|
|  |  | Tuesday, December 30, 2025 | |  | | Michael Nagle/Bloomberg via Getty Images | Good morning, Quartz readers! It’s Shannon Carroll with the Daily Brief. Today, the road to $5 trillion gets more crowded, OpenAI puts “Preparedness” on the org chart for a hefty chunk of change, Musk says silver’s rise is “not good,” and the DOJ reaches for fraud law to make inclusive hiring a legal risk. | | HERE'S WHAT YOU NEED TO KNOW | The “godfather of AI” puts a date on AI job losses:
soon. Geoffrey Hinton says models are improving fast enough to replace more white-collar work in 2026, and he worries safety and regulation are falling behind. | Silver hits a record — and Musk says the rally is “not good.” With China’s export restrictions starting Jan. 1, the metal’s record $80/ounce run is threatening to raise costs for solar, EVs, and data centers. | Intel has officially cashed Nvidia’s $5 billion check. The late-December closing makes Nvidia’s 4% stake official — and leaves Intel to prove that this “historic partnership” can turn filings into finished chips. | The DOJ is using fraud law to police DEI. The Trump administration is using a statute usually aimed at contractor scams to pursue cases tied to inclusive hiring at federally funded firms, raising the threat of steep penalties. | | | SPONSORED |  | 65% off a Lifetime of Learning | Start 2026 investing in yourself—and your career. With Babbel, 10 minutes a day is all it takes to add a new language to your skillset.
Get 65% off a Lifetime subscription and unlock bite-sized lessons, podcasts, and interactive speaking practice—practical tools that pay off in global opportunities and professional growth. | | |
| | MARKET: WEIGHT AND SEE | Trillion-dollar market caps used to feel like folklore — the sort of number you’d whisper, then immediately add,
“with a T.” In 2025, $1 trillion, $2 trillion, $3 trillion, and beyond became routine. A handful of tech companies ballooned into index-heavy monuments, fueled by explosive growth in consumer tech, cloud computing, and AI, plus enormous inflows from ETFs, money managers, pension funds, sovereign wealth funds, and the broader global investor class. The loudest proof arrived in October of this year, when Nvidia became the first publicly traded company to surpass a $5 trillion valuation, powered by its central role in AI workloads and the demand
for its chips. A few years ago, “hardware business” and “$5 trillion” didn’t belong in the same story; now, that’s basically the lede.
Beneath that, the $4 trillion tier filled out. Microsoft crossed $4 trillion midyear on accelerating cloud and AI services growth and strong enterprise demand for AI-integrated software. Apple rose above $4 trillion late in the year on the back of device sales and services revenue. Alphabet, late in the year, briefly brushed the $4 trillion mark, too,
reminding everyone it still owns a ridiculous share of the modern internet. The rest of the Magnificent Seven stayed enormous but a little less magnificent: Amazon sat around $2.4 trillion, with AWS and long-run bets in logistics and AI-driven commerce powering the story; Tesla and Meta remained firmly in the upper tier but below $2 trillion.
So who’s next to $5 trillion? The annoyingly sensible answer is the obvious one: the companies already at, or flirting with, $4 trillion.
Microsoft has the cloud machine, deep enterprise demand, and increasing AI integration that can keep compounding as long as earnings growth and multiple expansion hold. Apple’s ecosystem throws off recurring revenue from services and hardware, and AI-enhanced products could become a real catalyst if innovation (finally) accelerates enough to satisfy expectations. Alphabet has AI strength and a broad set of defensible franchises. And looming over the whole race is the slightly unsettling math of modern markets: scale attracts index weight, index
weight attracts passive ownership, and passive ownership can turn giants into self-reinforcing winners, amplifying fundamentals through the structure of modern capital markets — making the next $5 trillion behemoth far more likely to come from the giants with their names already in lights. Quartz’s Catherine Baab has more on the market’s $5 trillion era. | | | RECOMMENDED READING | The
future of tech, decoded.
Semafor Technology unpacks the ideas, innovations, and power shifts redefining the global tech landscape. From AI and machine learning to the startups and policies shaping
the industry, each briefing delivers clarity and depth on the tech transforming our world
|
| | | READY OR BOT | OpenAI just put a comp package on AI worries: $555,000 (plus equity) a year for a Head of Preparedness, the person tasked with turning “Well, what could go wrong?” from a conversation into a constraint. The posting
describes running the technical strategy behind OpenAI’s Preparedness Framework, the system that ties together capability evaluations, threat models, and mitigations. OpenAI calls it an “operationally scalable safety pipeline,” which is a very corporate way of saying the checks have to survive a product sprint. The job exists to make risk review repeatable, documented, and hard to wave away on launch day — factory-floor safety built for a product that ships fast and lands everywhere.
This person will “track” frontier capabilities that create “new risks of severe harm,” own evaluations and safeguards, and keep policy monitoring and enforcement in the loop across cybersecurity, biological and chemical risks, and AI self-improvement. CEO Sam Altman has framed this as a “stressful job,” promising the hire will “jump into the deep end pretty much immediately,” as OpenAI talks publicly about models that can mess with mental health and models that can surface “critical vulnerabilities.” Even the internal debate is on the
record: A former safety leader wrote in 2024 that “safety culture and processes have taken a backseat to shiny products,” and OpenAI’s updated framework leaves room to “adjust” safety requirements if a rival ships a high-risk model without similar protections.
The public, meanwhile, is done grading on a curve. More Americans are more concerned than excited about AI; want safety and data-security rules even if they slow development; and say trust in fair, unbiased AI is basically a
rounding error. OpenAI has also been tightening how ChatGPT behaves in sensitive situations, explicitly calling out “psychosis or mania” and “emotional reliance on AI” — as chatbots slide into a therapy-adjacent role that people didn’t exactly consent to beta-test — and the broader conversation turns toward lawsuits and consequences. If the next era of AI ships faster, the blame will also arrive faster — and OpenAI is hiring a name for the clipboard. Quartz’s Shannon Carroll has more on how “Preparedness” became a job, not a blog post. | | | SPONSORED |  | 65% off a Lifetime of Learning | Start 2026 investing in yourself—and your career. With Babbel, 10 minutes a day is all it takes to add a new
language to your skillset.
Get 65% off a Lifetime subscription and unlock bite-sized lessons, podcasts, and interactive speaking practice—practical tools that pay off in global opportunities and professional growth. | | |
| | MORE FROM QUARTZ |
|
|
|
