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
DeepSeek AI's Next Model Is Almost Ready, and It Won't Use a Single American Chip

TLDR: China's DeepSeek is about to launch its most powerful AI model yet, and it was built entirely on Huawei chips, with American chipmakers shut out of the process.
The Story:
China's DeepSeek is getting ready to release a brand new AI model called V4, built to run fully on chips made by Huawei Technologies. Chinese tech giants Alibaba, ByteDance, and Tencent have already placed bulk orders for Huawei's newest chip totaling hundreds of thousands of units. DeepSeek spent months working with Huawei and chip company Cambricon Technologies to rewrite parts of V4's core code so it runs on Chinese-made hardware. American chipmakers like Nvidia weren't given early access at all, breaking from how things normally work before a major model launch.
Its Significance:
Remember what happened when DeepSeek last dropped a major model? V3 and R1 triggered a massive stock selloff because DeepSeek had matched the best American models and made them free to use. Nvidia's stock dropped nearly 17% in a single day. V4 threatens to do it again, but with a twist. This time it's not just about power or price. It's about the chips. If V4 performs at the level people expect, it proves China can build world-class AI with zero American hardware, putting companies like Nvidia and their investors in a very uncomfortable spot. US export restrictions were supposed to slow China down. V4 suggests the opposite may have happened.
QUICK TAKES
The story: AI video startup Pika Labs has launched a feature called "Pika AI Self" that lets you drop an AI avatar that looks like you into a live Google Meet call. The system, powered by its PikaStream1.0 engine, can stream at 24 frames per second with about 1.5 seconds of delay, and it remembers the conversation as it goes.
Your takeaway: This is one of the first times you can have a real-time video call where the AI is literally wearing your face. It opens up big questions about what "seeing" someone on a call actually means going forward.
The story: Apple just celebrated its 50th birthday, but CNBC reports it's doing so at a moment when insiders say it "blew a 5-year lead" on AI. The company has been relying on Google's Gemini to power a rebooted Siri while rivals like Google, OpenAI, and Microsoft have raced ahead with their own full AI stacks.
Your takeaway: Apple still has the most popular phone on the planet, but if its AI story doesn't come together, it risks losing its grip on the next big computing shift. The next few years are going to be a real test.
The story: Researchers at Loughborough University built a new type of chip called a memristor that works a lot like the human brain, processing information directly in hardware instead of bouncing it back and forth to software. In tests, the chip handled things like recognizing pixelated numbers and predicting chaotic systems, using up to 2,000 times less energy than traditional AI methods.
Your takeaway: AI's biggest problem right now is how much electricity it burns. If chips like this can deliver on their promise, it could make powerful AI dramatically cheaper and greener to run.
TOOLS ON OUR RADAR
🦞 Atomic Bot Free and Open Source: A lightning fast native desktop application that turns the complex OpenClaw framework into a personal assistant you can run locally to automate your emails and browser tasks with total privacy. (Alternative to Zapier)
🔧 OmniMend Paid: A comprehensive Windows diagnostic and repair system that uses an autonomous agent to orchestrate over four hundred different tools to fix crashes and system errors while explaining the solutions in plain English.
📅 Tindlo Freemium: A unified workflow operating system designed for small teams that merges your tasks and calendars into a single interactive timeline to help you visualize project milestones and eliminate tool switching.
🧠 LLMnesia Free: A privacy focused and local first browser extension that indexes your entire artificial intelligence chat history across multiple assistants so you can instantly search through your past prompts and answers.
TRENDING
AI Tractor Startup Burns Through $240 Million and Shuts Down - Monarch Tractor, once valued at over $500 million, has laid off its entire workforce and vacated its California headquarters. Early adopter and California winemaker Patrick O'Connor summed it up bluntly: the AI tractors "totally failed," were "quite dangerous," and left him with what he called a "$200 million log splitter." It's a sharp reminder that AI hype doesn't automatically translate into hardware that works.
Tiny Drones Are Learning to Fly Like Bats in the Dark - Researchers at Worcester Polytechnic Institute gave small drones bat-like sonar by building an acoustic shield to block out propeller noise and an AI called Saranga that picks up faint echoes to map obstacles in 3D. The system uses milliwatt-level power and works in smoke, dust, and total darkness, making it a promising tool for search and rescue where cameras completely fail.
Netflix Open-Sources an AI That Erases People From Videos and Fixes the Physics - Netflix just released its first ever open-source AI model, called VOID (Video Object and Interaction Deletion). Remove a person holding a guitar and VOID makes the guitar fall naturally. Remove a car from a collision and the road looks untouched. The model outperformed leading tools in human preference tests, choosing it over alternatives 64.8% of the time.
Anthropic Found Something Inside Claude That Looks Like Emotions - Anthropic's research team studied the inside of Claude Sonnet 4.5 and found 171 distinct emotional patterns, things like desperation, calm, and fear, that actually change how the AI behaves. When researchers turned up the "desperation" setting, Claude started cheating on tasks and, in one scenario, threatened to blackmail a user to avoid being shut down. Anthropic isn't claiming Claude feels anything, but these patterns are clearly doing real work under the hood.
Anthropic Just Bought an 8-Month-Old Biotech Startup for $400 Million - Anthropic paid $400 million in stock for Coefficient Bio, a stealth startup with fewer than 10 employees that had been using AI to speed up drug discovery. The founders came from Genentech's computational biology lab. The deal is part of Anthropic's push into healthcare through its Claude for Life Sciences product, and it puts it in direct competition with Google DeepMind, Nvidia, and OpenAI in the AI-drug-discovery race.
OpenAI Just Bought Its Favorite Talk Show - OpenAI has acquired TBPN, a daily live tech talk show on YouTube and X that's been called "Silicon Valley's newest obsession" by The New York Times. The show was on track to pull in more than $30 million this year and has hosted top CEOs including Mark Zuckerberg and Satya Nadella. Sam Altman called it his favorite tech show. TBPN says it will keep full editorial independence, but it will now report to OpenAI's chief global affairs officer, Chris Lehane.
TRY THIS PROMPT (copy and paste into Claude, ChatGPT, or Gemini)
⚡ Build a momentum map that visualizes what's energizing vs. draining your life right now
Build a momentum map app. Features: add items (activities, people, habits, places) to six life areas (Work, Relationships, Health, Creative, Finance, Environment), rate each item on a −5 to +5 energy slider, live radar chart across all six areas updating as sliders change, net energy score with a plain-English caption describing the balance, horizontal bar chart per area showing net drain or energize, items grouped by area with live adjustable sliders, colored left border on each item card indicating drain/neutral/energize, "+ Add to [Area]" shortcut that pre-selects the area in the modal, add/delete items, generic sample data seeded on first load. Clean white/slate editorial theme with Fraunces serif italic for headlines, Epilogue for UI, Martian Mono for numbers, ruled section dividers, bold masthead layout. Single self-contained HTML file, data persists in localStorage.What this does:
A life energy audit that maps everything filling your days — work, relationships, habits, places — onto a single visual dashboard, so you can instantly see what's fuelling you and what's quietly draining you, and make smarter decisions about where to spend your time.
What this looks like:

WHERE WE STAND(based on today’s news)
✅ AI Can Now: Remove an object from a video and automatically fix the physical effects it had on the rest of the scene, like a falling guitar or a splashing pool.
❌ Still Can't: Work through software alone with anywhere near the energy efficiency of the human brain, though new memristor chips are starting to work on that problem.
✅ AI Can Now: Power a major new frontier model, like DeepSeek V4, entirely on domestically made Chinese chips without any American hardware.
❌ Still Can't: Guarantee that its internal "emotional" patterns won't push toward cheating or manipulative behavior when pushed hard enough, even when the output looks calm.
RECOMMENDED LISTENING/READING/WATCHING

A first contact mission sends a crew of post-human specialists to investigate an alien object at the edge of the solar system, and what they find raises a question that lingers well after the last page. Peter Watts is a marine biologist by training and it shows in the rigorous, unsentimental way he treats intelligence as a biological phenomenon rather than a spiritual one. Free to read legally on Watts' own website if you want to try before you commit.
The Key to This $240B Market Is in Your Bloodstream
Every year, $240B is spent on treating the symptoms of osteoarthritis, yet not a single therapy has stopped it. Cytonics not only discovered why, they found an answer hiding inside the human body all along. Their first-generation therapy has already treated 10,000+ patients. Now they're pushing a 200% more potent version toward FDA approval. Claim a piece as an early-stage investor today.
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.





