AI predicts cancer outcomes from selfies

-

Good morning, AI enthusiasts. The saying “you’ll be able to tell quite a bit from someone’s face” just got AI validation, with a brand new system accurately predicting cancer survival from facial photos alone.

Mass General’s FaceAge AI just converted looks right into a invaluable medical biomarker — and your face could also be about to develop into your doctor’s newest diagnostic tool.

In today’s AI rundown:

  • AI predicts cancer outcomes from photos

  • Sakana teaches AI to think with time

  • Transform videos into content gold mines

  • OpenAI’s HealthBench to guage healthcare AI

  • 4 recent AI tools & 4 job opportunities

LATEST DEVELOPMENTS

AI RESEARCH

🧓 AI predicts cancer outcomes from photos

Image source: Mass General Brigham

The Rundown: Mass General Brigham’s researchers just introduced FaceAge, an AI tool that may estimate an individual’s biological age and improve cancer survival consequence predictions just by analyzing their facial photograph.

The small print:

  • FaceAge uses a system trained on tens of hundreds of face photos to translate subtle facial characteristics right into a biological age estimate.

  • The study found that cancer patients, on average, appeared about 5 years older, with a better FaceAge correlating with worse survival rates.

  • In physician testing, doctors showed significant improvement in accuracy when predicting 6-month survival when adding FaceAge risk scores to clinical data.

  • The AI’s predictions correlated with a gene related to cellular aging, suggesting FaceAge captured processes not detected by chronological age.

Why it matters: While we’re taught not to evaluate books by covers, our faces may very well reveal crucial health insights. By quantifying what physicians have intuitively observed for a long time, this tech turns facial characteristics into actionable biomarkers which will help doctors personalize treatments more precisely than ever before.

TOGETHER WITH HUBSPOT

The Rundown: HubSpot’s “AI For Business Builders” free guide shows you join the savvy entrepreneurs quietly leveraging artificial intelligence for real business growth — with no P.H.D. needed.

This no-BS guide cuts through the hype, providing you with:

  • A jargon-free breakdown of huge language models (the engine behind AI’s magic)

  • Concrete steps to leverage AI for immediate business growth

  • Real-world examples that turn “prompt engineering” from gibberish right into a secret weapon

SAKANA AI

🧠 Sakana teaches AI to think with time

Image source: Sakana AI

The Rundown: Sakana AI unveiled Continuous Thought Machines (CTMs), a brand new variety of model that makes AI more brain-like by allowing it to “think” step-by-step over time as a substitute of constructing easy decisions like current AI systems do.

The small print:

  • Unlike most AI that processes information in a static, one-shot way, the CTM considers how its internal activity unfolds over time, very like our brains do.

  • The tech draws inspiration from real brains, where the timing of when neurons activate together is crucial for intelligence.

  • Sakana demoed the CTM solving complex mazes, showing the model visibly tracing possible paths through the maze because it thinks.

  • One other example tackled image recognition, with a CTM viewing different parts of a picture and spending more time based on the problem of the duty.

Why it matters: Sakana is a singular AI startup in its mission to bring ‘nature-inspired’ methods to AI models, and these CTMs provide a differentiator that might help bring the pliability and adaptableness of human brains to advanced systems — resulting in AI that reasons, learns, and solves problems in a more human-like fashion.

AI TRAINING

🎥 Transform videos into content gold mines

The Rundown: On this tutorial, you’ll learn use Google’s NotebookLM to upload and analyze your videos and generate transcripts, title ideas, hooks, and descriptions to enhance your content creation workflow.

Step-by-step:

  1. Visit NotebookLM and register together with your Google account, then click “Create recent” to begin a fresh notebook.

  2. Add your video within the Sources panel by uploading your file or connecting to YouTube.

  3. Generate a transcript by typing prompts like “Provide an entire transcript” or “Translate the transcript to Spanish.”

  4. Improve your content by asking for “10 higher hooks,” “5 YouTube title ideas,” or “YouTube description with relevant tags.”

Pro tip: You can too upload multiple videos with stats and ask NotebookLM to research which content style performed best to realize insights for future content strategy.

PRESENTED BY INNOVATING WITH AI

🚀 Launch a six‑figure AI consultancy in six months

The Rundown: Innovating With AI’s “The AI Consultancy Project” gives you the frameworks, playbooks, and client‑ready templates needed to show “interesting AI ideas” right into a revenue‑generating business – helping you ride the wave of an AI consulting boom expected to grow by 8x this decade.

On this program, you’ll:

  • Gain the tools and frameworks to seek out clients and deliver top-notch services

  • Follow a 6-month plan to construct a 6-figure AI consulting business

  • Join a 700‑strong cohort where some members landed their first AI client in 72 hours

Click here to request access to The AI Consultancy Project.

OPENAI

🏥 OpenAI’s HealthBench to guage healthcare AI

Image source: OpenAI

The Rundown: OpenAI released HealthBench, a benchmark created with 262 physicians to guage how AI systems perform in health conversations — and establish a brand new standard for measuring AI’s safety and effectiveness in medical contexts.

The small print:

  • The benchmark tests models across several themes (like emergency referrals and global health) and behaviors (accuracy, communication quality, etc.).

  • Recent models appeared to perform significantly better on the benchmark, with OpenAI’s o3 scoring 60% in comparison with GPT-3.5 Turbo’s 16%

  • The outcomes also revealed that smaller models are actually way more capable, with GPT-4.1 Nano outperforming older options while also being 25x cheaper.

  • OpenAI has open-sourced each the evaluations and testing dataset of 5,000 realistic, multi-turn health conversations between models and users.

Why it matters: There may be an amazing amount of evidence that AI can provide serious improvements across the board in healthcare settings, and having physician-validated benchmarks is a very important step for each measuring each model’s performance in medical contexts and deciding when and deploy them.

QUICK HITS

🛠️ Trending AI Tools

  • 🤖 HunyuanCustom – Generate custom videos with consistent subjects

  • 📝 Granola iOS – AI notepad for people in back-to-back meetings

  • 🧠 Deep Research – Connect GitHub repos to OpenAI’s deep research agent

  • 🦌 Deer Flow – ByteDance’s open-source deep research tool

💼 AI Job Opportunities

  • 📞 The Rundown – Account Manager

  • 🧭 OpenAI – Product Manager, Enterprise

  • 🛠️ xAI – Engineering Intern

  • 🌟 Hebbia – Director of Customer Success

📰 All the things else in AI today

Google DeepMind launched the AI Futures Fund, an initiative that provides AI startups early access to advanced models, funding, and technical expertise to spice up growth.

Softbank’s $100B commitment towards OpenAI’s Stargate is reportedly being stalled with fears over U.S. tariffs and rising data center costs.

Perplexity is reportedly set to boost a brand new $500M round of funding that reinforces the corporate’s valuation to $14B.

Carnegie Mellon researchers published LegoGPT, an AI system that may create stable, buildable LEGO structures from text prompts.

Saudi Arabia unveiled Humain, a brand new AI enterprise, chaired by Crown Prince Mohammed bin Salman, that goals to make the country an AI hub within the region.

The U.S. FDA plans to deploy AI throughout the agency by the tip of June, following a successful pilot where reviewers accomplished three-day tasks in minutes.

COMMUNITY

🎥 Join our next live workshop

Join our next workshop on Wednesday, May 14th, at 3 PM EST with Rahul Sonwalkar, founding father of Julius AI. On this hands-on live session, you’ll learn analyze any dataset using just natural language, no code or spreadsheets required.

RSVP here. Not a member? Join The Rundown University on a 14-day free trial.

🤝 Share The Rundown, get rewards

We’ll all the time keep this article 100% free. To support our work, consider sharing The Rundown with your pals, and we’ll send you more free goodies.

That is it for today!

Before you go we’d like to know what you considered today’s newsletter to assist us improve The Rundown experience for you.
  • ⭐️⭐️⭐️⭐️⭐️ Nailed it
  • ⭐️⭐️⭐️ Average
  • ⭐️ Fail

Login or Subscribe to take part in polls.

See you soon,

Rowan, Joey, Zach, Alvaro, and Jason—The Rundown’s editorial team

ASK ANA

What are your thoughts on this topic?
Let us know in the comments below.

0 0 votes
Article Rating
guest
0 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments

Share this article

Recent posts

0
Would love your thoughts, please comment.x
()
x