Good morning and thank you for joining us again!

Welcome to this daily edition of Beginners in AI, where we explore the latest trends, tools, and news in the world of AI and the tech that surrounds it. Like all editions, this is human curated and edited, and published with the intention of making AI news and technology more accessible to everyone.

THE FRONT PAGE

Your Doctor Can Now Show You Next Year's X-Ray Before the Damage Happens

TLDR: University of Surrey researchers built an AI that generates realistic X-rays of what your knee will look like a year from now, giving doctors and patients a visual preview of joint damage before it happens.

The Story:

The system, presented at MICCAI 2025, was trained on nearly 50,000 knee X-rays from about 5,000 patients. It uses a diffusion model (the same tech behind image generators like Midjourney) to create a "future" version of your current scan, then highlights 16 specific points in the joint where deterioration is most likely. The AI also produces a personalized risk score for disease progression. According to the researchers, it runs about nine times faster than comparable tools while remaining compact enough for real clinical workflows.

Its Significance:

Osteoarthritis affects over 500 million people globally and is the leading cause of disability in older adults. Current tools can only assess damage after it's already happened. The Surrey team argues that showing patients a side-by-side comparison of their current knee and their projected knee creates something abstract risk scores can't: motivation. When you can actually see what skipping physical therapy or ignoring weight management might cost you, the math changes. The researchers are now seeking clinical partners and believe the same approach could work for lung damage in smokers or heart disease progression.

QUICK TAKES

Open AI Models Match Performance at a Fraction of the Cost

The story: MIT researchers found that open AI models achieve about 90% of closed model performance but cost 87% less to run. Despite this, users still choose expensive closed models from OpenAI, Anthropic, and Google nearly 80% of the time.

Your takeaway: Companies are overpaying for AI by defaulting to big-name models instead of evaluating what they actually need. The researchers estimate businesses could save $25 billion annually by making smarter choices.

Bristol Myers and Microsoft Team Up on AI Lung Cancer Detection

The story: Bristol Myers Squibb will use Microsoft's FDA-cleared AI to analyze X-rays and CT scans for early signs of lung cancer. The partnership specifically targets rural hospitals and underserved clinics where screening rates are lowest.

Your takeaway: This isn't just about better tech - it's about getting cancer detection tools to communities that need them most. Over 80% of U.S. hospitals already use Microsoft's imaging network, so the infrastructure is ready.

Ukraine Will Let Allies Train AI on Its Combat Data

The story: Ukraine's new Defense Minister announced plans to share battlefield data - including millions of hours of drone footage - with allied nations to train their AI systems. He called the data trove one of Ukraine's "cards" in negotiations with partners.

Your takeaway: Nearly four years of war has given Ukraine something no one else has: real combat data at massive scale. Kyiv is turning that into leverage, offering allies access in exchange for continued support.

TOOLS ON OUR RADAR

☁️ Proton Drive* Freemium: Store files with end-to-end encryption from the makers of Proton Mail, Swiss privacy guaranteed.

🎨 Leonardo.AI Freemium: Generate stunning AI images with 150 free tokens daily across multiple models including FLUX.

⏱️ ActivityWatch Free and Open Source: Track how you spend time on your devices with this privacy-focused, locally-run time tracker, alternative to RescueTime.

🔐 Ente Auth Freemium: Cross-platform 2FA authenticator with end-to-end encrypted cloud backup, beautiful design meets bulletproof security.

TRENDING

Companies Plan to Double AI Spending in 2026 - A BCG survey of 2,400 executives found companies will boost AI budgets from 0.8% to 1.7% of revenues this year. Nearly all CEOs say they'll keep investing even if returns don't materialize quickly.

ServiceNow and OpenAI Sign Multi-Year Enterprise Deal - The partnership will embed OpenAI's models, including GPT-5.2, directly into ServiceNow's business software. The companies are building AI voice agents that can handle customer service calls and automate IT tasks.

Sandia Lab Uses AI "Labmates" to Supercharge Research - Physicists paired three AI systems together to design and run experiments on LED light control. The AI team achieved a fourfold improvement in about five hours - work that would have taken humans years.

64% of Teens Now Use AI Chatbots - A Pew Research study found 1 in 10 teens use chatbots daily, with many preferring AI conversations to talking with parents. Experts say the judgment-free nature of chatbots is a big draw.

Professor's Experiment: 57 of 60 Students Refused to Use AI on Exam - A Belgian university professor gave students the choice to use chatbots during an open-book exam - but required them to justify any AI-generated answers. Nearly all chose to go without.

TRY THIS PROMPT (copy and paste into Claude, ChatGPT, or Gemini)

Content Remix Machine: Turn one piece of content into ten with a format transformer and platform optimizer

What this does: Kills the "I need to create content for every platform" overwhelm. Drop in one solid piece of content and watch it transform into a week or two of platform-optimized posts, threads, carousels, and scripts. Work smarter, not harder.

What this looks like:

WHERE WE STAND(based on today’s news)

✅ AI Can Now: Run physics experiments autonomously, designing tests, analyzing results, and discovering approaches human researchers hadn't considered - all in a single afternoon.

❌ Still Can't: Match the real-world combat data Ukraine has collected over nearly four years of war. Allied nations are lining up to access this footage because no simulation or training dataset comes close.

✅ AI Can Now: Analyze routine medical scans to flag potential cancers, with FDA clearance, at scale across hospital networks - potentially catching cases that might otherwise be missed.

❌ Still Can't: Handle the precision demands of aircraft manufacturing without extensive human oversight. Airbus is only now beginning to test humanoid robots in assembly - and these bots still need maintenance every 48 hours.

FROM THE WEB

RECOMMENDED LISTENING/READING/WATCHING

Audio Drama: "The Machine Stops"

E.M. Forster wrote this in 1909, and it predicted video calls, instant messaging, online lectures, food delivery, and a humanity so dependent on technology they've forgotten how to live without it. When COVID lockdowns hit, this 115-year-old story suddenly became the go-to reference for "this is exactly what we're living through." It's the great-grandfather of every underground dystopia you've seen, from THX 1138 to WALL-E, and reading or listening to it now hits different when you realize Forster wrote it before World War I.

Thank you for reading. We’re all beginners in something. With that in mind, your questions and feedback are always welcome and I read every single email!

-James

By the way, this is the link if you liked the content and want to share with a friend.

*Some product links may be affiliate or referral links. This helps support the newsletter at no extra cost to you.