Original explainers & special topics from Nakitte
In mid-2026, two of the world’s largest technology companies released sustainability reports that function less as public relations documents and more as unvarnished data releases on the physical realities of the artificial intelligence buildout.
For nearly a decade, the dominant paradigm in frontier AI has been simple in principle and extraordinarily expensive in practice: predict the next token. Trillions of dollars in compute, energy, and engineering have been poured into scaling this approach.
The International Monetary Fund has delivered one of its clearest warnings yet on the structural consequences of tokenization. In a July 2 blog post drawing on deeper research published in April, IMF Financial Counsellor Tobias Adrian argues that moving financial assets and liabilities onto shared digital ledgers is not merely a technological upgrade.
While the Magnificent 7 produced uneven results and faced growing questions about the sustainability of their dominance, memory chipmakers posted returns that recalled the semiconductor supercycle of the late 1990s. Semiconductors and related hardware decisively won the quarter.
For years, the dominant narrative around artificial intelligence has been one of looming displacement. Headlines have warned of mass job losses, with estimates ranging from Goldman Sachs’ projection of up to 300 million full-time equivalent roles exposed globally to the IMF’s assessment that roughly 40% of jobs worldwide — and 60% in advanced economies — are exposed to AI.
The global semiconductor industry has entered a new regime. What began as a cyclical upswing in memory prices has evolved into a structural supply constraint centered on high-bandwidth memory (HBM), with spillover effects that are now hitting conventional DRAM, NAND flash, and downstream industries from PCs to automobiles.
The German automotive industry, long a pillar of Europe’s industrial might and a symbol of engineering excellence, is navigating one of its most severe crises in decades. In 2025 and into 2026, the sector has recorded sharp profit declines, accelerated job losses, and a dramatic erosion of its dominant position in key export markets
AI and high-performance computing workloads have pushed server and rack heat densities far beyond traditional air-cooling limits. Modern GPU-heavy racks now exceed 80–100 kW, with individual servers generating heat loads of 6–10 kW or more—up to 50× the thermal output of legacy CPUs in some accelerator configurations.
In the global contest for artificial intelligence leadership, attention fixates on GPUs, foundation models, and the hyperscalers pouring billions into training clusters. The more consequential constraint, however, is measured in megawatts, not teraflops.
In an era defined by supply-chain fragility, labor shortages, decarbonization mandates, and intensifying global competition, artificial intelligence has transitioned from experimental pilots to a core strategic lever in manufacturing.
Germany’s Energiewende (energy transition) stands as one of the most ambitious national decarbonization efforts in history. Launched with deep roots in the environmental movements of the 1970s and accelerated through landmark legislation in the 2000s and 2010s, it has delivered remarkable growth in renewable electricity while driving substantial greenhouse gas reductions.
The artificial intelligence boom has turned electricity into the new strategic bottleneck for Big Tech. While GPUs, models, and algorithms dominate headlines, the physical constraint of reliable, high-density, carbon-free power is reshaping capital allocation, site selection, and long-term competitive positioning.
Nearly nine decades after the *Hindenburg* disaster ended the commercial heyday of rigid airships, a quiet but determined revival is underway.
When Insilico Medicine announced in November 2024 that its generative AI-designed TNIK inhibitor, Rentosertib (formerly ISM001-055), produced dose-dependent improvements in forced vital capacity for idiopathic pulmonary fibrosis patients in a Phase IIa trial—with a +98.4 mL mean gain at the highest dose versus a -62.3 mL decline on placebo—the result carried weight far beyond one disease area.
Small Modular Reactors (SMRs) promise lower per-project capital outlays ($0.5–2+ billion range for typical deployments versus $6–12+ billion for large reactors), shorter construction timelines (target 2–4 years), and modular scalability.
Emerging markets face a stark energy trilemma: surging electricity demand from population growth, industrialization, and digitalization; heavy reliance on imported fossils or aging coal fleets; and the need for affordable, reliable, low-carbon power to support GDP expansion without locking in emissions or balance-of-payments shocks.
On 5 June 2026, President Recep Tayyip Erdoğan announced at the 3rd Global Islamic Economy Summit in Istanbul that three state-owned participation banks — Ziraat Katılım, Vakıf Katılım and Halk Katılım — would merge.
Washington — The U.S. labor market posted a stronger-than-expected performance in May, adding 172,000 nonfarm payroll jobs — nearly double the consensus forecast of around 85,000 — the Bureau of Labor Statistics reported on June 5.
The first quarter of 2026 delivered a stark illustration of how geopolitical shocks interact with structural demand in Middle East real estate. As tensions escalated into direct conflict involving Iran, Israel, and the United States, short-term sentiment indicators flashed warning signs — including sharp but temporary drops in certain Dubai real estate indices in March.
Broadcom Inc. (AVGO) reported another quarter of exceptional top-line expansion driven by its expanding role in artificial intelligence infrastructure, only for investors to punish the stock for what they viewed as insufficiently aggressive forward guidance.
In the first quarter of 2026, Alphabet, Amazon, Meta, and Microsoft alone deployed $130.65 billion in capital expenditures — more than three times the cost of the Manhattan Project in inflation-adjusted terms and 71% above the prior-year quarter.
As the company confidentially files for a potential 2026 IPO, its explosive growth from $1 billion to $47 billion in annualized run-rate revenue in roughly 17 months forces a hard look at whether private-market exuberance can survive public scrutiny.
Space Exploration Technologies Corp. is preparing what would be the largest initial public offering in modern market history.
