Sponsored by
Moda creates AI Decks that don't look AI-Generated
Every AI tool can create "AI slop". You know it the second you see it.
Moda is different: it learns your brand identity and applies it to every slide, document, and marketing asset your team makes. It learns your preferences so the designs get better over time.
The best part? Every layer is fully editable on a real canvas, and exports to powerpoint, PDF and more.
Moda has raised $7.5M from General Catalyst, Pear VC, and the founder of Dropbox.
Beginners in AI
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
AI Is Now Scanning 70 Million Child Soccer Players for Hidden Talent
LEAD STORY
TLDR: US Soccer is using AI to scan videos of tens of millions of young soccer players around the world, and the same approach is about to spread far beyond sports.
The Story:
US Soccer is using AI to scan videos of tens of millions of young soccer players around the world, according to a new report. The goal is to find kids with talent that human scouts can't reach, possibly years before they would appear organically through their exploits. About 50 to 70 million teenagers, boys and girls, play soccer on any given day. US Soccer COO Dan Helfrich said scouts miss 99.5% of those players because there aren't enough humans to watch every game. The AI looks at skill, technique, and movement to match kids to specific positions. CEO JT Batson said the program is still in an early test stage.
Its Significance:
This works for more than soccer. Casting directors could find their next lead in TikTok clips, police could spot a missing kid in thousands of hours of security footage, and hospitals could flag early signs of Parkinson's from short walking videos. What changes most is who gets found. Scouts and specialists used to only see people who lived close by or could afford to come to them. AI takes down that wall, so talent and warning signs show up in places nobody was looking.
QUICK TAKES
The story: Researchers at UW Medicine and Skape Bio used AI to design tiny proteins (under 100 amino acids) that can switch key cell receptors on or off, with results published in Nature. In one mouse study, an AI-designed protein worked as well as a real drug already in use but caused fewer side effects.
Your takeaway: GPCRs control almost every body function, from sight and smell to how hormones and medicines work. Being able to design new molecules that target them on demand could lead to better drugs for diseases that have no good treatment today.
The story: A new study from the University of Illinois and the National Weather Service shows how AI now translates U.S. weather warnings into Spanish, Chinese, Vietnamese, Samoan, French, and other languages in real time. About 68.8 million people in the U.S. speak a language other than English at home.
Your takeaway: Warnings used to depend on bilingual forecasters squeezing translations in between their regular jobs. AI now does it fast enough that non-English speakers get the same chance to take shelter or evacuate as everyone else.
The story: The Miller School of Medicine's pathology department ran a 100-Day Agentic AI Challenge that let doctors, residents, and staff with no coding background build their own AI agents using Microsoft Copilot Studio. One resident's billing-code assistant cut a 20-minute daily task down to seconds, and another agent translates technical pathology reports into a 6th-grade reading level so patients can actually understand their results.
Your takeaway: Most AI adoption stories start with "we hired engineers." This one starts with doctors using low-code tools to build their own assistants for the work they already do. That model spreads a lot faster than waiting on the IT department.
TOOLS ON OUR RADAR
🌐 WinSCP Free and Open Source: A reliable visual file transfer application that allows you to securely and safely copy massive folders from your personal computer to remote servers using a beautifully simple dual pane interface.
🔒 Notesnook Free and Open Source: A private digital notebook that seamlessly synchronizes your daily thoughts and important records across all your devices using absolute zero knowledge encryption to keep your data completely safe.
🗜️ Keka Free and Open Source: A beautiful and incredibly powerful file extraction utility for Apple computers that easily compresses massive folders to save space and seamlessly opens absolutely any obscure archive format you encounter.
🚀 Tailor AI Freemium: A beautifully dynamic marketing utility that seamlessly personalizes your website landing pages for different visitors and continuously runs advanced experiments to completely optimize your audience engagement.
TRENDING
China Has Approved the World's First Invasive Brain-Computer Chip — China's drug agency approved NEO, an implant from Neuracle and Tsinghua University that lets people with paralysis control a robotic glove with their thoughts. It beat Elon Musk's Neuralink to market and is now covered by China's public health insurance.
Mayo Clinic Study Shows AI Can Reveal Brain Tumor Risks Without Costly Genetic Testing — Mayo Clinic researchers trained AI on tissue samples and clinical data from 672 patients with meningiomas, the most common brain tumor in adults. The AI predicted whether the tumor would come back using standard pathology slides, skipping the expensive DNA tests many hospitals can't run. The study was published in The Lancet Digital Health.
MetaMask Launches AI Agent Wallet With Security Controls — Crypto wallet MetaMask released a new wallet built for AI agents to trade on their own across Ethereum, Hyperliquid, and other chains. It includes spending limits, allowlists for approved apps, two-factor checks for risky transactions, and up to $10,000 in MetaMask transaction protection. The wallet works with Claude Code, OpenAI Codex, Cursor, and other agent frameworks.
Frontier AI Models Can Find Crypto's Biggest Bugs. Experts Warn the Industry Isn't Ready — A security researcher used Anthropic's Claude Opus 4.8 to find a four-year-old flaw in Zcash that could have let attackers create unlimited fake ZEC tokens. The bug had survived four years of expert reviews. ZEC dropped about 38% after the disclosure.
More Travelers Using AI to Plan Summer Vacations — As summer travel picks up, families are using AI tools to build custom trip plans in seconds instead of clicking through travel sites for hours. Travel experts say AI works best when you give it your destination, budget, and personal interests up front.
Career Expert Tackles Job Search Struggles, From Experience to AI — CBS Sunday Morning talked with recent grads about today's hiring market, where one job listing can pull 300 applications and AI sorts the resumes before any human sees them. The unemployment rate for young workers is about twice the national average right now.
TRY THIS PROMPT (copy and paste into Claude, ChatGPT, or Gemini)
◆ Inventory your skills, contacts, audience, IP, and tools. See which combinations can be sold this month.
Build a single-file HTML app with vanilla HTML/CSS/JS. The Asset Inventory — find sellable combinations of skills/contacts/audience/IP/tools. Persist to localStorage key 'asset_inventory_v1'.
Aesthetic: deep navy (#0a0d18), gold and purple radial gradients. Inter sans 800 for headings, Source Serif 4 for body italic, JetBrains Mono for labels. Gold (#c8a050) primary, with category color-coding for asset chips: gold=skills, purple=contacts, cyan=audience, green=IP, pink=tools. Glowing gold action button.
Form: 5 sections, each with header showing ◆ icon — SKILLS / CONTACTS / AUDIENCE / IP / TOOLS — each a textarea with strong field-help text examples.
System instructions to the model: act as combinatorial revenue strategist. Find COMBINATIONS using at least 2 ingredients per offer. Don't suggest things requiring assets they don't have. Don't say "build audience first." Real offers, dollars, first moves. Return raw JSON: inventory_tags (object with skills/contacts/audience/ip/tools arrays of 1-4 word chip labels), combinations array (4-6 items each with name + ingredients array of asset+type + price + period + pitch + who_buys + delivery + first_step), underused_asset (name SPECIFIC asset they're sleeping on), this_month (2-3 sentences with **bold** around offer name and revenue number).
Render: inventory bar showing all five categories as color-coded chip rows. Each combo card displays numbered header + name + large gold price, "the recipe" row with color-coded ingredient chips joined by + signs, italic pitch, two-column who-buys / what-you-deliver, gold "first move this week" callout. Pink-bordered "asset you're sleeping on" card. Green-bordered "your move this month" card with bolded offer name and dollar number.What this does: Catalog what you've got in five categories — skills you can do, people you can reach, audience that already pays attention, IP you've already made, and tools you have access to. Get back your full asset stack rendered as color-coded chips, 4–6 specific sellable combinations (each showing the exact recipe of which assets combine, the price, who buys it, what you deliver, and one concrete first move for this week), the specific asset you're sleeping on, and one combination to focus on this month with a realistic revenue target. Saves to localStorage.
What this looks like:
WHERE WE STAND(based on today’s news)
✅ AI Can Now: scan video at scale and pull out specific skill, technique, and movement patterns across millions of athletes worldwide
❌ Still Can't: judge whether a player has the grit, work ethic, or coachability to actually develop into a pro
✅ AI Can Now: design tiny proteins under 100 amino acids that bind to specific cell receptors and switch them on or off in living human cells
❌ Still Can't: predict which of those designs will be safe and effective in humans without years of lab and clinical testing
FROM THE WEB
RECOMMENDED LISTENING/READING/WATCHING
A Korean Netflix film from Train to Busan director Yeon Sang-ho about a near-future research lab trying to build the perfect combat AI by uploading the consciousness of a legendary soldier into a robotic body, while her now-grown daughter runs the project. Got mixed reviews, but the central question of whether copying a parent's mind counts as honoring them or violating them is something that may well come up in our lifetimes.
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 * designated product links may be affiliate or referral links. As an Amazon Associate, I earn from qualifying purchases. This helps support the newsletter at no extra cost to you and Amazon makes a tiny hair less.
