They think they are pigs being fed. They are hogs going to slaughter.
By Scott Ortkiese | July 21, 2026 | Email: so@throughlinesynthesis.com
I wrote a longer piece last week, butressing what you already know, calledBrute Force Is Not Progress, arguing that generative AI in its current architecture is a shockingly inefficient use of memory, storage, electricity, water, land, and capital, and that the trillion dollar capex cycle behind it is being subsidized by ordinary Western consumers, ratepayers, and pension holders. That essay was written for a Western audience, in the language of American engineering criticism and American capital markets.
This companion is written for Russian and broader Eurasian readers. The engineering critique is the same. The consequences are not. A capex cycle this large, financed this aggressively, denominated in dollars, and dependent on chokepoint supply chains that Washington controls, does not just reshape American electricity bills.It reshapes the strategic terrain that Russia, China, India, Iran, and Turkey are already using to build a multipolar technology order.The question I want to work through here is what that terrain looks like from Moscow, and what Russia can and cannot honestly claim as a competitive position.
I write as a Houston based analyst with no illusions about Washington, and no interest in flattering any capital. The value of writing for a Russian audience is precisely that the analytical standard is higher than the propaganda standard. Russian readers know when they are being sold a story. So I will name the Russian weaknesses in the same voice I use to name the American ones.
The Western trillion dollar bill, in one paragraph
Combined 2026 capital expenditure by the top five American hyperscalers, Amazon, Alphabet, Microsoft, Meta, and Oracle, is now projected at roughly 725 to 785 billion dollars, up 77 percent year over year, with Barclays modeling 1.1 to 1.2 trillion dollars by 2028. Goldman Sachs projects 7.6 trillion dollars of cumulative AI related capex through 2031. JPMorgan projects 5.5 trillion dollars through 2030. The five largest hyperscalers are already outspending their own operating cash flow and issuing debt to cover the gap, a fact the Bank for International Settlements flagged publicly in June 2026. Roughly 70 percent of the world’s high end computer memory is being absorbed by these buildouts, DRAM and NAND prices are running 60 percent higher quarter over quarter, hard drive prices have more than doubled, laptop prices in some segments are up 50 percent, and Virginia’s data center corridor alone now consumes roughly 40 percent of that state’s electricity. Behind those numbers stands an architecture that Yann LeCun, one of the field’s founding figures, publicly says is not a path to superintelligence or even to human level intelligence.
That is the Western bill. Now the Russian angle.
The direct hit on Russia is small, and that is the first strategic fact
Sanctions and the 2022 payment system severance had already broken most direct Russian exposure to the Western AI capex cycle before it took off. Russian households and firms do not buy Nvidia GPUs at scale, do not pay OpenAI or Anthropic subscription fees at scale, and are not creditors of the hyperscaler bond issuance that JPMorgan expects to exceed 2.1 trillion dollars over five years. The channel that hits Western consumers hardest, the memory pass through into DRAM, NAND, and finished laptops, reaches Russia only through parallel imports at a markup, which was already the baseline after 2022.
That relative insulation is the first strategic fact worth naming. When a bubble this large deflates, or even when it merely corrects sharply, the balance sheet damage lands on Western hyperscalers, Western creditors, Western pension funds, Western ratepayers, and by extension on the political stability of Western coalitions. It does not land, in any equivalent way, on the Russian federal balance sheet or on the Russian household. That is not a Russian strategic achievement, it is a Russian strategic condition, produced by the same sanctions architecture that Washington sold as punishment.
Sanctions do not compound over the entire time horizon they are advertised to compound over. They frontload pain, and then they anchor the target economy outside the buyer of last resort. If the buyer of last resort then over invests in a single technology stack, the anchored economy carries less of the correction risk. That is the arithmetic Washington did not model when it designed the 2022 measures.
The Russian fiscal position is stronger than the Western one on debt, weaker than it looks on flow
Now the honest counterweight, because a Russian reader will not tolerate a Western analyst telling comfortable stories about Russian strength.
On the stock of debt, Russia is in a category almost no Western economy occupies. Russian federal debt closed 2025 at roughly 35.1 trillion rubles, about 16.5 percent of GDP, according to the Russian Finance Ministry, with the ceiling set at 20 percent through 2029. Trading Economics reports the December 2025 figure at 18.30 percent, and FRED at 17.85 percent for 2024. The comparable OECD average is roughly 110 percent of GDP. Russian public debt is overwhelmingly domestic and ruble denominated, with external federal debt at roughly 2.3 percent of GDP according to the French Treasury. International reserves reached 721.7 billion dollars by mid-2026, formally covering the entire stock of public debt several times over.
On the flow, however, the picture is thinner and worth stating plainly. Russia’s federal budget ran a 5.73 trillion ruble deficit in the first half of 2026, or 2.5 percent of GDP, already exceeding the full year plan of 1.6 percent set in the budget law. Reuters reported on July 16, 2026 that spending is now projected at 45.11 trillion rubles against unchanged revenue, pushing the 2026 deficit toward 4.83 trillion rubles versus the 3.79 trillion planned. The Center for Macroeconomic Analysis and Short-term Forecasting expects the full year gap to reach 3 percent of GDP, or 7 trillion rubles.
The National Wealth Fund’s liquid portion has been depleted from roughly 6.5 percent of GDP before February 2022 to 1.5 percent of GDP by July 1, 2026, or about 46.4 billion dollars, according to TASS citing the Finance Ministry. The Moscow Times, OSW, and independent analysts have documented that the budget rule was suspended for March through May 2026 and that the gap was financed by OFZ bond issuance largely absorbed by state banks using central bank refinancing. Finance Minister Anton Siluanov has publicly said the structural primary deficit will not close before 2029. Debt service, at 8.9 percent of federal spending in 2026 and roughly 1.5 percent of GDP, is now twice the pre-war share, and running at nearly double the OECD average burden despite a fraction of the OECD’s debt stock, because Russian yields are elevated by war risk and by the central bank’s tight policy stance.
None of that is an argument for panic. Russian debt is unlikely to trigger a 1998 style crisis at these levels, and every serious analyst including The Moscow Times and Pravda concedes that point. But the reader deserves the honest version of the story, which is that Russia enters this cycle with the lowest debt to GDP ratio of any major economy and with the least room to increase it, because the flow deficit is widening while sanctions have raised the marginal cost of every additional ruble of borrowing. Russia’s strategic advantage on debt is real. It is not infinite, and it is being spent, quietly, on the war and on import substitution simultaneously.
Compared against Western hyperscalers whose capex now consumes 85 to 90 percent of operating cash flow, and against Western sovereigns whose debt runs above 100 percent of GDP with debt service rising faster than growth, Russia’s balance sheet still looks better. Compared against Russia’s own three year plan, it looks tighter than at any point since 2022. Both statements are true.
Where the multipolar node actually sits
The interesting question is not whether Russia can outbuild American hyperscalers on GPU count. It cannot, and no serious Russian planner claims otherwise. The interesting question is where Russia sits inside a Eurasian and BRICS plus stack that does not need to match American capex to compete with it on cost, sovereignty, and specific verticals.
Four data points from the last month sketch the answer.
First, the BRICS Academic Conference in Moscow in July 2026 named connectivity, resilience of transport and trade corridors, critical supply chains, and AI cooperation as its explicit agenda. Russia’s BRICS Council reports that the national AI strategy targets an 11.2 trillion ruble contribution to Russian GDP by 2030, with a Kremlin task force coordinating domestic generative models and a stated policy that dependence on foreign LLMs is unacceptable.
Second, the BRICS Standards Bodies meeting in Bengaluru in mid-July 2026 opened a dedicated workshop on AI standards with experts from India, Russia, China, and Brazil. India used the meeting to push common standards and rules for safe, trustworthy, and responsible AI systems, and the political intention on the Indian side is a BRICS wide AI standards MoU that competes with the American, EU AI Act, and OECD frameworks.
Third, at the World AI Conference in Shanghai on July 17, 2026, Xi Jinping pitched China as the leader of a new global AI order, launching a new world body positioned to serve developing countries and to route around Western export controls. Reuters covered it plainly. The New York Times covered it as a threat, noting that China, Russia, and others are actively working to inflame debate over American AI data center policy in domestic Western media, which is another way of saying the multipolar side has figured out that ratepayer politics in Virginia, New Jersey, and Ohio is now a useful lever.
Fourth, in May 2026, India and Russia advanced a critical minerals cooperation framework covering exploration, processing, and technology transfer for lithium and rare earths, with a draft agreement circulated and signing expected within months. Indian Rare Earths Limited is reportedly evaluating the Tomtor deposit in Yakutia in cooperation with Rosneft. Tomtor is one of the world’s largest undeveloped rare earth deposits. The Kolmozerskoye lithium mine in Murmansk resumed active operations in early 2026 after a thirty year pause, with commercial scale targeted for 2030.
Add Turkey and Iran to the picture and the node is not hypothetical, it is operational. Turkey sits on the Middle Corridor and controls the Bosporus and Ankara’s independent AI, drone, and defense industrial policy. Iran sits on the North South Transport Corridor and on the Persian Gulf, and it now trains substantial technical capacity in Chinese and Russian institutional partnerships. The China Russia India Turkey Iran configuration your question named is the operational spine of the emerging Eurasian technology bloc. It does not need to match American hyperscaler capex to matter. It needs standards, common protocols, cross border payment infrastructure, and physical connectivity, all of which are being built.
What Russia can honestly claim on technology, and what it cannot
Technology sovereignty is where Russian rhetoric runs ahead of Russian industrial capacity, and the honest reader should be told which is which.
What Russia can honestly claim is real. Sberbank’s GigaChat and Yandex’s YandexGPT are functioning sovereign LLM stacks used in production by Russian government, banking, and consumer services. The 2025 Data Economy and Digital Transformation of the State national project funds domestic infrastructure, applied research, cybersecurity, and workforce development. Putin’s July 2026 order for a national AI task force focused on domestic data centers and dedicated power sources, including small nuclear plants, is a coherent policy, not a headline. Russia has world class mathematical and cryptographic talent, a legitimate quantum research program at Rosatom and Skoltech, and a well documented capability in specialized narrow AI for defense, energy, and industrial applications.
What Russia cannot honestly claim is that it can match American or Chinese frontier model performance on general purpose LLMs in the short term. It also cannot claim self sufficiency in the compute stack. The chip design and fabrication chokepoints run through Taiwan (TSMC), the Netherlands (ASML’s EUV monopoly), and secondarily South Korea, Japan, and the United States. Sanctions have limited Russian access to leading edge nodes below roughly 28 nanometers, and parallel imports at markup do not substitute for domestic fab capacity. The Moscow Times reported on June 24, 2026 that Russia has suspended 38 data center projects worth 168.6 billion rubles, or 2.26 billion dollars, over the past three years, because of high borrowing costs and grid limits. That is the truthful ceiling on the Russian domestic buildout under current conditions.
Rare earths are the same story on a longer horizon. Russia holds the world’s second or fourth largest declared reserves depending on classification standard, roughly 3.8 million tonnes on the USGS estimate and up to 28.5 million tonnes on the Russian Ministry of Natural Resources estimate. Russia produces roughly 2,600 tonnes per year, less than one percent of global mine output, against Chinese production of roughly 270,000 tonnes. Ksenia Shoigu, head of the Mendeleev Valley Foundation, disclosed at Innoprom in July 2026 that Russian dependence on imports of pure rare earth metals reaches 90 percent in certain segments and 75 percent overall, despite the reserve base. The foundation has built 12 full cycle technological chains and signed nine agreements with 13 technology partners, targeting lithium for energy storage, ultra pure silicon, and rare earth applications in AI hardware. Deputy Minister of Industry Mikhail Yurin was blunt at the same event: the market niche is protected not by the rare earth itself but by the finished high tech product that contains it.
Read those two admissions together. Russia is the world’s largest rare earth reserve holder that cannot yet close its own supply chain. That is a solvable problem on a five to seven year horizon, and the India Russia framework is precisely the vehicle for solving it, because India needs Russian reserves and Russia needs Indian and Chinese processing know how and capital. But the honest current state is a reserves story rather than a production story.
The strategic corollary is that Russia’s technology competition strategy cannot be built on matching American capex. It has to be built on three things. One, on the low debt, low external exposure position I described above, which gives Moscow room to run counter cyclical AI investment when Western capex corrects. Two, on the Eurasian standards, payments, and physical connectivity spine now being assembled through BRICS plus, which reduces the marginal cost of every Russian participant in the bloc. Three, on the resource endowment, which is real and which becomes a Russian strategic asset only if Moscow closes the processing gap that Shoigu and Yurin publicly named.
Energy is the specific place where Russia is best positioned
If there is one dimension of the AI capex cycle where Russia’s underlying position is genuinely strong, it is energy.
The American and Chinese hyperscaler buildouts are running into a physical wall, and the wall is called electricity. Global data center power demand is on track to exceed 1,000 terawatt hours in 2026 and reach 945 terawatt hours by 2030 on the IEA base case. In the United States, Goldman projects 66 gigawatts of data center load by 2027, up from 31 in 2025, against a 2,600 gigawatt permitting backlog for new generation. Some American operators are repurposing jet engines to bridge the gap. Ireland already consumes 21 percent of its metered electricity through data centers. Even China, which controls its own grid buildout and can commission new capacity faster, published Reuters reported guidance on June 25, 2026, encouraging green power links for data centers because of the emerging squeeze.
Russia sits on the largest natural gas reserves in the world, on the second largest coal reserves, on substantial hydro capacity in Siberia, and on a mature civilian nuclear industry through Rosatom, which is exporting reactors to twelve countries and is one of the few vendors on earth building SMR capable projects on commercial timetables. Russian electricity generation grew 1.8 percent year over year in Q1 2026. Siberian generating capacity is chronically under utilized because Russian population density does not consume it locally, which historically has been an economic weakness, but which under an AI compute economy becomes an asset. Cold climate, cheap power, and low population density are the ideal siting profile for hyperscale data centers, and Russia has all three at scale.
The natural counterparty is China, whose own grid is being drawn thin by domestic AI capex and whose western provinces sit directly across the border from Russian generation surplus. Vedomosti reported on July 16, 2026 that Russian energy exporters to China including Rosneft, Surgutneftegaz, Gazprom Neft, Gazprom, Novatek, and SUEK expect continued growth in supply volumes. Ground News flagged in late June that China and Russia are already in active talks after a temporary halt in power supplies, meaning the electricity export channel is live and negotiated. If the Eurasian node reaches full form, Russian generation and Russian cold climate siting could support Chinese, Indian, and Central Asian AI compute in a way that no other combination in the Eastern Hemisphere can.
That is a durable Russian asset, and it is not priced into the multipolar narrative yet. It should be.
Rare earths, palladium, and the mineral leverage worth being sober about
On the mineral side, Russia has real leverage in a few specific columns and imported dependency in most others. Palladium is the cleanest case. Russia produces roughly 75,000 kilograms per year, about 39 percent of world output, and the United States is 36 percent net import reliant on palladium, per the USGS 2025 commodity summary and the wafergraph 2026 supply chain analysis. That is real geopolitical weight in every catalyst and sensor market. On gallium and germanium, both critical for GaAs and GaN semiconductor substrates that increasingly matter for AI accelerator packaging, China holds roughly 99 percent of gallium and imposed export controls in August 2023. Russia is not a factor in those columns.
On rare earths broadly, the picture I described above is the operative one. The reserves are real. The processing gap is real. The India Russia framework is the most credible near term route to close the gap, because Indian financial capacity and processing ambition align with Russian reserves and refining infrastructure, and because Chinese dominance on the refining side pushes both parties toward a hedged partnership rather than dependency on Beijing. The Tomtor deposit in Yakutia, the Kolmozerskoye lithium deposit in Murmansk, and the Angara Yenisei hub together are the physical assets that will determine whether Russia becomes a genuine mineral node or remains a reserves story into the 2030s.
The realistic timetable is 2028 for a complete Russian rare earth technological line and 2030 for large scale Russian lithium output. That is not a criticism, it is a physical constraint that mining and refining projects impose on every jurisdiction. It is the timetable Russian planners quote themselves.
What a Western capex correction would do to Russia and to the Eurasian node
Now the counterfactual, because your original question deserves it. What happens to Russia if the Western AI capex cycle corrects sharply?
Four channels matter.
The first channel is commodity demand. AI compute is electricity, and electricity is oil, gas, and increasingly nuclear fuel. A Western capex correction would soften global electricity demand growth, which would soften global gas and coal demand at the margin. That is negative for Russian energy revenue in absolute terms, but positive relative to Western hydrocarbon producers whose cost base is higher. Russia’s Urals crude sensitivity is real, and Model Diplomat’s July 2026 analysis flagged the fiscal pressure at 51 dollar Urals. But relative competitiveness against North American shale and Middle Eastern producers with lower breakeven prices is a different question, and the multipolar routing of Russian energy toward China, India, and Turkey mostly bypasses the Western correction channel entirely.
The second channel is capital flow. A Western hyperscaler correction, whether in equity valuations or in the credit stack behind data center financing, would trigger risk off across US investment grade and high yield credit, then across emerging markets. Russia is largely disconnected from those flows since 2022, which insulates the ruble and the OFZ curve from the direct channel. It does not insulate the yuan, which matters because Russian reserves and Russian trade settlement now run heavily through the yuan. If a Western correction hits Chinese growth, it hits Russia via the yuan and via reduced Chinese demand for Russian energy and minerals. That is the second order transmission worth naming.
The third channel is technology cost. A Western correction would collapse the pricing moat that hyperscalers currently defend, and it would accelerate the adoption of DeepSeek and other Chinese and open source models at roughly one twentieth to one thirtieth of American frontier model costs on comparable regional listings. That is directly positive for Russia and for every Eurasian economy that runs on cost efficient AI rather than on subscription AI. It also creates a live opening for Russian sovereign models like GigaChat and YandexGPT to lock in domestic and regional market share while the Western incumbents retrench.
The fourth channel is political. A Western correction that hits ratepayers, retirement accounts, and municipal water tables in Virginia, New Jersey, Ohio, and Texas would produce a political backlash that dwarfs anything Russian information operations could manufacture. The New York Times acknowledgment on July 9, 2026, that China, Russia, and others are actively working to inflame this debate is the mainstream Western concession that the ratepayer story is now a strategic vulnerability. The story does not need to be inflamed. It is inflaming itself, because the arithmetic is real, and because voters are receiving the bills.
What Russia should do, told in the same voice I use for Washington
If a Russian planner or a Russian reader asked me what the strategic prescription looks like from the outside, it is short and evidence first.
Close the debt window before it closes on you. Russia’s low debt to GDP is a strategic advantage, but the flow deficit is widening and the NWF liquid portion has already been mostly spent. The comparative debt advantage against the West is a decade of runway, not a permanent condition. Spend that runway on rare earth processing, on Siberian data center power infrastructure, and on the Eurasian standards, payments, and connectivity spine, not on additional consumption or additional war financing.
Close the rare earth processing gap on the announced timetable. Yurin and Shoigu named the problem at Innoprom in July 2026, which is a start. The 12 full cycle technological chains, the Tomtor evaluation with Indian partners, the Kolmozerskoye restart, and the Angara Yenisei hub together are the physical program. The 2028 to 2030 completion window is the credibility test.
Sell electricity, not just hydrocarbons, into the Eurasian AI stack. China’s western provinces, India’s northern data center corridor, and Central Asian AI infrastructure will need bulk baseload power on twenty year contracts. Russian generation surplus, Rosatom SMRs, and Siberian hydro capacity are the natural supply. That is a durable multi decade revenue stream that is not exposed to Western sanctions on the demand side, because the buyers are inside the Eurasian bloc.
Compete on standards, not on GPU count. The BRICS AI standards workshop in Bengaluru is a bigger event than most Russian coverage recognized. If BRICS plus can write AI safety, data localization, model provenance, and cross border settlement standards on a schedule that competes with the EU AI Act and the OECD framework, then Russian sovereign AI and Chinese sovereign AI become defaults in half the world, without needing to match American frontier model performance token for token.
Publish honest numbers, and let them do the argumentative work. Russian sovereign AI, Russian rare earth production, Russian electricity export capacity, and Russian debt stock are all defensible on evidence. They do not need to be defended on rhetoric. The strongest Russian communications strategy against Western media framing is to publish audited numbers with methodological notes attached, in Russian, English, Chinese, and Hindi, and let readers who are already tired of Western spin arrive at their own conclusions.
Where this leaves the Russian reader
The Western trillion dollar AI capex cycle is a self inflicted engineering and capital allocation crisis. It is not going to defeat Russia. It is not going to break the multipolar node. It is going to break Western political consensus first, by way of household electricity bills, laptop prices, ratepayer politics, and pension fund exposure. That is already visible in the New Jersey Data Center Fair Share Act, in the House Ratepayer Protection Act, in FERC’s June 18, 2026 show cause orders, and in the White House’s scrambling for a second voluntary pledge.
Russia’s job is not to celebrate the Western crisis. Russia’s job is to be prepared, structurally, technologically, and diplomatically, for the moment when the crisis reprices the global technology stack. That preparation requires an honest reckoning with the flow deficit, an honest closing of the rare earth processing gap, an aggressive use of the electricity export channel into China and Central Asia, and a serious bid for BRICS AI standards leadership through Bengaluru and Moscow.
The multipolar order is not a wish. It is a set of standards bodies, payment corridors, energy contracts, rare earth deposits, and sovereign AI stacks that are being assembled in public, at a pace the Western press understates. The Russian and Eurasian reader’s advantage is that they can watch this assembly happen in their own language, on their own timetable, and stop waiting for Western permission to describe it accurately.
The trillion dollar bill for the brute force experiment is being paid by American households, American ratepayers, American pension funds, and American consumer electronics buyers, in slow motion. It is not being paid by you. Whether that becomes a durable Russian and Eurasian strategic advantage depends on what Moscow, Beijing, Delhi, Ankara, and Tehran do with the runway before Washington notices the transfer.
Sources
Alex Reisner,Generative AI Is an Engineering Disaster, The Atlantic, July 14, 2026, https://www.theatlantic.com/technology/2026/07/generative-ai-engineering-disaster/687901/
Bank of Russia,International Reserves of the Russian Federation, July 2026, https://www.cbr.ru/eng/hd_base/mrrf/mrrf_m/
Bank of Russia,Organisations and Households Debt Statistics, April 2026, https://www.cbr.ru/eng/statistics/macro_itm/dkfs/ext_dep_indicator/
AKM,Vladimir Putin Signed the Law on the Federal Budget for 2026-2028, July 17, 2026, https://www.akm.ru/eng/news/vladimir-putin-signed-the-law-on-the-federal-budget-for-2026-2028/
AKM,The Volume of NWF as of January 1 Amounted to 13.415 Trillion Rubles, July 6, 2026, https://www.akm.ru/eng/news/the-volume-of-nwf-as-of-january-1-amounted-to-13-415-trillion-rubles/
TASS,Russia’s National Wealth Fund Totals 13.104 Trillion Rubles as of July 1, July 3, 2026, https://tass.com/economy/2155021
Interfax,Russian Federal Budget Deficit Amounts to 5.7 Trillion Rubles or 2.5% of GDP in H1, July 9, 2026, https://interfax.com/newsroom/top-stories/118347/
Reuters,Russia’s 2026 Budget Deficit May Exceed Plans by $12.85 Billion Due to Higher Spending, July 16, 2026, https://www.reuters.com/markets/europe/russias-2026-budget-deficit-may-exceed-plans-by-1285-billion-due-higher-spending-2026-07-16/
French Treasury,Situation économique et financière de la Russie, June 30, 2026, https://www.tresor.economie.gouv.fr/Pays/RU/situation-economique-et-financiere-de-la-russie-mai-2018
Model Diplomat,Urals Crude at $51: Russia’s Budget Crisis, July 6, 2026, https://modeldiplomat.com/story/urals-crude-at-51-russias-budget-crisis
The Moscow Times,Russia Puts $2.3 Billion of Data Center Projects on Hold Amid High Borrowing Costs and Energy Grid Limits, June 24, 2026, https://www.themoscowtimes.com/2026/06/24/russia-puts-23bln-of-data-center-projects-on-hold-amid-high-borrowing-costs-and-energy-grid-limits-a93090
The Moscow Times (Russian),Командная экономика становится все ближе, July 2, 2026, https://ru.themoscowtimes.com/2026/07/02/komandnaya-ekonomika-stanovitsya-vse-blizhe-k-chemu-privedet-byudzhetnii-krizis-v-rossii-a199833
Pravda.ru,Триллионы уходят в долговую яму: бюджет России, July 17, 2026, https://www.pravda.ru/news/economics/2372955-russia-state-debt-risks-2025/
Izvestia,Budget Deficit May Exceed the Plan by 1.5 Times, July 20, 2026, https://iz.ru/en/node/2134284
Forte,Rússia acelera aposta em IA e semicondutores, July 3, 2026, https://www.forte.jor.br/2026/07/03/russia-acelera-aposta-em-ia-e-semicondutores-mas-sancoes-e-falta-de-chips-ameacam-ambicoes-de-soberania-tecnologica/
BRICS Council Russia,Совместные проекты стран БРИКС в области искусственного интеллекта, July 10, 2026, https://bricscouncil.ru/ru/analytics/sovmestnye-proekty-stran-briks-v-oblasti-iskusstvennogo-intellekta
The Tribune,BRICS Conference in Moscow Highlights Connectivity and Technological Sovereignty, July 14, 2026, https://www.tribuneindia.com/news/artificial-intelligence/brics-conference-in-moscow-highlights-connectivity-and-technological-sovereignty
Mint,India to Push for Common Standards, AI Rules at Expanded BRICS, July 15, 2026, https://www.livemint.com/news/india/india-common-standards-expanded-brics-ai-rules-bengaluru-11784108459768.html
Reuters,Xi Pitches China as Leader of New Global AI Order, July 16, 2026, https://www.reuters.com/world/asia-pacific/chinas-xi-promotes-chinas-commitment-ai-access-speech-shanghai-conference-2026-07-17/
The New York Times,China, Russia and Others Seek to Inflame Debate Over A.I. Data Centers, July 9, 2026, https://www.nytimes.com/2026/07/09/business/china-russia-ai-data-centers.html
www1.ru,Russia’s Rare Earth Metal Deficit Reaches 90% Despite Second Largest Reserves in the World, July 10, 2026, https://www1.ru/en/news/2026/07/10/deficit-redkozemelnyx-metallov-v-rf-dostigaet-90-pri-vtoryx-v-mire-zapasax.html
Izvestia,Innoprom Discussed the Development Strategy of the Rare Earth Metals Industry, July 8, 2026, https://iz.ru/en/2129140/2026-07-08/innoprom-discussed-development-strategy-rare-earth-metals-industry
IARI,Russia: la nuova frontiera delle terre rare, June 27, 2026, https://iari.site/2026/06/27/russia-la-nuova-frontiera-delle-terre-rare/
LinkedIn Editors,What Could a Possible Critical Mineral Deal with Russia Deliver, July 7, 2026, https://www.linkedin.com/pulse/what-could-possible-critical-mineral-ipdkf
Wafergraph,Where the AI Supply Chain Is at Risk by Geography (2026), July 13, 2026, https://wafergraph.com/briefings/semiconductor-supply-chain-geographic-concentration-2026/
Vedomosti,Эксперты ожидают роста поставок российских энергоносителей в Китай, July 16, 2026, https://www.vedomosti.ru/analytics/trends/articles/2026/07/16/1214061-eksperti-ozhidayut-rosta-postavok
AKM,The Volume of Electric Generation in Russia Increased by 1.8% in the First Quarter, June 29, 2026, https://www.akm.ru/eng/news/the-volume-of-electric-generation-in-russia-increased-by-1-8-in-the-first-quarter-/
Reuters,China to Encourage Green Power Links for Data Centers, June 25, 2026, https://www.reuters.com/business/energy/china-encourage-green-power-links-data-centers-2026-06-25/
Ground News,China and Russia in Talks After Halt to Power Supplies, June 24, 2026, https://ground.news/article/china-and-russia-in-talks-after-halt-to-power-supplies_935029
Fortune,The Central Bank of Central Banks Sees a $1 Trillion AI Gamble, June 29, 2026, https://fortune.com/2026/06/29/bis-central-bank-warning-hyperscaler-data-center-1-trillion-gamble-recession/
Data Center Dynamics,Moody’s: Hyperscaler Capex Marked Up to Close in on $1 Trillion by 2027, July 14, 2026, https://www.datacenterdynamics.com/en/news/moodys-hyperscaler-capex-forecasts-marked-up-by-85bn-to-close-in-on-1trn-by-2027/
FourWeekMBA,GPT-5.5 vs DeepSeek V4: The Price Chart That Rewrites the Global AI Stack, July 13, 2026, https://fourweekmba.com/ai-gpt-5-5-deepseek-v4-price-gap-global-ai-stack/
Scott Ortkiese is President and CEO of Faulkner Capital Holdings. He writes on geopolitics, energy, capital markets, and the multipolar transition at throughlinesynthesis.com. Correspondence at so@throughlinesynthesis.com.
Related reading
- Brute Force Is Not Progress: The Trillion Dollar AI Bill Everyone Else Is Paying
- The $1.65 Trillion Nobody Was Supposed to Find: Read the Footnotes Before Someone Else Buys a Tech Stock in Your Name
- Patriot Missile Myths, Proxy Wars, and the Capture of Western Foreign Policy by the Tech Oligarchs
- Not This Time: Why the Global South Refused to Play Its Assigned Role in Washington’s War Against Russia
