It's Friday, November 13th: Welcome to The Stress Test 🔍
Two CEOs, two scandals, two very different fallouts. This week's Stress Test is on why Jensen Huang's exit rattled all of AI, and what it says about how much the industry leans on one company.
🌟 Community Spotlight: Nebius
Interested in building production AI agents for generative media? Join Nebius's new course to learn how to build, deploy, evaluate, optimize, and monitor AI systems that orchestrate GPU-powered image, video, and audio models on Nebius Serverless AI. The course begins November 16.
🔍THE STRESS TEST
The Coldplay CEO Was a Meme. Jensen Huang's Exit Was an Industry Event.
One safety story a week, pressure-tested for what's actually happening underneath the surface.
You probably saw the memes. In July 2025, a kiss cam at a Coldplay concert caught Andy Byron, CEO of the data company Astronomer, with his arms around the company's head of HR. Within days he had resigned and a co-founder had taken over as interim CEO. About a week later, Astronomer was poking fun at the whole thing in an ad starring Gwyneth Paltrow, Coldplay frontman Chris Martin's ex-wife. Hardly anyone outside the company had to change anything.

This week Nvidia had its own CEO scandal. On Sunday, Jensen Huang acknowledged being unfaithful to his wife, and the board opened an investigation into reports he had been targeted by a foreign intelligence effort. By Thursday morning he had resigned, and every AI lab running on Nvidia chips had a stake in a board decision it had no say in.
The headline version is that one man's personal life shook AI. Let's stress-test the headline.
Start with Huang. He ran Nvidia for 33 years, and Fortune reported that he concentrated decision-making in his own hands. Until Thursday the board had never had a chair, and the company had no clear public succession plan. Its own annual report says it is "highly dependent" on its executive team and that "failure to ensure effective succession planning" could "hinder our strategic planning, execution and long-term success," CNN reported. So part of the headline holds up.

Then look at what the board did. Within a day it named CFO Colette Kress, at Nvidia since 2013, interim CEO, appointed its first-ever chair, and said it expects operations and the product roadmap "to continue without interruption." The people to run a transition were there all along. What was missing was a plan anyone outside the boardroom could see.
Now look past Nvidia. The reason this wasn't just a business story is that Nvidia has been estimated to control about 90% of the market for high-powered GPUs used in AI data centers, and many AI companies have standardized on its CUDA software, per Fast Company. Labs are diversifying, and Anthropic says it runs on Google, Amazon, and Nvidia chips, but that shift is slow.
In trust and safety, the first question you ask about any system is what happens when one piece fails. AI companies ask it about their models all the time. The supply chain under those models got far less of that scrutiny.
So Huang's exit was the trigger, but the reason it mattered is that the industry concentrated its most important input in one company, and that company concentrated its leadership in one person. Astronomer could joke about its scandal because so little depended on it in comparison.
The thing to watch is Nvidia's November 17 earnings call, its first without Huang, and whether labs start talking about supplier risk the way they talk about model risk.
The longer version, including what Nvidia's own filings warned about, why switching chips isn't as simple as buying different ones, and what the Coldplay CEO actually tells us, is on Hidden Layer this week.

Each week, we highlight AIC chapters doing groundbreaking work with their members around the world. Tag us on socials to be featured!
🇸🇬 Singapore | Code Meets Animation Puts AI Motion Tools in Builders' Hands

Image from Emily Nguyen
On Thursday, November 12, AI Collective Singapore teamed up with LottieFiles, 65Labs, Lorong AI and community partner Noticebob for Code Meets Animation at Lorong AI @ One-North. The hands-on afternoon focused on taking AI-generated animation into real products. The rundown included a first look at LottieFiles' AI tools, a CodeWithAI talk on using AI as a creative and coding partner, and a live Lottie AI Challenge where attendees made animations against the clock for prizes.
Emily Nguyen gave her first-ever demo at the event, showing Sound of Color, an app that turns music into visuals inspired by synesthesia. She was candid about the realities of live demos, too: hers took 20 minutes to load. Another attendee wrote that LottieFiles' Creator MCP stuck with him most, since it gives an AI the editing tools for an animation, which he said leads to better quality, smaller files and fewer tokens.
Thank you to LottieFiles, 65Labs, Lorong AI and Noticebob for building this with our Singapore chapter. Taipei is next: our Taipei chapter hosts its own Code Meets Animation with LottieFiles on Tuesday, November 17.
💻 Online | Lunch & Hack Explores the Browser as an AI Runtime

Image from Adelina Martiniuc
On Thursday, November 12, our biweekly virtual Lunch & Hack asked what changes when AI becomes a native capability of the browser. Opera's Product Advocacy Director walked through Web AI, which makes it possible to run AI directly in the browser, and HTML-in-Canvas, which opens up new ways to build interactive web experiences inside 2D and 3D environments. For the hack, attendees were invited to build an AI-native web experience of their own in Opera One with the Web AI API turned on.
Adelina Martiniuc, CEO of RoryPlans.ai, spoke alongside Opera and posted about the session afterward. The full recording is on our YouTube channel, and its description sums up the appeal: models run on your own computer's hardware, and private data stays in the tab. Past sessions, upcoming topics and a form to propose new ones live at RoryPlans.ai/Lunch. Thank you to Opera for joining us.
🎙️ This Week on The AI Collective Podcast
Week four of the show, hosted by AJ Green and Catherine McMillan, moved from chip design to community to agent security. On Monday, Hannah Earley, Founder and CTO of Vaire Computing, joined to talk about reversible chips that recycle energy usually lost as heat. Monday's lineup also listed guests from qBotica, on moving from automation to agentic AI, and Google DeepMind, on developer experience for generative AI.
On Tuesday, Silka Hsu, Head of Partnerships at Boardy, talked about Boardy and leading partnerships, with the lineup also listing Tezlo, a voice-first AI companion for Tesla owners, and Utiva, on scaling AI startups in emerging markets. Wednesday's lineup featured GenseeAI on protecting what AI agents can access and Humanoid's CFO on leading a European startup through hypergrowth. Thursday was a shorter news show, with a rundown that included n8n Labs' Nod, Runway's Praxis-1, the Gemini Decisions API and AWS Emulator reaching general availability.
Today at 9 AM PT, the lineup features guests from Okta, on making data ready for enterprise AI agents, and initializ, on authorizing AI agents before they act. The show is live every weekday at 9 AM PT on LinkedIn, YouTube, X and Instagram, and every episode is on our YouTube channel.
🫵 Want your message in front of 500,000 AI builders?
Our partners and sponsors get exclusive placements across the newsletter and access to AIC's in-person network: demo nights, dinners, hackathons, and forums across 180+ chapters.
For all inquiries, send us a note at [email protected].
The AI Collective is built by volunteers across 180+ chapters in 40 countries.
Thank you to the thousands of volunteers around the world who make this work possible. We truly could not do this without you.
🧑💻 About the Editors


About Lindsay Gross
Lindsay is an AI engineer, researcher, and writer focused on how AI systems behave in practice and what it takes to make them safe. Her work sits at the intersection of AI safety, governance, and product design. At AIC and in her newsletter, Hidden Layer, she writes about the questions that matter most as these systems scale.

