Viral coding agents have turned software delivery from months into weeks and weeks into days. Major firms now sell AI solutions teams that help every other company board the same express — and the logical endpoint is software that no longer arrives as SaaS, but as an in-house capability that no longer needs the SaaS industry's engineers.
Latest from AI & Compute
Defining analyses in AI & Compute
Drone strikes physically damaged AWS data centers in the UAE and Bahrain, knocking 109 cloud services offline and cascading into banking, payments, and logistics across the Gulf. Simultaneously, seven P&I clubs cancelled war risk coverage for the Strait of Hormuz, freezing 20% of global oil transit without a single mine being laid.
Nvidia dominates the AI narrative, but a structural shift is underway beneath it. Hyperscalers are designing custom chips — and Broadcom is the architect translating those designs into silicon. With AI revenue doubling year-over-year, six major XPU customers, and a line of sight to $100 billion in chip revenue by 2027, Broadcom is becoming the indispensable backbone of AI infrastructure.
Nvidia posted the largest clean beat in semiconductor history — $68.1 billion in Q4 revenue, 73% year-over-year growth, and guidance that crushed every estimate. The stock fell 5.5%, its worst day in ten months. The paradox reveals a market that no longer rewards AI momentum; it demands proof that the $700 billion capex deluge will ever pay off.
After a year celebrating massive AI infrastructure spending, Wall Street abruptly shifted in early February 2026. Alphabet's announcement of $185 billion in annual capex—more than double 2025—forced investors to confront whether the industry's $660 billion buildout represents rational investment or developing bubble.
10 additional stories
While U.S. hyperscalers committed $660 billion to a GPU arms race, Chinese researchers developed algorithmic innovations achieving similar results at a fraction of the cost. DeepSeek's success represents more than competitive threat—it's a fundamental challenge to assumptions underpinning the West's AI development strategy, with implications for capex sustainability, export control effectiveness, and trillion-dollar market valuations.
Truist downgraded Northrop Grumman, citing margin risk at a moment when defense contractors face cost inflation, program schedule pressure and increasing prime–subcontractor friction. The move could presage a modest sector-wide repricing if profit deterioration proves persistent rather than idiosyncratic.
ASML stands to gain if DRAM makers adopt EUV at scale and TSMC resumes a fresh ordering cadence: the company’s extreme‑ultraviolet tools sit at the intersection of memory cyclical recovery and foundry investment, creating a two‑way upside via supplier cap‑ex cycles that ripple through the semiconductor supply chain.
The CME’s second successive margin increase on silver and gold futures this month has shrunk permitted leverage, lowered speculative volume and pushed dealer desks to reprice risk. Traders are stripping back positions; liquidity is fragmenting across venues; and hedging costs are rising for miners and jewelers.
Meta's purchase of Manus—an AI agent studio—bundles code, workflows, and people into a package that accelerates agent capabilities across social products. The deal shifts the binding constraint from model scale to platform integration and talent, creating both a moat and a flashpoint for rivals and regulators.
Large language models have broken through the barrier between computation and creativity, discovering new theorems and algorithms that expand human mathematical knowledge. From FunSearch to AlphaEvolve, AI systems are now partners in mathematical exploration.
The tech industry's AGI obsession represents a catastrophic misallocation of resources, chasing artificial minds while wasting billions of real ones. An examination of why the most advanced intelligence we know isn't coming from data centers—it's already here.
NVIDIA's CUDA platform has dominated GPU computing for nearly two decades. But that dominance may be the company's greatest vulnerability as AI-powered compilers threaten to make accumulated expertise irrelevant.
