Good Morning from San Francisco,
🍎 Oops, they did it again. Apple's grand AI promises just pulled a Houdini - vanishing faster than your battery life at 1%. While Siri takes another gap year, Google's Gemini actually delivers, and Huawei's phones are stealing Chinese hearts with AI that, surprise surprise, works.
💫 Intel just hired the unlikeliest CEO imaginable: Lip-Bu Tan built his fortune helping startups dodge Intel's shadow. His previous company's stock soared 3,200%. Now he faces an ironic challenge - rescuing the behemoth he spent decades outmaneuvering. 🎯
Stay curious,
Marcus Schuler
Apple's AI Plans Go Dark: The Siri Features That Never Were
Apple just deleted its AI promises. A YouTube ad vanished. Website claims disappeared. Reality bit hard.
The missing Siri upgrade promised magic: finding old flight details, scanning messages for meeting notes, taking smart actions across apps. One snag: No one outside Apple ever saw it work.
Meanwhile, competitors ship real features. Google's Gemini handles complex tasks. Amazon's Alexa summarizes books. Samsung and Huawei pack their phones with working AI.
The timing stings. iPhone sales in China dropped 18% last quarter as Huawei reclaimed the lead. Consumers chose working features over empty promises.
John Gruber of Daring Fireball cuts deep: "This didn't happen to Apple. Decision makers within the company did it." He compares this to Apple's dark pre-Jobs era, when marketing dreams replaced shipping products.
Apple's new timeline? Somewhere between late 2025 and never. The same features they advertised for iPhone 16 might miss iPhone 17.
Why this matters:
- The company that perfected "It just works" now advertises features that don't exist. That trust gap won't close easily.
- In tech, second place rarely matters. Just ask BlackBerry how catching up worked out.
Read on, my dear:
- John Gruber (Daring Fireball): Something Is Rotten in the State of Cupertino
- implicator.ai: Apple's AI Dreams Hit Reality: The Siri Features That Weren't
AI Photo of the Day
Prompt:
extreme beauty female
The Startup King Takes Intel's Throne: Lip-Bu Tan's Impossible Task
Intel just picked its most unlikely CEO ever. Lip-Bu Tan built his fortune backing tiny startups that competed with Intel. Now he must save the chip giant from itself.
The timing feels surreal. Intel's stock dropped 60% this year. Its AI strategy flopped. Its factories sit half-empty while competitors surge ahead.
Enter Tan, a Malaysian-born venture capitalist who turned $370 million startup bets into billion-dollar wins. His secret? Finding small teams who could outmaneuver giants like Intel.
Consider his track record:
👉 Annapurna Labs, which Amazon bought to replace Intel chips
👉 Invested in Nuvia, now powering Qualcomm's attack on Intel's PC business
👉 Just this week, funded an AI chip startup backed by Intel's rival AMD
The man moves fast. On Monday, he invested in a startup competing with Intel. By Wednesday, Intel named him CEO.
His last turnaround tells the story. As CEO of Cadence Design Systems, Tan focused on one thing: helping companies design chips they'd make at TSMC, Intel's biggest rival. The stock jumped 3,200%.
Now he faces his biggest challenge. Intel employs 130,000 people, runs massive factories, and bleeds cash trying to catch up in AI.
His letter to employees hints at his plan: "Restore Intel's position as a world-class products company, establish ourselves as a world-class foundry."
Translation: Fix the core business first. Stop dreaming about AI domination.
Wall Street loves it. Intel stock jumped 12% on the news. But Tan's real test starts March 18th, his first day on the job.
Why this matters:
- Intel's new CEO made his career helping companies escape Intel's grip. Now he must convince those same companies to trust Intel again.
- In an industry obsessed with AI hype, Tan brings old-school focus on execution. Sometimes the best innovation is just making things work.
Read on, my dear:
reuters: Intel's new CEO Lip-Bu Tan has a history as a successful underdog
AI & Tech News
OpenAI Wants Trump to Tame State AI Rules
OpenAI asked the Trump administration to protect AI companies from state regulations in exchange for sharing their technology with federal agencies. In a policy proposal Thursday, the ChatGPT maker warned that hundreds of pending state bills could slow U.S. AI development while China speeds ahead. The company suggested that firms who voluntarily work with the U.S. AI Safety Institute should get immunity from state rules - making federal oversight their only concern.
US Tech Firms Face Trump's New Tariff Storm
Trump slapped fresh tariffs on China, Mexico, and Canada in January, forcing US tech firms to swallow hefty import taxes or jack up prices. The move hits electronics particularly hard - China still supplies 78% of US smartphones and 79% of laptops, while companies like Austere scramble to stockpile inventory before costs climb even higher.
CISA Staff Struggle as Political Storm Hits Security Work
Wired reporter Eric Geller reveals how mass layoffs and leadership turmoil have crippled America's cyber defense agency CISA, with over 300 experts pushed out and critical partnerships fraying. The agency's acting director Bridget Bean now prioritizes pleasing Trump over protecting staff, while remaining employees juggle double workloads and watch international threats slip through widening cracks in US digital defenses.
AI Firm Takes Shot at Drug Research, Backed by Big Money
Insilico Medicine raised $110 million from investors betting on AI-powered drug discovery, pushing its value past $1 billion. The startup, which mixes US research with Chinese labs, has caught attention by using AI to speed up the slow, expensive process of finding new medicines - though the field's bold promises remain unproven.
Alibaba's AI App Gets a Brain Boost
Alibaba supercharged its Quark mobile app with new AI powers, cramming a chatbot, reasoning engine, and task manager into one sleek package. The revamped app, powered by Alibaba's Qwen AI model, can now write articles from images and even generate meeting minutes for those who dozed off during the presentation.
Gates Sees AI Taking Over Everything (Except Baseball)
Bill Gates thinks AI will spread like PCs did - from corporate offices into every home. Speaking on a podcast, the Microsoft founder predicts we'll all soon have personal AI agents, though he mercifully spares us from watching robots play America's pastime. Despite recent doom-scrolling about AI's limits, tech leaders like Sam Altman and Eric Schmidt insist the technology isn't hitting any walls.
Paris Tech Hub Serves Pizza with a Side of Startup Success
Financial Times reporter Éanna Kelly reveals how Station F, Paris's mammoth 34,000-square-meter startup incubator, has transformed from housing locomotives to launching unicorns, with its companies raising over €1 billion annually - more than the entire Italian tech scene. The former train station now packs its space with cargo-container meeting rooms and a Space Invaders arcade, while its restaurant La Felicità keeps founders fueled with pizza and Tuesday karaoke sessions - because apparently, even AI entrepreneurs need their ABBA fix.
AI Joins Academia, Struggles With Bibliography Like Everyone Else
An AI system just wrote a scientific paper good enough to fool peer reviewers. The machine did everything - from crafting hypotheses to running code to making those fancy graphs scientists love.
Japanese startup Sakana's AI Scientist-v2 submitted three papers to a prestigious machine learning workshop. One made the cut, scoring better than many human submissions. The accepted paper explored neural network training, though the AI apparently struggled with basic academic etiquette, crediting modern researchers for decades-old work. (Even robots, it seems, can't resist a little academic politics.)
The system wrote the entire paper without human tweaks. Company researchers just picked which papers to submit - like academic talent scouts for artificial intelligence.
But before you imagine robot professors taking over campus, note the caveats: This was a workshop paper, not a major conference publication. And Sakana withdrew it before publication, perhaps sparing us all from AI's first faculty lounge drama.
Why this matters:
- An AI just passed peer review, academia's gold standard. That's like a robot getting tenure, minus the parking spot.
- The real question isn't whether AI can play scientist - it's whether it can make discoveries that matter. Let's see it tackle cancer before we hand over the lab keys.
Read on, my dear:
AI Decoded 🔓: How Sora, Veo 2 and Gen-3 Work
The AI revolution storms into video production. Three heavyweights battle for supremacy - with tools that sound ripped from a sci-fi movie.
- Turns text into videos up to one minute long
- Runs on DALL·E 3 tech and a "diffusion transformer" (sounds complex because it is)
- Still trips over basic physics - water splashes look like slow-motion pudding
- Ships with built-in guardrails: No violence, sex, or hate speech
- Cranks out two-minute videos
- Guards its tech specs like nuclear codes
- Limits access to US users over 18 - everyone else waits outside
- Google plays hard to get, again
- The underdog with sharp teeth
- Targets pros: filmmakers and content creators
- Claims "photorealistic humans" - sure, whatever you say
- Did its copyright homework
These tools mark a turning point: Video production goes mainstream. Soon, nobody will need a film studio - just a smartphone and a good idea. Well, maybe that's stretching it a bit.
