In partnership with
Startups who switch to Intercom can save up to $12,000/year
Startups who read beehiiv can receive a 90% discount on Intercom's AI-first customer service platform, plus Fin—the #1 AI agent for customer service—free for a full year.
That's like having a full-time human support agent at no cost.
What’s included?
6 Advanced Seats
Fin Copilot for free
300 Fin Resolutions per month
Who’s eligible?
Intercom’s program is for high-growth, high-potential companies that are:
Up to series A (including A)
Currently not an Intercom customer
Up to 15 employees
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 published with the intention of making AI news and technology more accessible to everyone.
THE FRONT PAGE
Your Insurance Denial Might Come From AI—With No Doctor Review
TLDR: Seventy-one percent of health insurers now use AI for prior authorization decisions, but the algorithms operate as "trade secrets" with no FDA oversight—while patients appealing just 1 in 500 denials face systems designed to delay care until they can't afford the fight or die from their condition.
The Story: A 2025 survey from the National Association of Insurance Commissioners revealed that 71% of health insurers use AI for "utilization management"—the process deciding whether your doctor's recommended treatment is "medically necessary" and gets covered. Indiana University law professor Jennifer Oliva told PBS NewsHour that unlike clinical AI tools doctors use for diagnosis, insurance algorithms face zero FDA regulation and companies claim they're proprietary trade secrets. In 2023 alone, 73 million Americans on Affordable Care Act plans had claims denied, yet less than 1% appealed—partly because insurers spend an average of 1.2 seconds reviewing each case, according to ProPublica's 2023 investigation of Cigna. California passed a law in 2024 requiring licensed physicians to supervise AI coverage decisions, but most states leave insurers free to define "medical necessity" themselves.
Its Significance: Oliva's research in the Indiana Law Journal shows insurers strategically deploy AI knowing the appeals process takes years—so when algorithms deny care to patients with terminal illnesses, companies bet those patients will die before winning their appeals. The data backs this up: patients with chronic illnesses face higher denial rates. Now patients are fighting back with their own AI—services charging $40-50 to generate appeals using the same algorithmic logic insurers use to deny claims. Companies like Claimable charge around $40 to generate customized appeal letters that compile clinical research and analyze other patients' successful appeals—and they've overturned roughly 1,000 denials so far. North Carolina nonprofit Counterforce Health offers the service free, using AI to scan denial letters, comb through policy documents, and draft appeals in minutes rather than the hours most patients spend. Software engineer Holden Karau built FightHealthInsurance.com after her own denial battles, offering a free platform that's especially helpful since most patients who appeal actually win—but historically only 1 in 500 even tries. The uncomfortable reality? We've created an AI arms race where algorithms deny care and other algorithms fight back, while the question of whether profit-driven AI should make life-or-death medical decisions in the first place remains unresolved.
QUICK TAKES
The story: Figure AI's former head of product safety filed a lawsuit claiming he was fired for warning executives that their humanoid robots could fracture human skulls. He says the company weakened its safety plan right after closing a $39 billion funding round that included investments from Nvidia, Microsoft, and Jeff Bezos.
Your takeaway: As humanoid robots move closer to real-world deployment, this lawsuit shows the tension between rapid commercialization and making sure the technology is actually safe.
The story: Alibaba's Qwen app crossed 10 million downloads in its first week after relaunching this month. The free app runs on the company's Qwen3 model and competes directly with ChatGPT by offering research tools, image creation, and plans to connect shopping, maps, food delivery, and more services.
Your takeaway: Chinese tech giants are making aggressive moves into consumer AI with free apps backed by open-source models, challenging American companies on both price and features.
The story: The IRS is rolling out Salesforce AI agents across multiple divisions after cutting its workforce by 25% earlier this year. The AI tools will help process customer requests faster, but they can't make final decisions or distribute funds without human review.
Your takeaway: Government agencies are turning to AI to handle reduced staffing, raising questions about whether automation is replacing workers or just helping the remaining staff keep up.
The story: AIG, Great American, and WR Berkley are asking regulators to exclude AI-related claims from corporate insurance policies. They call AI systems a "black box" that's too unpredictable to insure, pointing to cases like Google's AI falsely accusing a company and triggering a $110 million lawsuit.
Your takeaway: When the companies whose job is managing risk say AI is too risky to cover, that's a major red flag about how ready this technology really is for widespread business use.
The story: Lenovo is holding 50% more component inventory than usual because AI data centers are buying up memory chips and driving prices higher. The world's biggest PC maker is stockpiling to avoid shortages and keep costs down.
Your takeaway: The AI boom isn't just affecting tech companies—it's creating supply chain ripples that could raise prices for everyday electronics like laptops and phones.
TOOLS ON OUR RADAR
🔨 Counterforce Health
[Free]: Generate AI-powered insurance appeal letters that cite medical evidence and policy language to fight denied claims.📐 Claimable
[Paid]: Appeal denied health insurance claims in minutes using AI to craft evidence-based letters that cite medical studies and policies.📐 Fathom
[Freemium]: Stop taking meeting notes forever with AI that records, transcribes, and summarizes Zoom, Teams, and Google Meet calls automatically.🛠️ Gamma
[Freemium]: Create polished presentations, documents, and websites from text prompts—no design skills required, just describe what you want.
TRENDING
AI Improves Early Alzheimer's Detection by Combining Multiple Scans – Researchers developed an AI system that integrates MRI, PET scans, clinical data, and genetic indicators to detect Alzheimer's earlier while reducing the number of expensive imaging tests patients need.
AI Agents Could Shape Future Economies – Economists published research examining how AI agents that can plan and execute complex tasks might interact with humans, reshape markets, and transform business organizations over the next decade.
Text Tool Blocks AI From Reading Your Writing – A new free tool called Gibberifier inserts invisible characters between letters to confuse AI systems while keeping text readable for humans, helping prevent plagiarism and waste AI processing tokens.
Tech Companies Train AI on Your Public Posts and Profile Data – Meta, Google, and LinkedIn use public posts, photos, and profile information to train AI tools, with limited opt-out options and Meta providing none at all for U.S. users.
AI Simplifies Medical Reports From 7 Minutes to 2 Minutes Reading Time – A study found AI-simplified CT scan reports cut reading time from seven minutes to two, with 80% of patients calling them easier to understand, though 6% contained factual errors requiring human review.
TRY THIS PROMPT (copy and paste into Claude, ChatGPT, Grok, Perplexity, Gemini)
Skill Acceleration Roadmap: Build a 30-day learning sprint to master any skill faster using science-backed techniques
Design a personalized 30-day learning roadmap that uses accelerated learning techniques to help me make real progress on a new skill.
**My context**:
- Skill I want to learn: [e.g., Python, guitar, public speaking]
- Current level: [e.g., complete beginner, some basics, intermediate]
- Daily time available: [e.g., 30 minutes, 1 hour, 2 hours]
Create:
1. **Research Learning Science** - Search for latest research on skill acquisition, deliberate practice, and memory retention. Find 3 proven accelerated learning techniques.
2. **Build Interactive Sprint Tracker** - Create a 30-day dashboard featuring:
• Daily practice milestones with progressive difficulty
• Spaced repetition reminder system
• Skill checkpoint tests every 7 days
• Progress visualization and streak counter
Make it motivating and easy to follow.
3. **Video Micro-Lessons** - Generate a 45-second daily practice routine video. Find 3 YouTube channels with the best tutorials for my chosen skill.
4. **Resource Library** - Present in visual cards:
• Top 3 learning platforms or courses
• Practice exercises ranked by difficulty
• Community forums or accountability groups
• Week-by-week skill milestonesWhat this does: Combines learning science research with an interactive 30-day tracker to create a structured, science-backed path for skill development, complete with daily practice routines and curated resources tailored to your schedule.
WHERE WE STAND
✅ AI Can Now: Run full government operations with reduced human oversight, as the IRS deploys AI agents after cutting 25% of its workforce.
❌ Still Can't: Be reliably insured by companies whose entire business is calculating risk, with major insurers calling AI systems "too much of a black box."
✅ AI Can Now: Make critical healthcare decisions about treatment approvals and hospital stays that directly affect patient outcomes and costs.
❌ Still Can't: Guarantee safe physical interactions, as humanoid robot companies face lawsuits over machines powerful enough to fracture human skulls.
✅ AI Can Now: Dominate global supply chains so thoroughly that the world's biggest PC maker must stockpile 50% more inventory to handle the squeeze.
❌ Still Can't: Function independently from human labor, still requiring content moderators and data workers—many in developing countries—to train and filter outputs.
FROM THE WEB
RECOMMENDED LISTENING/READING/WATCHING
If you want to see AI-human integration go horribly wrong, watch Upgrade. Grey Trace is a technophobe mechanic who ends up paralyzed after an attack that kills his wife. A tech genius offers him an experimental chip implant called STEM that restores his movement and can take full control of his body. What starts as a revenge thriller takes a fun twist that I won’t ruin with spoilers here.
Made on a $3 million budget, the fight choreography alone is worth the watch if you just want to turn your brain off and let the machine take over.
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.
