Tech News

April 04, 2026 at 07:01 AM

Artificial intelligence is reshaping education, healthcare, law enforcement, and politics all at once, with some genuinely inspiring breakthroughs buried alongside some genuinely troubling ones. The good news first: 🚀 THIS IS COOL Professors at Columbia, Georgia Tech, Arizona State, and Stanford are building AI tutors that do the opposite of what ChatGPT does—instead of giving answers, they ask better questions. Dan Wang's app "Caisey" forces students to argue through case studies before class, recreating the centuries-old Oxbridge tutorship model at scale. Similar tools are helping engineering students solve complex problems, health sciences students practice with simulated patients, and foreign language learners get personalized feedback. The philosophy here is radical: slow students down, make them think harder, deepen understanding. It's a rebuke to the efficiency-obsessed AI narrative and a reminder that education's real purpose isn't information delivery—it's intellectual growth.

But the same week Arizona cold case units cracked decades-old investigations using DNA advances and cell phone forensics, a Utah healthcare provider won regulatory approval to let an AI chatbot prescribe psychiatric medication. 🤔 THINK ABOUT IT How comfortable are you with an algorithm deciding whether you get SSRIs refilled, even if it's "only" lower-risk medications? Legion Health's Doctronic chatbot will ask 15 questions and cost $20 per refill—cheaper and faster than a doctor's appointment, which is exactly the problem psychiatrists like Harvard's John Torous are flagging. The system will eventually phase out human review, which means real autonomy for an AI system making decisions that affect brain chemistry. The justification is administrative burden relief, but you can taste the financial incentive underneath.

💰 MONEY MOVES Meanwhile, OpenAI just bought TBPN, a buzzy online talk show, which CNN frames as the AI giant buying influence rather than just content—a pattern dating back to 1926 when RCA created NBC partly to sell radios. Around the same time, Anthropic (the AI safety company) launched "AnthroPAC," a corporate political action committee to ramp up election spending. Both moves signal that major AI firms are playing political infrastructure, not just building models. On the investment side, income investors suddenly have options: the FT Vest Technology Target Income ETF (TDVI) offers an 8% yield by bundling dividend-paying tech stocks with covered calls, while Archer Aviation—the electric air taxi company—has crashed 62% from its 2025 highs and remains a pre-revenue money-burner with a $729 million operating loss last year. The consensus is clear: walk away from aviation hype, embrace stable tech dividends.

Here's what nobody's talking about enough: algorithms are already deciding your life in ways you can't see. Social media feeds curate what you see, recommendation engines shape what you buy, dating apps narrow your romantic options before you choose, navigation apps route you automatically, and streaming platforms guide your next binge—all powered by systems learning from your behavior and anticipating your decisions. It's not malicious, just inevitable. The convenience is undeniable. The loss of control is real.

Finally,

One other signal: Chinese tech firms are openly marketing detailed intelligence on U.S. military movements in Iran to buyers. That's a different category of problem entirely, but it's worth noting that while American companies are debating AI ethics and buying podcasts for influence, other players are moving faster and asking fewer questions.

April 03, 2026 at 07:00 AM

# Tech News Briefing — April 3, 2026

Four pioneers—Geoffrey Hinton, Yann LeCun, Yoshua Bengio, and Fei-Fei Li—are being celebrated this week as the architects who refused to quit on artificial intelligence when the world had already written it off. During the so-called "AI winters," when funding evaporated and the field seemed dead, these researchers kept publishing papers almost nobody read, training students in supposedly obsolete techniques, and asking questions their peers found either too abstract or too ambitious to bother with. 🚀 THIS IS COOL In 2024, Hinton received the Nobel Prize in Physics for his foundational work on neural networks and backpropagation—a 1986 paper that became one of the most cited works in computer science history despite attracting almost no attention when it was first published. The irony is sharp: the technology that now powers search engines, medical diagnostics, and image recognition systems was built on decades of quiet, unglamorous faith in ideas that seemed ridiculous at the time.

💰 MONEY MOVES Meanwhile, the AI boom is reshaping capital flows in ways that should make you pay attention. OpenAI just acquired TBPN, an online tech talk show with a loyal Silicon Valley following, marking an aggressive move in its escalating competition with Anthropic for enterprise customers. That acquisition signals something larger: control over the narrative around AI matters as much as the technology itself. At the same time, defense startups are eyeing what they're openly calling an "Iran war windfall" as the U.S. and Gulf states turn to technology for military modernization, while Chinese chip firms are posting record revenue driven by surging domestic AI demand—bolstered, conveniently, by U.S. trade restrictions that force local manufacturers to become more competitive. The geopolitics of AI is no longer theoretical.

Then there's the uncomfortable stuff.

🤔 THINK ABOUT IT On the medical technology side, something genuinely transformative just happened quietly. Doctors at Johns Hopkins University used digital "twins"—virtual replicas of patients' diseased hearts—to safely test treatments before operating on the real thing. 🚀 THIS IS COOL The FDA allowed this digital twin technology to guide treatment for ten patients with ventricular tachycardia, a notoriously difficult-to-treat arrhythmia responsible for roughly 300,000 U.S. deaths annually, and the results were promising enough to publish in the New England Journal of Medicine this week. The technology lets doctors see exactly where to ablate damaged tissue by watching electrical waves swirl across a simulated heart—basically treating the digital version first to see what actually works. This represents a fundamental shift in how medicine could work: predict, simulate, validate, then execute. If this scales, it changes everything about how we approach complex surgical interventions.

The darker tech stories are brewing too.

Finally, there's the White House news app delivering what the Washington Post calls a "rose-tinted view" where Trump is always winning, Chinese IVF clinics are embroiled in embryo mix-up scandals, and Guyana's Small Business Bureau just launched a $1.2 million green technology fund per applicant. The tech world is fragmenting: some innovations (digital heart twins) are genuinely saving lives, while others (defense startups eyeing profit from regional conflict) are accelerating old patterns of warfare, and still others (government information apps) are becoming propaganda tools. 🤔 THINK ABOUT IT When artificial intelligence was abandoned for decades, the heroes were the ones who kept working anyway—but now that AI is reshaping every industry from healthcare to defense to media, who decides whether we're using these breakthroughs to solve problems or exploit them?

April 02, 2026 at 05:01 PM

OpenAI is making a surprise pivot into media. The AI giant acquired TBPN, a tech talk show founded in late 2024 by entrepreneurs John Coogan and Jordi Hays, which has built a devoted following in Silicon Valley through CEO interviews and industry coverage. 💰 MONEY MOVES OpenAI didn't disclose financial terms, but the acquisition represents a striking strategic shift for a company that just shelved its Sora video-generation tool to focus on the lucrative enterprise AI coding market. The move signals OpenAI's intention to shape the narrative around AI's societal impact while competing fiercely with Anthropic for institutional customers and mindshare. Sam Altman himself has appeared on the show, alongside Meta's Mark Zuckerberg, Microsoft's Satya Nadella, and filmmaker James Cameron—suggesting TBPN already had credibility with tech's power players.

Meanwhile, the broader AI hype machine itself is becoming a problem. Federal Reserve economists Miguel Faria-e-Castro and Serdar Ozkan published analysis suggesting that AI optimism is driving inflation in the short term, regardless of whether the technology actually delivers on its promises. The mechanism is straightforward: when households see AI adoption as a "news shock," they interpret it as a future income boost and spend more today. Businesses do the same, increasing investment in AI infrastructure and tools based on the assumption of massive productivity gains. The result? 💰 MONEY MOVES Demand outpaces supply, prices rise, and we get what the Fed researchers call an "inflationary surge in aggregate demand"—even if the productivity miracle never materializes. They compare the current moment to dotcom optimism at the turn of the century, when computers were everywhere except, ironically, in the productivity numbers. The irony cuts deep: we might be overheating the economy based on a technology bet that could take years or decades to pay off, if it ever does.

🚀 THIS IS COOL On the infrastructure side, Marvell Technology surged 7 percent on news that NVIDIA is making a $2 billion strategic investment in the company, weaving Marvell into the heart of NVIDIA's AI factory ecosystem through the NVLink Fusion partnership. Marvell's data center revenue already hit $6.1 billion in fiscal 2026, up 46.5 percent year-over-year and representing 74 percent of total company revenues—proof that the chipmaker has already transformed itself into an AI infrastructure specialist. The deal also follows Marvell's $3.25 billion acquisition of Celestial AI in February, adding photonic fabric technology that complements the silicon photonics work with NVIDIA. This is a vertically integrated competitive moat being built in real time, with NVIDIA's $2 billion validation signal telling every hyperscaler globally that Marvell belongs in their next-generation data centers.

But not every tech stock is riding the AI wave. 💰 MONEY MOVES Align Technology was removed from the FTSE All-World Index in March, which could trigger mechanical selling from index-tracking funds and reshape how institutional investors view the company's role in global portfolios. Align projects modest low-single-digit revenue growth through 2028, with Q1 2026 guidance of $1.01 to $1.03 billion—a far cry from the explosive growth expectations dominating Silicon Valley. The company faces sustained pricing pressure and weaker demand for elective treatments like Invisalign, and the index removal creates short-term trading pressure precisely when the company needs to prove it can execute. Meanwhile, a Wall Street giant issued a doomsday warning for Micron Technology, another darling of the AI chip boom that has tumbled from its pedestal since Trump's reelection fueled hopes for semiconductor dominance.

🤔 THINK ABOUT IT If Fed economists are right that AI hype alone can drive inflation independent of actual productivity gains, what happens when the gap between promise and reality becomes undeniable? Does the bubble deflate slowly or suddenly? And when it does, which companies—those betting on real infrastructure like Marvell, or those dependent on narrative momentum like the media outlets being acquired to shape that narrative—will actually survive intact?

April 02, 2026 at 07:00 AM

Apple just became a half-century-old tech giant, and the story of how a college dropout and a Hewlett-Packard engineer accidentally changed the world on April Fools' Day 1976 is still wilder than most startup mythologies. Steve Jobs and Steve Wozniak founded Apple Computer Co. with a handshake and a two-page partnership agreement that valued the company so modestly that their original 10 percent stakeholder, Ron Wayne, sold his share for $2,300—a decision that would have netted him roughly $370 billion at today's $3.7 trillion market valuation. 🚀 THIS IS COOL The Macintosh's introduction in January 1984 didn't just change computing; it changed how corporations sell products, turning a 60-second Ridley Scott commercial into the template for Super Bowl advertising that still dominates culture four decades later. Apple's journey from a garage startup to the world's most valuable company involved a brutal public firing of Jobs in 1985, a desperate return in 1997, and then an innovation sprint that somehow produced the iPod, iPhone, and iPad within a single decade. That's the preamble to what's happening right now in the chip and AI infrastructure wars.

💰 MONEY MOVES Nvidia just wrote a $2 billion check to Marvell Technology, and this isn't charity—it's strategic ecosystem-building that sent Marvell's stock up 7 percent in a single trading session. The deal integrates Marvell's custom AI chips into Nvidia's NVLink infrastructure, which means customers get cheaper, more specialized hardware that still plays nicely with Nvidia's expensive processors. Marvell's data center revenue hit $6.1 billion last fiscal year, up 46.5 percent year-over-year, and the company projects hitting $15 billion in revenue by 2028. For context, this partnership essentially cements Marvell as a legitimate player in the AI infrastructure arms race, not just a secondary chipmaker hoping for scraps. Meanwhile, Micron Technology bounced back 9 percent after getting hammered by fears over Google's TurboQuant memory-compression algorithm, which theoretically could reduce AI memory needs by up to 83 percent. 🤔 THINK ABOUT IT But here's the thing: analysts and Micron itself are pushing back, arguing that TurboQuant mainly improves inference efficiency, which could actually *accelerate* demand by letting AI run on edge devices that previously couldn't handle it. Micron's high-bandwidth memory is sold out through all of 2026, and as the only U.S.-based HBM manufacturer, the company has structural advantages that no algorithm easily displaces.

The market's starting to realize that the "Great Rotation" away from AI winners might actually be creating buying opportunities elsewhere. Nintendo's Switch 2 is already the fastest-selling console ever, and with memory chip prices declining after OpenAI scaled back its spending commitments, the company's previous bear case—rising input costs—has evaporated. 💰 MONEY MOVES Nintendo's pulled in 99 percent revenue growth compared to last year through nine months of fiscal 2026, selling over 17 million hardware units, which sets the company up for a massive expansion beyond gaming into movies and theme parks. JPMorgan added three tech stocks to its favorite list in April: JFrog and Palo Alto Networks as AI adoption beneficiaries, plus Aramark, a facilities provider that somehow weathered the volatility better than the trendy names. The underlying pattern here is that investors are getting savvier about separating genuine AI infrastructure plays from the hype cycle—and 🚀 THIS IS COOL the technology underneath is legitimately impressive, from custom chipsets to compression algorithms that could democratize AI access.

But there's a darker side to this tech boom that nobody's talking about at investor conferences. 💰 MONEY MOVES Iranian forces attacked Amazon's AWS data center in Bahrain on April 2, marking the second assault on Amazon's Middle Eastern infrastructure, with the Revolutionary Guard having formally designated 18 American tech companies—including Apple, Microsoft, Nvidia, Google, Meta, Tesla, and others—as legitimate military targets on March 31. Iran claims these data centers support U.S. military and intelligence operations, and the attacks have triggered widespread outages across the UAE and surrounding territories, affecting financial institutions and government organizations. This isn't just a geopolitical sideshow; it's a direct threat to the infrastructure layer that powers every AI model, every cloud service, every digital transaction. The tech industry's entire growth story depends on uninterrupted access to data centers, and for the first time in decades, that assumption is being tested by kinetic warfare.

Franklin Templeton is doubling down on crypto by acquiring 250 Digital, a spinoff from venture firm CoinFund, signaling that traditional asset managers are now treating digital assets as essential rather than experimental. Meanwhile, the AgeTech industry is quietly booming, with robot roommates designed to keep seniors healthy and independent—a market driven by aging demographics that no algorithm will solve, only augment. 🤔 THINK ABOUT IT We're watching three parallel stories unfold simultaneously: the consolidation of AI infrastructure around a few dominant players like Nvidia and Marvell, the geographic and geopolitical fragmentation of that infrastructure due to Middle East tensions, and the expansion of tech into literally every corner of human life, from eldercare to entertainment. The question isn't whether technology will keep transforming the world—Apple's 50-year track record proves it will. The question is whether the tech industry can keep the lights on when nation-states start treating data centers as military targets.

April 01, 2026 at 05:01 PM

Nvidia is dropping $2 billion into Marvell Technology, and this isn't just a friendly investment—it's a strategic play that reveals where the AI chip market is actually headed. The two companies announced they'll work together to make Marvell's custom AI chips compatible with Nvidia's data center infrastructure, effectively expanding Nvidia's ecosystem while giving customers real optionality. 💰 MONEY MOVES For Marvell, which reported $8.2 billion in full-year revenue and is projecting $15 billion by fiscal 2028, this partnership essentially guarantees that its products won't be orphaned in a rapidly consolidating market. The custom chip movement—where companies build specialized silicon instead of buying off-the-shelf solutions—is accelerating, and Nvidia just signaled it's not threatened by that trend; it's going to own it.

But here's where things get weird: while Wall Street is celebrating the AI boom, the Federal Reserve is quietly warning that all this hype might be actively hurting the economy right now. Fed economists from the St. Louis branch published analysis suggesting that AI optimism is functioning like a "news shock"—when households and businesses hear about transformative technology on the horizon, they start spending and investing today in anticipation of future gains, which drives up demand and inflation before the technology actually delivers anything. 🤔 THINK ABOUT IT The dotcom bubble worked the same way—"Computers are everywhere except for in the productivity number," as one Fed researcher put it. We're potentially repeating that pattern, except this time with AI. Consumer price index is up 2.4% year-over-year, and while nobody can quantify exactly how much of that is AI hype versus other factors, the Fed's warning is a shot across the bow: the technology might be overheating an economy that doesn't have room for extra heat right now.

The human reality check is rolling in from multiple angles. Microsoft is hitting rough waters—Melius Research's Ben Reitzes bluntly stated there's no room in the budget to pay extra for Copilot, the company's flagship AI product, which suggests customers aren't convinced the productivity gains justify the premium. 🚀 THIS IS COOL Meanwhile, a massive study published in JAMA involving Mass General Brigham found that AI scribes in healthcare do work, but not as dramatically as promised: they freed up doctors to see one additional patient every two weeks, saving an average of 16 minutes daily. Primary care physicians benefited most with 27 minutes saved, but they still spent more than two and a half hours documenting visits daily even with the technology. That's meaningful but hardly revolutionary, and it definitely doesn't match the hype about AI liberating doctors from administrative burden.

Consumer-facing AI is flopping even harder. Customers are reportedly furious with AI-powered customer service chatbots, particularly when seeking refunds—turns out generative AI designed to please humans is actually terrible at handling angry people looking for their money back. That's a massive problem for companies trying to deploy AI to cut costs, because if the technology creates customer friction instead of smoothing it, the ROI evaporates. 💰 MONEY MOVES Franklin Templeton just acquired 250 Digital, a cryptocurrency unit spun out of CoinFund, expanding its digital assets push at the exact moment when regulators are getting twitchy about unproven tech infrastructure. The crypto industry is following the same hype cycle as AI, which should worry everyone paying attention to historical patterns.

On the hardware side, things are moving faster. 🚀 THIS IS COOL Donut Lab's solid-state battery for the Verge TS Pro electric motorcycle just passed real-world charging tests, hitting 100+ kW of power for five minutes using air cooling instead of liquid cooling—the system charged from 10% to 50% in five minutes, putting it among the fastest-charging vehicles ever made. If this scales to cars, it's genuinely transformative. Micron stock soared 9% on Wednesday as investors reassessed whether Google's new TurboQuant memory-compression algorithm (which can reduce AI model memory needs by up to 83%) would actually hurt memory chip demand. Analysts are pushing back on the panic, noting that Micron's high-bandwidth memory is sold out for all of 2026 and that compression might actually accelerate demand by enabling AI on edge devices. Micron is the only U.S.-based HBM manufacturer, giving it a structural advantage that's hard to algorithm away.

The Bank of England added another warning this week: rapid AI adoption by financial institutions could become a stability threat, potentially triggering shocks in private credit markets that ripple everywhere. 🤔 THINK ABOUT IT We're simultaneously watching the most hyped technology in decades struggle to deliver on its promises in real-world deployments while hardware infrastructure accelerates and the Fed worries the hype itself is the real economic threat—not the technology's failure, but its success at convincing people to spend money today on speculative future gains. That's the tension the market needs to navigate, and right now nobody's pricing in what happens if the productivity gains never materialize at the scale everyone's betting on.

April 01, 2026 at 07:00 AM

TWSC showcased its full-stack AI storage solutions at the MemoryS 2026 Summit in Shenzhen last week, signaling a major shift in how companies are approaching the infrastructure behind artificial intelligence. The Taiwan-based storage firm presented a comprehensive ecosystem spanning data centers and edge-side AI applications, built on self-developed controllers, proprietary firmware, and scenario-specific optimization. 🚀 THIS IS COOL TWSC's breakthrough with its first self-developed enterprise SSD controller—paired with DDR5 RDIMM modules hitting speeds of 6400MT/s—represents the kind of hardware-firmware-algorithm synergy that AI workloads demand as they strain existing infrastructure with unprecedented concurrency and latency requirements. This matters because AI isn't just about the chips running models anymore; it's about the entire plumbing underneath.

Speaking of that plumbing, Nvidia just made a strategic bet that underscores exactly how competitive this space is getting. 💰 MONEY MOVES On Tuesday, Nvidia announced a $2 billion stake in chipmaker Marvell Technology and a partnership to integrate Marvell's custom AI chips and networking products into Nvidia's ecosystem using its NVLink Fusion technology. The move sent Nvidia shares up 5.6%, marking their second-best day of the year—a reaction that matters more than it might seem on the surface. Nvidia's largest customers, including Amazon and Microsoft, are increasingly designing their own custom silicon. Rather than fight this trend, Nvidia is essentially saying: build your chips, but use our networking fabric to talk to everything else. It's a brilliant pivot from defending the castle to becoming the castle's architecture. For investors who've watched Nvidia's stock languish despite steady good news, this partnership suggests the company isn't just a GPU vendor anymore—it's becoming infrastructure.

Meanwhile, OpenAI is capitalizing on the insatiable demand for AI compute and development. 💰 MONEY MOVES The company added another $12 billion to its latest funding round, bringing total capital raised to eye-watering levels. What's notable here isn't just the number but the mechanism: OpenAI extended participation to individual investors through banks for the first time, raising over $3 billion from retail investors who previously couldn't get in. This democratization of AI funding is significant. It suggests that investors—whether institutional or individual—are fully convinced we're in a foundational technology shift. The question becomes whether OpenAI can translate all that capital into revenue growth that justifies the valuations, or whether we're watching another cycle of world-changing technology paired with cash that's slightly too easy to spend.

On the academic side, IBM and ETH Zurich announced a 10-year collaboration focused on next-generation AI quantum algorithms targeting both scientific and enterprise use cases. 🚀 THIS IS COOL The partnership includes new professorships and dedicated research projects, signaling IBM's intention to shape the future talent pipeline and computing architectures simultaneously. For IBM investors watching the stock at $242.39 with a three-year return of 104%, this deal is less about immediate payoff and more about optionality—betting that the convergence of quantum computing and AI will create new categories of enterprise problems worth billions. It's a long-term play in a world obsessed with quarterly earnings, which means IBM is either brilliantly patient or quietly burning runway on research that competitors might commercialize first.

Back on the domestic front, the Trump administration reorganized health IT regulation by reverting the federal health IT office back to its Office of the National Coordinator name and streamlining its focus to external coordination rather than internal HHS technology oversight. It's a technical bureaucratic move with practical consequences: the chief technology officer, chief data officer, and chief AI officer roles moved back under the chief information officer. This matters for health data interoperability, which has been a genuine problem for patients trying to access their own medical records—one of the few areas where government and business agree something's broken and needs fixing.

🤔 THINK ABOUT IT If custom AI chips are becoming table stakes and every major tech company is building them, what does that do to Nvidia's 80% market share in GPU compute, and how long does a $2 billion partnership actually buy you when the next generation of architectures might render today's integrations obsolete? The infrastructure layer is where real power concentrates, but right now everyone's racing to redefine what that layer actually is.

March 31, 2026 at 05:01 PM

OpenAI just added another $12 billion to its funding round, bringing the total haul to a staggering $122 billion. That's not a typo. The scale of capital flowing into AI infrastructure right now is almost incomprehensible—it dwarfs what most Fortune 500 companies raise in their entire existence. But the real story isn't just the number itself; it's what it signals about who controls the future of artificial intelligence and how much faith investors still have despite a volatile quarter for tech stocks.

Speaking of AI dominance, 💰 MONEY MOVES Nvidia just dropped a $2 billion strategic investment in chipmaker Marvell Technology on Tuesday, and the market loved it. Nvidia stock jumped 5.6%, while Marvell surged 7%, signaling that investors see this as a genuine inflection point rather than defensive positioning. The partnership integrates Marvell's custom AI chips and networking products into Nvidia's NVLink ecosystem, effectively letting Nvidia's largest customers—Amazon, Microsoft, Google, Meta—build custom silicon that still plays nice with Nvidia's dominant computing fabric. It's a brilliant play: Nvidia acknowledges that custom chips are the existential threat to its GPU throne, then figures out how to profit from them anyway. CEO Jensen Huang framed it as inevitable—all the world's data centers are being replaced with this new form of computing, and Nvidia wants to own the connective tissue.

🚀 THIS IS COOL What makes the Marvell deal architecturally interesting is the silicon photonics angle. Marvell acquired Celestial AI's photonic fabric technology, and now that capability sits directly inside Nvidia's ecosystem for the first time. That's infrastructure-layer differentiation that goes way beyond the GPU war. Marvell itself is no startup betting the farm on AI—the company just reported 42% fiscal 2026 revenue growth with data center revenue hitting $1.518 billion and representing 73% of total revenue. Nvidia is doubling down on real momentum, not speculation.

The market's overall mood on Tuesday, though, had less to do with AI chipsets and more to do with geopolitics. 💰 MONEY MOVES Wall Street surged hard—the Dow jumped over 1,100 points, the S&P 500 rose 2.9%, and the Nasdaq climbed 3.8%—all because traders got hopeful that the Iran war might be ending. Iran's president signaled willingness to negotiate, Trump said the hard part is done, and suddenly oil futures tanked. Brent crude fell 2.8% to around $104 per barrel, which matters because the global inflation fears that have haunted markets for weeks started looking slightly less scary. This was the upbeat end to a brutal quarter—both the S&P 500 and Dow had posted their worst quarter since 2022 before Tuesday's rally. 🤔 THINK ABOUT IT How much of tech's gains this quarter actually came from AI fundamentals versus just relief that a regional conflict might cool down?

Meanwhile, Jack Dorsey is pitching a vision of AI-powered organizations that replace middle managers entirely, weeks after his company Block cut nearly half its staff. And in Europe, a French consumer group just sued Ubisoft over the shutdown of online racing game "The Crew," raising questions about what happens to digital products when companies flip the kill switch. NASA also lined up multiple launch windows in April for Artemis II, and the Trump administration renamed the federal health IT office back to the Office of the National Coordinator, reversing Biden-era organizational changes that had bundled in chief technology, data, and AI officer roles.

The through-line here is consolidation—of capital into mega-funded AI companies, of computing power into ecosystems controlled by a few giants, of decision-making away from human managers. The money is real, the technology is moving fast, and the winners are already starting to look inevitable.