In partnership with
The AI your stack deployed is losing customers.
You shipped it. It works. Tickets are resolving. So why are customers leaving?
Gladly's 2026 Customer Expectations Report uncovered a gap that most CIOs don't see until it's too late: 88% of customers get their issues resolved through AI — but only 22% prefer that company afterward. Resolution without loyalty is just churn on a delay.
The difference isn't the model. It's the architecture. How AI is integrated into the customer journey, what it hands off and when, and whether the system is designed to build relationships or just close tickets.
Download the report to see what consumers actually expect from AI-powered service — and what the data says about the platforms getting it right.
If you're responsible for the infrastructure, you're responsible for the outcome.
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
Claude AI's Company Anthropic Loses Pentagon Battle and Can't Work With the US Government
TLDR: A federal appeals court ruled that the Department of War can keep Anthropic — the company that makes Claude — on a national security blacklist, even as a separate court has blocked the government from banning Claude everywhere else.
The Story:
Anthropic was labeled a "supply chain risk" by the Department of War earlier this year — a designation previously used only against foreign adversaries. The label came after Anthropic refused to remove guardrails that the military requested in order to deploy the software. On April 8, a three-judge panel of the U.S. Court of Appeals for the D.C. Circuit declined to block the designation, saying the financial harm to Anthropic was not enough to outweigh the government's interest in managing its AI supply chain during an active military conflict.
Its Significance:
This case is the first time a U.S. company has been put on a supply chain blacklist under laws designed to stop foreign sabotage. Anthropic says the designation could cost it billions of dollars and damage its reputation with partners who work with the military. The outcome could decide whether AI companies are free to set rules for their own products, or whether governments can demand full, unrestricted access as a condition of doing business. Should war be an exception to the rule? These answers could soon be decided by the courts. The ruling does not affect the general public's ability to use the software.
QUICK TAKES
The story: An analysis by AI startup Oumi found that Google's AI Overviews — the summaries that pop up at the top of search results — are accurate only about 91% of the time. That sounds decent until you scale it: Google handles around 5 trillion searches per year, which works out to tens of millions of wrong answers every hour.
Your takeaway: The study also found that with Google's newer Gemini 3 model, the AI cites sources that don't actually support what it's saying 56% of the time — worse than before. If you're using Google for important questions about your health, money, or news, double-checking the answer elsewhere is a smart habit.
The story: Anthropic recently restricted its Claude Mythos Preview model — which can find software security holes faster than almost any human — to a small group of trusted tech companies. Now OpenAI is doing something similar with its GPT-5.3-Codex model, which it has rated "high" for cybersecurity risk, limiting access through a special invite-only program.
Your takeaway: This looks like the start of a new trend: some AI models will simply not be available to the general public. If a model is powerful enough to hack critical systems at scale, expect companies to keep it locked behind special access programs. Going forward, not all AI tools will be for everyone.
The story: Matei Zaharia, co-founder of Databricks, won the 2026 ACM Prize in Computing — often described as computing's equivalent of a Nobel Prize, with a $250,000 award — for building Apache Spark, a tool that made working with enormous amounts of data affordable for any company. In a TechCrunch interview, Zaharia said he believes AGI (artificial general intelligence) is already here, just in a form we don't fully recognize yet.
Your takeaway: Zaharia's argument is that we keep measuring AI against human standards, which misses the point. AI doesn't need to think like us to be transformative. His bigger concern isn't whether AI is "smart enough" — it's whether we're building the right tools to use it safely and accurately.
TOOLS ON OUR RADAR
🎨 Krita Free and Open Source: A professional grade digital painting application created by artists that provides advanced brush engines and illustration tools for concept art and comics without any monthly subscriptions. (Alternative to Adobe Photoshop)
🌍 GeoSpy Paid: A visual intelligence platform that uses artificial intelligence to analyze the pixels in your photographs and accurately predict their geographic location without relying on any hidden metadata.
Skyvern Freemium: A browser automation agent that uses computer vision to autonomously navigate complex websites and complete multi step web tasks completely on its own.
🗺️ Map This Freemium: A spatial data utility that leverages artificial intelligence to instantly convert your spreadsheets and text files into beautifully customized interactive maps for presentations and analysis.
TRENDING
Hacker Allegedly Steals 10 Petabytes of Secret Data from China's Top Supercomputer — An anonymous hacker going by "FlamingChina" claims to have broken into China's National Supercomputing Center in Tianjin and stolen a massive trove of data, including files marked "secret" in Chinese, missile schematics, and aerospace simulations. Cybersecurity experts who reviewed samples say the data appears genuine. The hacker allegedly got in through a compromised VPN and used a network of automated programs to quietly siphon data over six months.
Intel and Google Expand Partnership to Build More Powerful AI Chips — Intel and Google announced an expanded deal to use Intel's Xeon and Xeon 6 processors for AI workloads, and to jointly develop custom infrastructure chips. The move reflects a growing shift away from training AI models — which requires massive GPU clusters — toward deploying them, which calls for more flexible, general-purpose computing.
AWS Boss Explains Why Investing Billions in Both Anthropic and OpenAI Is Fine — AWS CEO Matt Garman told the HumanX conference that Amazon's $50 billion investment in OpenAI — on top of its $8 billion stake in Anthropic — isn't a conflict of interest. He said AWS has always competed with its own partners and that the future of AI will likely be "multi-model," with different AI tools used for different tasks on the same platform.
Oprah Daily: Should You Use AI for Health Questions? — AI is becoming a go-to source for health information, with more than 40 million people asking ChatGPT health questions every day. Oprah Daily takes a look at what AI can and can't do when it comes to medical questions, including when it's helpful, when it falls short, and what to watch out for before trusting an answer with your health.
NVIDIA Highlights Real-World Robots During National Robotics Week — NVIDIA is spotlighting AI-powered robots moving from simulated training into real industries this week. Examples include solar-powered farm robots using vision AI to remove weeds without chemicals, and autonomous robots completing utility-scale solar panel installations. NVIDIA says advances in simulation tools, synthetic data, and foundation models are helping robots handle complex, unpredictable environments faster than ever before.
Factory Launches Desktop App for Running Multiple AI Coding Agents at Once — Factory's new macOS and Windows app lets developers run several AI "Droids" at the same time, each working on different coding tasks. Each Droid keeps its own memory between sessions — installed packages, cloned repos, running services — so it picks up right where it left off, rather than starting from scratch each morning.
TRY THIS PROMPT (copy and paste into Claude, ChatGPT, or Gemini)
Build a personal expense tracker app that logs every purchase, shows category breakdowns, and monitors your monthly budget.
Build me a single-file HTML expense tracker app using React 18 (loaded from unpkg CDN) and Babel standalone. Include these features:
1. Monthly budget overview panel with total spent, remaining balance, and a progress bar that turns red above 90%
2. Category breakdown showing spending by category (Housing, Food
3. Add expense form with description field, amount, and category dropdown with emoji i
4. Scrollable expense log showing each entry with category color, emoji, date, and
5. A tips section at the bottom with financial advice
Use a dark teal color scheme: background #061a1a, accent #14b8a6, text #ccfbf1. PreWhat this does: A fully functional expense tracker that runs in your browser with no setup. Add expenses with description, amount, and category. The budget progress bar updates in real time, and category totals recalculate automatically.
What this looks like:
WHERE WE STAND(based on today’s news)
✅ AI Can Now: Find software security holes in major operating systems faster than nearly any human expert, prompting companies to restrict access to these tools entirely.
❌ Still Can't: Be trusted without fact-checking. Google's AI Overviews give wrong or unsupported answers tens of millions of times per hour at current search scale.
✅ AI Can Now: Power autonomous robots doing full, real-world jobs, from weeding farms to installing solar panels — without human operators guiding every step.
❌ Still Can't: Reliably verify its own outputs in real-time, meaning a human still has to review what it produces before it can be trusted in high-stakes environments like military operations.
FROM THE WEB
No idea what's going on in the clip, but it's visually striking and can easily pass for a professional trailer.
RECOMMENDED LISTENING/READING/WATCHING
Three MIT students on a cross-country road trip follow a trail of hacker breadcrumbs to a remote location in the Nevada desert, and wake up somewhere that doesn't make sense. It's a slow burn that earns its third act, and works best if you go in knowing as little as possible. Low budget but confident in its direction, with a visual style that shifts deliberately as the story unfolds.
Hiring in 8 countries shouldn't require 8 different processes
This guide from Deel breaks down how to build one global hiring system. You’ll learn about assessment frameworks that scale, how to do headcount planning across regions, and even intake processes that work everywhere. As HR pros know, hiring in one country is hard enough. So let this free global hiring guide give you the tools you need to avoid global hiring headaches.
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.
