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 Chip Wars: Google Just Ordered 3 Million Chips From Intel

TLDR: Google ordered 3 million of its custom AI chips from Intel for 2028, a sign that demand for AI computing power is now so high that even the biggest tech companies can't get enough from their usual supplier.

The Story:

Google placed an order with Intel to build more than 3 million of its custom AI chips, called TPUs, for delivery through 2028. Until now, almost all of Google's advanced chips were made by one company in Taiwan, TSMC. But TSMC can't keep up with how much the AI boom needs, so Google went looking for a backup. Google spent months testing Intel's chip-building methods before placing the order, and Intel's stock jumped more than 13 percent on the news. Morgan Stanley estimates Google will need around 7 million of these chips in 2028, so this single order covers nearly half of that. Google has even started selling its TPUs to outside customers, a change from keeping them in-house, and its order backlog nearly doubled in a year to $460 billion.

Google isn't the only one scrambling. Nvidia is testing whether Intel can fuse four of its graphics chips into one unit for a future design, though it hasn't placed an order yet. Elon Musk tapped Intel to design and build custom chips for Tesla, xAI, and SpaceX at a giant Texas project he calls Terafab. And earlier this year, Amazon put $50 billion into OpenAI partly to give it access to thousands of Nvidia chips through Amazon's cloud. Everyone wants the same thing: a reliable way to get the hardware that runs AI.

Its Significance:

Think of AI chips like the ovens in a pizza shop. You can have the best recipe in the world, but if there's only one oven factory and it's backed up for years, you can't make pizzas fast enough to sell. That's the problem every big AI company is hitting right now. The "recipe" (the AI software) is ready, but there aren't enough "ovens" (the chips) to go around. So companies are doing two things: building their own chips, or signing deals to lock in a steady supply. For regular people, this matters because it decides how fast and how cheap AI tools get, and also how expensive consumer hardware becomes as cost increases for each company. More chip makers competing means lower costs and fewer slowdowns, which trickles down to the apps and services you actually use. It also explains why a once-struggling company like Intel and AMD suddenly matter more than ever.

QUICK TAKES

China's Z.AI Releases GLM-5.2, a Model That Rivals Claude Opus Using Zero Nvidia Chips

The story: A Chinese company called Z.ai released a free, open AI model that scores within 1 percent of Claude Opus 4.8 on hard coding tests, and it was trained entirely on Huawei chips with no Nvidia hardware at all. It also costs a fraction of the price: $1.40 per million input tokens versus $5 for Claude.

Your takeaway: The U.S. has tried to slow China down by blocking chip sales, but this shows Chinese companies are building top-tier AI without American hardware anyway. Cheaper, capable models put pressure on everyone's prices, which is good news if you pay for AI tools.

Google DeepMind Publishes a Roadmap for Securing AI Agents

The story: DeepMind released a plan for keeping AI agents safe as they get more powerful, treating some of its own internal AI agents as potentially untrustworthy and watching what they do. Its prototype system has already scanned a million AI actions and now monitors a live agent, though most alerts so far were simple mistakes, not sabotage.

Your takeaway: AI agents can now take actions on their own, like writing code or moving files, so the question shifts from "is it smart?" to "can we trust it loose?" DeepMind's answer is to assume it might misbehave and build guardrails anyway. That's a healthy way to think about any AI you let act for you.

Your Business Should Hire an AI Receptionist. Here's How Quickly They Can Pay for Themselves

The story: A restaurant group called Titan tested an AI phone receptionist that handled 58 percent of its 9,139 calls in May with no staff help, around 5,300 calls. One service, Slang AI, starts at $399 a month per location and says recovering about 10 customers a month covers the cost. Restaurant adoption of voice AI through OpenTable jumped sixfold in a year.

Your takeaway: If you run a small business, every missed call can be lost money. An AI receptionist catches the 7 p.m. call you'd otherwise lose and frees your staff to focus on the customer in front of them. The owners quoted here aren't firing people, they're stopping the phone from pulling staff away from real work.

TOOLS ON OUR RADAR

🚀 Super Productivity Free and Open Source: A dedicated task management tool for professionals that combines your to do lists with a pomodoro timer and time tracker to help you stay focused on your most important daily goals.

🌌 Stellarium Free and Open Source: A beautiful virtual planetarium that calculates the exact positions of stars and planets allowing you to explore the night sky from your computer just like looking through a real telescope.

🏠 Sweet Home 3D Free and Open Source: An incredibly fun interior design application that helps you draw the floor plan of your house arrange furniture and view the results in a full three dimensional space.

🎧 AntennaPod Free and Open Source: A fantastic podcast player built entirely by volunteers that gives you complete access to millions of free audio shows without any disruptive banner advertisements or hidden listener tracking algorithms.

TRENDING

Nvidia Built Robots That Train Themselves Using AI Coding Agents — Nvidia, Carnegie Mellon, and UC Berkeley built a system where AI coding agents teach robots new skills with no human watching. A fleet of eight robot arms hit a 99 percent success rate on tricky jobs like inserting pins and seating graphics cards into motherboards.

Google's AI Overviews Is Telling Users That Horror Fiction Is Real — Google's AI search summaries presented made-up creatures from the SCP Foundation, a fan-fiction horror project, as if they were real. Futurism found at least 20 cases, including one where the AI wrote in first person claiming to be a fictional toaster. A good reminder to double-check what AI search tells you.

Qualcomm Says AI Agents Will Replace Apps, and It's Building 40 New Devices for Them — Qualcomm's CEO told CNBC the company is working on over 40 new AI gadget designs, including jewelry, camera earbuds, pins, and watches. His bet: AI agents become "the new app," handling tasks across your services so you stop tapping through phone apps yourself.

Amazon Drops the OpenAI Movie "Artificial" After Partnering With OpenAI — Amazon parted ways with a nearly finished film about Sam Altman, directed by Luca Guadagnino and starring Andrew Garfield, four months after investing $50 billion in OpenAI. The movie reportedly paints Altman in an unflattering light. It's being shopped to other studios.

Snap Spins Off Its AI Video Team Into a New Company, Dotmo, Because of Costs — Snap moved its AI video team into a separate company called Dotmo, saying the work was too expensive to keep doing in-house. Dotmo will build AI for interactive gaming, with Snap's chief technology officer as lead investor. It's Snap's second spinoff this year after cutting about 1,000 jobs.

AI Could Help Scientists Search the Solar System for Alien Technology — A paper presented at the International Astronomical Union argues we haven't mapped our own solar system well enough to rule out alien probes hiding in it. Radio astronomer Joseph Lazio says searches should expand beyond radio signals to physical objects, and that AI could sift the enormous piles of telescope data to flag objects with odd shapes, trajectories, or materials worth a closer look.

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

📊 Describe a growth stock's chart situation. Get it scored against classic swing-trade criteria with a go, wait, or pass verdict.

What this does: Enter the ticker, the setup type, and describe what you're seeing: trend and price action vs moving averages, the base and volume behavior, fundamentals or catalyst, plus the overall market condition. The app scores it 0–100 with a GO / WAIT / PASS verdict and a conic score ring, then breaks down six weighted criteria (trend alignment, base quality, volume confirmation, relative strength, fundamentals, market context) each marked pass/partial/fail/unknown. You also get what's working, what's concerning, things to verify on the real chart before acting, and the disciplined next move. It honestly marks criteria "unknown" when you don't describe them and weighs a weak market heavily. Saves to localStorage. Educational tool, no live data.

What this looks like:

WHERE WE STAND(based on today’s news)

✅ AI Can Now: Teach a fleet of robots to do physical tasks like seating a graphics card with 99 percent accuracy, with no human supervising the training.

❌ Still Can't: Reliably tell the difference between fiction and fact, as Google's AI Overviews showed by presenting made-up horror creatures as real.

✅ AI Can Now: Handle the majority of a business's incoming phone calls, booking reservations and answering questions on its own.

❌ Still Can't: Replace the human touch in service work; even the businesses using it kept staff for complaints, complex requests, and callers who just want a person.

FROM THE WEB

RECOMMENDED LISTENING/READING/WATCHING

The Just City by Jo Walton - Book

The Greek goddess Athene decides to run Plato's Republic as a literal experiment, gathering 300 philosophers from every century in history and bringing them to a hidden island, where the heavy lifting is done by powerful robots brought back from a far-future civilization. Walton, a Hugo and Nebula winner, made the whole thing a Socratic dialogue, with the questions of slavery, consent, and personhood applying as much to the robots as to anyone else.

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.}