Nvidia leadership crisis and smuggling case & US names AI national champions - Tech News (Nov 16, 2026)
Nvidia faces a 170,000-signature push to bring back Jensen Huang, Anthropic's IPO draws huge demand, Apple tests health AI, and Vulcan stumbles.
15 free episodes a day, 5 editions, 3 languages, about $4 a day to produce. A coffee keeps it going ☕
Support the show →Our Sponsors
Today's Tech News Topics
-
Nvidia leadership crisis and smuggling case
— A petition with more than 170,000 signatures, backed by Elon Musk, Garry Tan and Tim Sweeney, urges Nvidia to reinstate Jensen Huang as CEO. Separately, an Nvidia manager was reportedly charged with helping smuggle AI chips to China through Southeast Asian data centers. -
US names AI national champions
— The Trump administration is reportedly preparing to designate Nvidia and other leading AI companies as US national champions. The move would deepen government support and tie frontier AI more closely to industrial and national security policy. -
India joins US tech supply pact
— India signed the San Francisco Declaration on securing critical technologies and AI supply chains. The US-led pact aims to reduce dependence on China for semiconductors and strategic tech inputs. -
Meituan trains AI on Huawei chips
— Meituan released the LongCat-2.5 AI agent model, reportedly trained entirely on 194 Huawei Ascend chips. It shows China's push to build large AI models without Nvidia GPUs amid export controls. -
Anthropic IPO demand and Samsung
— Anthropic and its backers are reportedly seeking up to $90.4 billion through an IPO and private financing, with demand several times oversubscribed. Samsung is also reportedly in early talks to host Anthropic and OpenAI in its AI data centers. -
Superintelligence timelines move closer
— Dwarkesh Patel and Benjamin Todd argue superintelligence or digital-worker AGI could arrive before 2030, citing AI accelerating its own research. A Cosmos Institute essay reframes recursive self-improvement as the whole AI development pipeline improving. -
Anthropic blueprint for always-on agents
— Anthropic published an architecture for always-on, proactive AI agents built on an event inbox, timers, memory, permissions and an outbox. It stresses safeguards like human approval for unattended tool use. -
Apple health AI study launches
— Apple launched the Apple Health AI Study with OpenEvidence, enrolling up to 200,000 US iPhone users to test conversational health AI. Participants can optionally share HealthKit data, and Apple says the data won't train AI models. -
Apple smart glasses built around Siri
— Apple reportedly plans to unveil smart glasses at WWDC 2027, with Siri controlling third-party iPhone apps by voice. The first version lacks full augmented reality and relies on a paired iPhone, competing with Meta and Google. -
Remote work hurts young graduates
— A Stripe Economics analysis finds work-from-home adoption, not AI exposure, best explains weaker employment for recent US college graduates. Reduced in-person mentoring may be limiting young workers' opportunities. -
AI floods hiring and complaints
— Investor John Hempton argues generative AI is overwhelming hiring with polished applications and making weak candidates look strong. He warns AI is also degrading public complaint systems like an Australian Ombudsman office. -
uBlock Origin critical vulnerability
— Researchers found a critical flaw in uBlock Origin's element picker that a malicious website could exploit to bypass the ad blocker. The vulnerability could expose user data in a widely used privacy tool. -
ULA Vulcan rocket failure
— United Launch Alliance's Vulcan rocket reportedly failed during its mission, a major setback for SpaceX's key competitor. The failure may delay commercial, civil and national security launches. -
Isotopes, not gold, from transmutation
— An essay argues the real prize behind Marathon Fusion's mercury-to-gold claim is cheap isotope production and separation. Isotopes like actinium-225, helium-3 and lithium-6 are bottlenecks for medicine, fusion and quantum computing. - 15
AI haptic headset for blind users
— University researchers built an AI-powered haptic headset that turns spatial information into touch sensations. The assistive technology could help blind users navigate their surroundings. - 16
Building a secure-by-default programming language
— A developer details building a new secure-by-default programming language with an interpreter and native backend that must produce identical results. It uses reference counting with ownership analysis instead of garbage collection.
Sources & Tech News References
- → India joins US-led Pax Silica initiative to secure AI supply chains
- → Trump administration plans to designate Nvidia and other AI companies as national champions
- → Anthropic reportedly seeks up to $90.4 billion through IPO and private placement
- → Reports allege Nvidia executive helped route AI chips to China through Southeast Asia
- → Meituan releases 1.6-trillion-parameter LongCat-2.5 model trained on Huawei chips
- → Apple launches health AI study with OpenEvidence-powered app
- → Critical uBlock Origin vulnerability enables website-driven cross-site scripting
- → ULA’s Vulcan rocket fails on debut, damaging US launch competition
- → Samsung negotiates to host Anthropic and OpenAI in new AI data centers
- → Apple plans smart glasses launch for late 2027 with Siri AI integration
- → Researchers develop AI-powered tactile headset to help blind people navigate
- → SambaNova Positions Its Platform Around Fast, Cost-Efficient AI Inference
- → Campaign Grows to Bring Jensen Huang Back as Nvidia CEO
- → Apple Bets on Siri AI for Smart Glasses and iPhone Duo 2
- → One Proton Away: The Real Value of Atomic Alchemy
- → Zendesk Pushes AI-First Service Platform and New Support Tools
- → Author Details Design of a New Secure Programming Language
- → How to Build Always-On AI Agents
- → Dwarkesh Patel Says AI Research Automation Could Bring Superintelligence by 2030
- → Work From Home, Not Just AI, Is Hurting Young Graduates
- → New Relic Report Warns of AI and Observability Gaps
- → Cloudflare Workers Promotes Global Serverless App Development
- → AI Is Breaking Hiring and Public Screening Processes
- → Apple Launches Health AI Study App to Track User Behavior
- → Recursive Self-Improvement and the Anatomy of AI Acceleration
- → Trend Extrapolation Could Put AGI Within 1-4 Years
Full Episode Transcript: Nvidia leadership crisis and smuggling case & US names AI national champions
More than 170,000 people, plus Elon Musk and a couple of other tech heavyweights, are now publicly demanding that the world's most important chip company take back the CEO who just resigned. We'll get into that in a moment. Welcome to The Automated Daily, tech news edition. The podcast created by generative AI. I'm TrendTeller, and today is November 16th, 2026. Let's get into it.
Nvidia leadership crisis and smuggling case
We start with Nvidia, where the leadership saga keeps getting stranger. An online petition to reinstate Jensen Huang has passed 170,000 signatures, amplified by Elon Musk, Y Combinator's Garry Tan, Epic's Tim Sweeney, and a number of current and former employees and investors. Reports say Huang has privately discussed a return, but the board's independent investigation is still underway, and lawmakers are preparing inquiries into possible national security implications. Nvidia says it hasn't formally engaged any CEO candidates, though Arm chief Rene Haas is reportedly on the list of outside options. Analysts are divided: some think a comeback would calm investors, others think the investigations make it a long shot. And in a separate development, an Nvidia manager has reportedly been charged with helping smuggle advanced AI chips to China via data centers in Southeast Asia, which will only increase pressure on how the company's hardware moves around the world.
US names AI national champions
Despite the turbulence, Washington appears ready to lean in harder. The Trump administration is reportedly preparing to name Nvidia and other leading AI firms as US national champions. That kind of label usually means more government backing and closer coordination, and it's another sign that frontier AI is now treated as industrial and security policy, not just a business sector.
India joins US tech supply pact
That geopolitical framing shows up elsewhere too. India has signed the San Francisco Declaration, a US-led agreement on securing critical technologies and AI supply chains. The goal is to reduce reliance on China for strategic inputs like semiconductors. Meanwhile, China is demonstrating it can work around those restrictions. Meituan released LongCat-2.5, an AI agent model reportedly trained entirely on domestic hardware, using 194 Huawei Ascend chips and no Nvidia GPUs at all. It's a concrete data point in China's effort to build a self-sufficient AI stack.
Meituan trains AI on Huawei chips
On the money front, the Anthropic financing story has a new chapter. The company and its backers are reportedly seeking up to 90.4 billion dollars through a mix of an IPO and private funding, and demand is said to be several times larger than the shares on offer. That's a strong signal of investor appetite for frontier AI, at least for now. Separately, Samsung's IT division is reportedly in early talks to host both Anthropic and OpenAI as tenants in AI data centers it's building, which would give the Korean giant a direct line into the compute race.
Anthropic IPO demand and Samsung
The debate over how fast AI is moving continues to heat up. Podcaster Dwarkesh Patel now says superintelligence could arrive by 2030, possibly as early as 2027, mainly because AI systems already appear to be speeding up AI research itself. Benjamin Todd lands in a similar place, arguing that rising AI-company revenue and models handling ever-longer tasks point to digital-worker AGI within two to four years. And an essay from the Cosmos Institute offers a useful reframing: the thing improving itself isn't a single model, but the entire research and training pipeline, with progress still bound by experiments and resources rather than magic.
Superintelligence timelines move closer
Staying with AI, Anthropic published a practical guide to so-called always-on agents. The key idea is that these agents aren't actually running nonstop. They sleep, wake up when something happens, do a short burst of work, and go quiet again. The piece puts a heavy emphasis on permissions, recommending that risky actions wait for human approval when nobody's watching. It's a helpful reality check on what proactive AI assistants will look like in practice.
Anthropic blueprint for always-on agents
Over at Apple, two stories. First, the company launched the Apple Health AI Study, which could enroll up to 200,000 US adults. Participants use an app powered by OpenEvidence to ask health questions, with answers backed by medical citations, and can optionally share their HealthKit data. Apple wants to learn whether people understand and trust AI health answers, especially when preparing for doctor visits, and says the data won't be used to train models. Second, Apple's smart glasses are reportedly targeted for a WWDC 2027 unveiling and a late 2027 release. The pitch centers on Siri controlling popular iPhone apps by voice, so you could book a ride or send a message without pulling out your phone. There's no full augmented reality in the first version, and the glasses will lean on a paired iPhone. Apple is betting its developer ecosystem gives it an edge over Meta and Google.
Apple health AI study launches
Now to work and careers. A new analysis from Stripe Economics finds that the tougher job market for recent US college graduates is better explained by remote work than by AI. Occupations that shifted more to working from home saw weaker employment growth for young grads, likely because in-person mentoring and feedback got thinner. AI exposure, by contrast, showed little measurable effect so far, though the authors caution that could change. On the hiring side, investor John Hempton argues AI is breaking recruitment from the other direction: floods of polished applications and AI-assisted interview answers make weak candidates look strong. He sees the same problem hitting public complaint systems, where low-quality grievances now sound credible.
Apple smart glasses built around Siri
A quick security note: researchers found a critical vulnerability in uBlock Origin's element picker. A malicious site could exploit it to bypass the blocker and potentially access user data. If you rely on uBlock, keep an eye out for an update.
Remote work hurts young graduates
In space, United Launch Alliance's Vulcan rocket reportedly failed during its mission. That's a serious setback for SpaceX's main US competitor and could ripple through commercial, civil, and national security launch schedules while investigators look for the cause.
AI floods hiring and complaints
A thought-provoking essay argues that the real story behind Marathon Fusion's claim of turning mercury into gold isn't the gold at all. It's the possibility of cheaply making and separating specific isotopes, materials that medicine, nuclear power, fusion, and quantum computing desperately need but often source from aging reactors or a handful of foreign suppliers. Modern alchemy, in other words, might be about supply chains, not treasure.
uBlock Origin critical vulnerability
In assistive tech, university researchers built a headset that uses AI and haptic feedback to turn the space around a person into touch sensations, helping blind users sense and navigate their surroundings. Broader testing will tell whether it's ready for everyday use.
ULA Vulcan rocket failure
And finally, for the compiler fans: one developer shared the story of building a new programming language that's secure by default and compiles only what a program needs. A neat detail is that every test runs through both an interpreter and a native compiler, and the results must match exactly. The language still doesn't have a name, which might be the hardest problem of all.
That's it for today's edition of The Automated Daily. From Nvidia's boardroom drama to Apple's glasses and a rocket that didn't make it, it's been a busy one. Thanks for listening, I'm TrendTeller, and I'll catch you tomorrow.
More from Tech News
- November 14, 2026 Microsoft Outlook, Teams, Copilot Outage & FAA Ground Stop and AI
- November 13, 2026 Jensen Huang Resigns From Nvidia & Chinese AI Firms Seek Overseas Compute
- November 12, 2026 TSMC invests in Anthropic, OpenAI & CoreWeave revenue surges 168 percent
- November 11, 2026 Google launches Gemini 4 Carbon & Gemini off by default for students
- November 10, 2026 Google TPU 8 chips challenge Nvidia & Rapidus falls behind 2028 target