — without a single labeled example — by inventing its own problems and solving them. It even beat models trained on thousands of expert-made datasets.

🩺 Doctor Bot, Meet Your Exam. OpenAI just launched HealthBench, a 5,000-case test to see how well AI handles real-world medical chats — graded with rubrics from 262 doctors across 60 countries. Models like GPT-4.1 are already outperforming some humans… but still flunk hard questions.

🧠 AI Learns to Think in Time. Sakana AI dropped the Continuous Thought Machine — a new brain-inspired model that syncs neuron activity over time to "think" step-by-step, solving mazes and classifying images like a tiny digital cortex.

Today’s Toolbox

🛒 Perplexity Pro + PayPal – Shop smarter, straight from your AI chat

👉 Now you can ask a question and buy the answer — Perplexity Pro lets you research, book, and pay in one go, using PayPal or Venmo.

🪄 Pro tip: Planning a trip? Just type “Find me the cheapest nonstop to Austin next weekend,” and check out right there — no tab-hopping, no passwords.

🎙️ Tactiq – Live meeting notes without typing a word

👉 Instantly captures and summarizes your Zoom or Google Meet calls so you can focus on talking, not typing.

🪄 Pro tip: You can highlight key moments in real-time just by clicking during the call.

💬 Kustomer IQ – AI that handles “Where’s my order?” so you don’t have to

👉 Kustomer IQ auto-replies to common support questions, summarizes chats, and helps human agents respond faster — across email, chat, and socials.

🪄 Pro tip: Use it to cover nights and weekends without needing to hire round-the-clock staff. Your customers get answers; you get sleep.

🧪 Test the Prompt

A playground for your imagination (and low-key prompt skills).

Each send, we give you a customizable DALL·E prompt inspired by a real-world use case — something that could help you in your business or job if you wanted to use it that way. But it’s also just a fun creative experiment.

You tweak it, run it, and send us your favorite. We pick one winner to feature in the next issue.

Bonus: you’re secretly getting better at prompt design. 🤫

👑 The winner is…

Last week, we challenged you to test GPT-4o’s visual generation skills with this prompt.

Here’s the WINNER:

Congrats to Lucas from Montgomery!🥳

Want to be featured next? Keep those generations coming!

🎨 Prompt: The City of Ideas

“Design a massive, layered city built entirely from a single person’s creative output. The city is structured around a core tower made of [unexpected creative material], and its surrounding districts are connected by bridges shaped like [symbolic object]. Light pulses through the structures as new ideas flow in.
Style: ultra-detailed worldbuilding concept art, dramatic contrast lighting, aerial wide shot.”

We’ll be featuring the best generations in our next edition!

_FEEDBACK