<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Vicente Scoutt on Medium]]></title>
        <description><![CDATA[Stories by Vicente Scoutt on Medium]]></description>
        <link>https://medium.com/@vicente_scoutt?source=rss-24567b27da74------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*N8B8jJm-fYDLVzPaVcsFxA.png</url>
            <title>Stories by Vicente Scoutt on Medium</title>
            <link>https://medium.com/@vicente_scoutt?source=rss-24567b27da74------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 21:06:32 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@vicente_scoutt/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[NRED May Be Building Two Businesses Under One Ticker]]></title>
            <link>https://medium.com/the-investors-handbook/nred-may-be-building-two-businesses-under-one-ticker-e21c6eee388f?source=rss-24567b27da74------2</link>
            <guid isPermaLink="false">https://medium.com/p/e21c6eee388f</guid>
            <category><![CDATA[business]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[copper]]></category>
            <category><![CDATA[mining]]></category>
            <category><![CDATA[investing]]></category>
            <dc:creator><![CDATA[Vicente Scoutt]]></dc:creator>
            <pubDate>Tue, 21 Jul 2026 16:32:39 GMT</pubDate>
            <atom:updated>2026-07-22T01:42:19.360Z</atom:updated>
            <content:encoded><![CDATA[<p><strong><em>Wilmac remains an exploration story. MetalCore is beginning to raise a very different set of questions.</em></strong></p><p>Most junior mining companies can be explained in a few sentences. They acquire a property, identify targets, raise money, and drill. After that, the geology decides how interesting the story becomes.</p><p>NRED no longer fits quite as neatly into that description.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/436/1*E2y6vyttExh-XX6NyKTVoQ.png" /></figure><p>The company still has a recognisable mineral exploration business. Wilmac covers approximately 16,078 hectares in British Columbia’s Quesnel porphyry belt, about 10 kilometres west of the operating Copper Mountain Mine. The property is being advanced as a broader copper-gold-platinum exploration system, with associated components also being assessed through historical records, sampling, and geophysics.</p><p><a href="https://www.newsfilecorp.com/release/301870/NovaRed-Mining-Announces-Planned-2026-Field-Program">Its 2026 field program looks much like the work expected from an exploration company. NRED has collected 970 soil samples and outlined four IP/AMT survey grids at North Lamont, West Lamont, Wilmac, and Plume.</a> Initial drilling is planned for the fall, subject to the required drill permit being obtained or transferred and approved.</p><p>This part of the company develops slowly and physically. A soil anomaly has to be followed on the ground. A geophysical interpretation has to be tested beneath the surface. Eventually, a drill bit must pass through the target and produce an answer that cannot be generated by software.</p><p>MetalCore introduces a different rhythm.</p><p>The platform has grown beyond 4.1 million analytical records across Canada and the United States. Its dataset includes roughly 1.5 million geochemical records, 1.4 million oil and gas wells, more than 800,000 deposit records, close to 200,000 geothermal records, and additional magnetic, satellite, land-tenure, and geophysical information.</p><p><a href="https://novaredmining.com/news/novared-mining-to-close-metalcore-early-access-waitlist-in-48-hours-as-platform-dataset-surpasses-4.1-million-records">About 920 people registered for an early-access cohort limited to 1,000 users.</a></p><p>Registrations do not yet make MetalCore a software business. Still, they suggest that the platform is attracting interest beyond the people directly working on Wilmac.</p><p>That is where the NRED story becomes less conventional.</p><p>MetalCore could remain an internal tool used to evaluate the company’s own properties. It could also develop into something external: paid property assessments, project licences, subscriptions, partnerships, or access for larger exploration teams. The exact model is still unclear, and that uncertainty is part of what makes it worth watching.</p><p>A mining project and a software platform create evidence in very different ways.</p><p>At Wilmac, progress will be measured through completed surveys, stronger targets, drill holes, mineralisation, and geological interpretation. MetalCore will have to prove itself through actual use. Did a customer pay for a pilot? Did the platform change where a company sampled or drilled? Did someone return with a second property?</p><p>The two sides may also strengthen each other.</p><p>Wilmac gives MetalCore a practical environment rather than a staged demonstration. The platform has already brought together information covering 10 known mineral occurrences, approximately 19 historical assessment reports dating from 1968 to 2025, 38 regional geochemical samples, and aeromagnetic data.</p><p>That information can be compared with the company’s soil program and planned geophysics. Later, drilling can test whether the targets selected through the combined dataset were worth pursuing. It creates a useful loop: historical information shapes fieldwork, and new field results return to improve the next interpretation.</p><p>MetalCore could also influence how NRED evaluates opportunities beyond Wilmac. A platform capable of comparing large land packages and fragmented historical datasets may help the company review potential acquisitions faster or identify geological patterns that are difficult to see across separate archives.</p><p>Recent appointments suggest that management is giving the technology side more structure. NRED named Valentin Saitarli, who has more than a decade of experience in machine learning and computer vision, as chief technology officer. The company also entered a 30-day standstill agreement with EyeX to examine a possible interest in AI surveillance technology for industrial and mining environments.</p><p>The proposed corporate name, NRED Intelligent Mining Inc., subject to approval, offers another clue. It suggests that management wants the market to view technology as part of the company’s identity rather than as a small internal experiment.</p><p><a href="https://novaredmining.com/">I would not value the two sides separately yet. The exploration program still needs geological results, while MetalCore still needs evidence of commercial adoption. But they are already creating different possible paths for the company.</a></p><p>Wilmac offers exposure to the search for copper, gold, platinum, and associated components. MetalCore could eventually become a product used by other teams trying to search more efficiently.</p><p>The next few milestones should make the picture clearer. A completed geophysical program or drill campaign would move the mineral side forward. A paid pilot, external licence, or commercial partner would do the same for MetalCore.</p><p>Until then, NRED sits in an unusual position for a junior explorer. Part of the company is trying to discover value underground. Another part is trying to build a tool that could help decide where exploration capital goes before drilling begins.</p><p>📩 <strong>Join </strong><a href="https://investorshandbook-newsletter-c72f94.beehiiv.com/"><strong>Investor’s Handbook Digest</strong></a><strong> — get the best investing, markets, and wealth-building insights each week.</strong></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e21c6eee388f" width="1" height="1" alt=""><hr><p><a href="https://medium.com/the-investors-handbook/nred-may-be-building-two-businesses-under-one-ticker-e21c6eee388f">NRED May Be Building Two Businesses Under One Ticker</a> was originally published in <a href="https://medium.com/the-investors-handbook">Investor’s Handbook</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[50 Millions De Dollars Pour Des Puces Qui Font Circuler La Lumière]]></title>
            <link>https://medium.com/no-time/50-millions-de-dollars-pour-des-puces-qui-font-circuler-la-lumi%C3%A8re-c0e49bc89847?source=rss-24567b27da74------2</link>
            <guid isPermaLink="false">https://medium.com/p/c0e49bc89847</guid>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[investing]]></category>
            <category><![CDATA[infrastructure]]></category>
            <category><![CDATA[industry]]></category>
            <category><![CDATA[ai]]></category>
            <dc:creator><![CDATA[Vicente Scoutt]]></dc:creator>
            <pubDate>Wed, 08 Jul 2026 19:09:19 GMT</pubDate>
            <atom:updated>2026-07-08T19:09:19.846Z</atom:updated>
            <content:encoded><![CDATA[<p><strong><em>Coherent au Texas n’est pas seulement une expansion semiconductor. C’est un rappel que l’AI infrastructure ne repose pas uniquement sur les GPU, mais aussi sur la photonics, les cleanrooms, les wafers, l’optical networking et une production très physique.</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EtcHT0cMEzTc3I3TCOVrNA.png" /></figure><p>Quand les gens parlent d’AI infrastructure, ils imaginent généralement des serveurs. Puis des GPU. Puis des data centres, de l’électricité, du cooling et NVIDIA.</p><p>Mais entre un AI model et un data centre, il existe une autre couche dont on parle beaucoup moins : l’optical networking. Les données ne doivent pas seulement être calculées. Elles doivent circuler rapidement entre chips, racks, servers et facilities. Et c’est précisément là que l’économie “digitale” se heurte soudain à la lumière, aux wafers, aux cleanrooms et aux usines.</p><p><a href="https://www.coherent.com/news/press-releases/a-chip-letter-of-intent-for-50m-to-expand-world-leading-manufacturing-facility-for-ai-infrastructure">Coherent a annoncé une CHIPS letter of intent de $50 million pour agrandir son semiconductor manufacturing facility à Sherman, Texas. Cet argent doit soutenir la growing demand for optical networking technologies qui alimentent les AI datacenters, tout en renforçant le partnership de Coherent avec NVIDIA.</a></p><p>À première vue, c’est simplement une nouvelle CHIPS Act de plus. Mais les détails sont plus intéressants que le titre. L’expansion doit double manufacturing production space et quadruple wafer production capacity. Il ne s’agit donc pas d’une belle présentation autour de l’AI, mais de surface industrielle réelle, de wafer output, de tools, de cleanroom workflow, d’emplois techniques et d’une supply chain qui doit suivre le data centre buildout.</p><p><a href="https://www.nist.gov/chips/coherent-texas-sherman">NIST décrivait déjà le Sherman site comme une existing 700,000 square-foot facility, où le proposed CHIPS investment doit soutenir la modernization and expansion of a state-of-the-art manufacturing cleanroom.</a> L’objectif est d’ajouter des advanced wafer fabrication equipment pour la world’s first 150mm indium phosphide manufacturing line, afin de produire des InP optoelectronic devices at scale. Ces devices sont utilisés dans les datacom and telecom transceivers, l’AI infrastructure, l’advanced sensing, le médical, l’automotive et la consumer electronics.</p><p>Voilà une autre façon de regarder l’AI. Pas un GPU cluster de plus. Pas un campus rendering de plus. Mais de l’indium phosphide, des optical transceivers, des cleanroom tools, des wafers, des lasers, de la photonics et du manufacturing yield. Les AI systems ne consomment pas seulement du compute. Ils font circuler en permanence d’énormes volumes de données à travers une optical layer. Si cette couche est faible, le compute coûteux devient moins utile.</p><p><a href="https://blogs.nvidia.com/blog/coherent-texas-ai-optical/">Dans son propre article, NVIDIA relie directement l’expansion de Coherent à Sherman à l’AI infrastructure. La société écrit que Coherent a annoncé un $50 million CHIPS Act grant pour l’expanded Sherman facility, après avoir déjà reçu environ $17 million de support du Texas CHIPS program et de la Sherman Economic Development Corporation.</a> NVIDIA mentionne aussi son engagement plus large à produire up to $500 billion of AI infrastructure in the U.S. through industry partnerships avec de nouveaux sites en Arizona et au Texas.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7C47_Fqw4ViOr-rY3i4RbQ.png" /></figure><p>Il faut éviter d’exagérer. Une seule usine ne résout pas toute l’AI supply chain. Mais elle montre très bien à quel point l’AI boom est devenu physique. Pour qu’un modèle fonctionne plus vite, il ne faut pas seulement plus de chips. Il faut plus d’optical links, de photonic components, de wafers, de testing equipment, de technicians, d’engineers, de cleanroom capacity et de process control.</p><p><a href="https://www.semiconductor-today.com/news_items/2026/jun/coherent-170626.shtml">Coherent affirme qu’une fois le projet terminé, le Sherman site devrait créer more than 1,000 jobs, dont more than 550 direct advanced manufacturing, engineering and technical roles</a>. Ce n’est pas une “digital economy” abstraite. Ce sont des personnes près des tools, des production lines, des process recipes, des maintenance schedules, des yield problems et des quality checks.</p><p>Une cleanroom n’est pas simplement une pièce propre. C’est du HVAC, de la filtration, du humidity control, du vibration control, de la power stability, des sensors, des valves, des pumps, de l’exhaust, des safety systems et du backup power. La wafer fabrication ne pardonne pas le chaos. Un petit écart peut signifier un yield plus bas, des delays et une supply chain plus coûteuse.</p><p>La photonics a aussi son côté matériel. Optical networking sonne léger, presque immatériel, parce qu’il s’agit de lumière. Mais pour que la lumière fonctionne dans l’AI infrastructure, il faut de l’epitaxy, de la lithography, du packaging, de l’alignment, des lasers, des modulators, des detectors, des substrates, des testing stations et des équipements précis. Cette lumière passe quand même par une usine.</p><p>Dans ce type de nouvelles, le cuivre n’est pas le personnage principal, mais il est tout près. Semiconductor fabs, cleanrooms, optical networking factories et AI data centres ont besoin de transformers, switchgear, power distribution, grounding, motors, pumps, control wiring, sensors, cable trays, chillers et backup systems. Même quand la core technology est la photonics ou l’InP, l’usine physique autour d’elle vit grâce à l’électricité, au refroidissement et au métal.</p><p><a href="https://novaredmining.com/news/novared-mining-provides-update-on-2026-exploration-program-at-its-wilmac-copper-gold-project">À l’extrémité la plus précoce de cette chaîne, il n’y a pas seulement les grands copper producers. Il y a aussi des exploration stories où le futur métal ressemble encore à un field program, de la geophysics et des drill targets. Un de ces exemples est NovaRed. </a>Dans ce thème, il me semble plus logique de ne pas répéter uniquement le data-platform angle, mais de regarder la partie terrain : sur le Wilmac Copper-Gold Project, la société prévoyait expanded soil geochemistry et four IP/AMT geophysical surveys sur les North Lamont, West Lamont, Wilmac et Plume grids, pendant que le drill permit process avançait vers un initial drill program.</p><p>C’est loin d’une indium phosphide wafer line à Sherman. Mais la direction est la même : la digital economy devient plus lourde. Plus d’AI ne signifie pas seulement plus de chips. Cela signifie plus d’optical transceivers, plus de fabs, plus de cleanroom equipment, plus de power infrastructure, plus de cooling, plus de cables et plus de specialized materials. Et bien avant tout cela, quelqu’un doit trouver et préparer les ressources de base sur lesquelles repose la partie physique du système.</p><p>L’expansion de Coherent est intéressante précisément parce qu’elle se tient entre deux mondes. D’un côté — AI, NVIDIA, data centres et optical networking. De l’autre — une usine au Texas, des cleanrooms, des wafers, des engineers, des tools et de la production capacity. Le marché aime la première partie parce qu’elle semble immense. Mais la deuxième décide si cette échelle peut réellement être produite.</p><p>L’AI est souvent décrite comme une software revolution. Mais chaque nouvelle de ce type casse un peu cette image. Si une AI plus grande nécessite de nouvelles InP lines, des cleanroom expansions, des optical components et des milliers de manufacturing jobs, alors ce n’est plus seulement du software.</p><p>Ce sont des usines, de la lumière, du métal, de l’électricité et une discipline physique très complexe.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c0e49bc89847" width="1" height="1" alt=""><hr><p><a href="https://medium.com/no-time/50-millions-de-dollars-pour-des-puces-qui-font-circuler-la-lumi%C3%A8re-c0e49bc89847">50 Millions De Dollars Pour Des Puces Qui Font Circuler La Lumière</a> was originally published in <a href="https://medium.com/no-time">No Time</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[A Small Reactor Ends With Big Infrastructure]]></title>
            <link>https://medium.com/@vicente_scoutt/a-small-reactor-ends-with-big-infrastructure-b00d3f43f6e2?source=rss-24567b27da74------2</link>
            <guid isPermaLink="false">https://medium.com/p/b00d3f43f6e2</guid>
            <category><![CDATA[energy]]></category>
            <category><![CDATA[infrastructure]]></category>
            <category><![CDATA[industry]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[technology]]></category>
            <dc:creator><![CDATA[Vicente Scoutt]]></dc:creator>
            <pubDate>Thu, 02 Jul 2026 18:47:51 GMT</pubDate>
            <atom:updated>2026-07-02T18:47:51.670Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/910/1*vm0sgp9XA2U9GwU-4SYiWQ.png" /></figure><p><strong>Oklo is getting attention because of a nuclear milestone. But between the reactor and electricity for AI, there is not a headline — there is fuel, a turbine, cables, transformers, and approval.</strong></p><p>Oklo again looks like a very convenient story for the market. A small reactor, clean baseload, data centers, Sam Altman in the background, nuclear renaissance, the AI power problem. Everything that is easy to sell to investors right now is gathered in one place.</p><p>But there is a trap here. The market often hears “reactor going live” and almost automatically imagines electricity for a data center. In reality, a long industrial corridor stands between those two things.</p><p><a href="https://www.barrons.com/articles/oklo-stock-nuclear-reactor-going-live-6ab671a3">Barron’s writes that Oklo’s first reactor is expected to reach criticality this month. The important detail: the Groves Isotope Test Reactor in Texas will not produce electricity</a>. It is a test reactor for radioisotopes used in medicine, manufacturing, research, space, and national security. So the headline sounds like a nuclear power story, but physically, this is not yet a power plant for an AI campus.</p><p>Oklo announced on July 1, 2026, that the U.S. Department of Energy approved the Documented Safety Analysis for the Groves Isotope Test Reactor under DOE’s Reactor Pilot Program. That is an important step because the DSA includes hazard analysis, safety controls, and operating requirements needed before startup. But final pre-startup review, operational authorization, fuel loading, and only then criticality still remain.</p><p>That is why this news is more interesting than the OKLO stock move itself. It shows that a nuclear story does not scale like software. Even a small advanced reactor means fuel, a safety case, a regulatory pathway, site work, heat removal, control systems, sensors, pumps, valves, cables, security, and a team that cannot afford to work “roughly correctly.”</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/838/1*jP1QaOxfu5NuVrnMqxsoTA.png" /></figure><p><a href="https://www.nrc.gov/reactors/new-reactors/advanced/who-were-working-with/pre-application-activities/okla-aurora-powerhouse">Groves is one layer. Aurora is another. The NRC describes Aurora Powerhouse as a liquid metal-cooled, metal-fueled fast reactor with a maximum power level of 75 MWe</a>. That is closer to the story that interests AI infrastructure, industrial load, and data centers. But even here, the word “small” is a little misleading. The reactor itself may be small compared with a traditional nuclear plant. The infrastructure around it is still large.</p><p>The least glossy part of this story is the conventional island. A reactor by itself does not yet mean electricity in the grid. It needs a steam turbine, generator, auxiliaries, power conversion, and grid interface. <a href="https://www.powermag.com/in-a-first-for-advanced-nuclear-siemens-energy-turbine-package-advances-for-oklos-aurora-inl/">POWER wrote that Siemens Energy is manufacturing the turbine-generator package for Oklo’s first Aurora powerhouse at Idaho National Laboratory: an SST-600 steam turbine, an SGen-100A generator, and associated systems</a>. This is no longer a presentation. It is procurement, factories, mechanical interfaces, and equipment that has to operate for years.</p><p>That is why the nuclear trade around AI power is not as clean as it looks. Data centers want reliable electricity. But reliable electricity does not appear from the word nuclear alone. The plant has to be built, cables have to be brought in, switchgear, transformers, protection systems, control cabinets, cooling equipment, and a substation have to be installed, and agreements have to be made with the grid or a specific industrial customer.</p><p>And this is where copper appears again, not in the headline, but in the real system. A nuclear plant, data center, grid interconnection, and industrial load need conductors, busbars, cables, transformers, motors, sensors, relays, grounding, and cooling systems. If small nuclear, AI data centers, battery storage, EV charging, and grid upgrades all grow at the same time, they are not just competing for electricity. They are expanding the physical demand for copper.</p><p>That is why the future shortage of copper matters here. New nuclear capacity may look like an answer to the power problem, but it still has to be connected with metal. And new copper does not appear after one regulatory approval. It has to be found, tested, drilled, financed, permitted, and moved through concentrate, smelting, refining, and finally into material for industry.</p><p>Against that backdrop, NovaRed looks not like a ready-made solution, but like an early-stage copper-gold exploration example worth watching because of a possible copper supply squeeze. Wilmac in British Columbia has an option structure for a 70% interest and sits in the Quesnel porphyry belt, roughly 10 km west of Hudbay’s producing Copper Mountain Mine. This is not proof of a discovery and not a production story. But this is exactly the level where future copper begins: land, geology, field work, permits, drilling, and capital.</p><p><a href="https://novaredmining.com/">NRED has several specific early signals that still need to be tested</a>. North Lamont returned copper-in-soil up to 1,125 ppm, and the 2026 program includes expanded soil sampling and four IP/AMT geophysical surveys across North Lamont, West Lamont, Wilmac, and Plume. Initial drilling depends on an approved drill permit. The early-stage risk is obvious, but if AI power, nuclear buildout, grid upgrades, and data centers all put pressure on physical copper at the same time, projects like this become more visible long before production.</p><p>Oklo could become an important part of the future power story. But Groves reminds us of a simple thing: a nuclear headline is not yet electricity for AI. Between them stand fuel, a safety review, a turbine package, a cable, a transformer, a substation, a regulator, and time.</p><p>The reactor may be small. Everything that has to be built around it is not.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b00d3f43f6e2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Green Metal Has Found a Second Buyer]]></title>
            <link>https://medium.com/@vicente_scoutt/green-metal-has-found-a-second-buyer-0d32d4ef1321?source=rss-24567b27da74------2</link>
            <guid isPermaLink="false">https://medium.com/p/0d32d4ef1321</guid>
            <category><![CDATA[semiconductors]]></category>
            <category><![CDATA[stocks]]></category>
            <category><![CDATA[materials]]></category>
            <category><![CDATA[canada]]></category>
            <category><![CDATA[copper]]></category>
            <dc:creator><![CDATA[Vicente Scoutt]]></dc:creator>
            <pubDate>Thu, 25 Jun 2026 19:04:36 GMT</pubDate>
            <atom:updated>2026-06-25T19:04:36.761Z</atom:updated>
            <content:encoded><![CDATA[<p><strong><em>Canada sells critical minerals as the foundation for batteries, EVs and clean energy. But when the Pentagon enters the same story, the metal starts to sound like more than just climate raw material.</em></strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UTAbbxMxJmtofO7D5lvqkA.png" /></figure><p>There is one uncomfortable truth in critical minerals: metal does not ask who will take it after extraction. Copper can go into an electrical grid, a transformer or a data center. Or it can go into military electronics. Graphite can become part of an EV battery, or it can end up in a defense supply chain. Tungsten, cobalt, nickel, rare earths — all these words look good in clean-tech presentations, but they are just as necessary for drones, radars, missiles, submarines and armor.</p><p><a href="https://progressive.international/wire/2026-06-25-elbows-up-canada-is-letting-pentagon-take-unprecedented-stakes-in-canadian-mines/ru/">That is why the Progressive International / The Breach piece stands out. It is not simply criticizing mining. It shows how quickly the language of “energy transition” can turn into the language of “national security.”</a></p><p>According to the article, in 2024–2025 the Pentagon invested more than $78 million in six companies with Canadian mining projects. The list includes Fortune Minerals, Lomiko Metals, Fireweed Metals, Northcliff Resources, Nano One Materials and Electra Battery Materials. Trilogy Metals stands out separately — $35.6 million for an approximately 10% stake. Another detail is the U.S. DOE’s 5% equity stake in Lithium Americas through warrants.</p><p>This is no longer just “we need raw materials.” It is “we want a seat inside the future route of those raw materials.”</p><p><strong>A critical mineral does not become peaceful just because it is called green.</strong></p><p>The official logic is easy to understand. China controls many parts of the supply chain. The U.S. and Canada want to depend less on Beijing. That is where the language comes from: friend-shoring, North American supply chains, strategic autonomy, the clean energy economy. On a presentation slide, it all looks clean: batteries, EVs, grid storage, jobs, domestic processing.</p><p>But real metal rarely stays inside one slide.</p><p>If the Pentagon enters a mining project, it becomes hard to pretend this is only about climate. The Defense Production Act does not exist for pretty ESG reports. Its logic is simple — to secure the material base for the American industrial and defense system.</p><p>And this is where Canada starts to look a little uncomfortable. When Chinese state capital enters critical minerals, it is called a security risk. When American state capital makes a similar move, it is called partnership. Maybe the U.S. really is a closer ally. But the difference in language is too obvious to ignore.</p><p><strong>Sovereignty does not begin with a flag above a mine. It begins with the question of who controls the future metal.</strong></p><p>I would not reduce this story to “mining versus nature.” That is a weak angle. Without mining, there will be no power grids, data centers, batteries, transformers, EVs, cables, wind turbines or even defense industry. The question is not whether extraction is needed. The question is who it really serves, and for what purpose.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/849/1*Lg0NU2Lm-u-l4-CDGZt7GA.png" /></figure><p>In this wider copper/materials layer, it is worth looking not only at major mines, but also at early-stage projects that are still shaping the future supply map. Through the <a href="https://novaredmining.com/">Wilmac Copper-Gold Project in British Columbia, NRED</a> is working in exactly that early part of the chain — 16,078 hectares in the Quesnel porphyry belt, approximately 10 km west of the Copper Mountain Mine. In 2026, the company plans expanded soil sampling, IP/AMT surveys across the North Lamont, West Lamont, Wilmac and Plume grids, as well as an initial drill program subject to an approved permit. This is not a defense story and not a production claim — it is an example of how future copper begins long before the cable, battery or substation.</p><p>If communities are told, “this project is needed for clean energy,” that is one conversation. If there is also a defense supply chain behind the project, that is a different conversation. And people have the right to know that before a road, mine site, processing plant or offtake agreement becomes a fact.</p><p>Especially when Indigenous lands are involved. The article mentions the La Loutre graphite project, the Ambler access road and the Sisson tungsten project. In these cases, the words “critical minerals” can quickly become pressure: faster because of China; faster because of climate; faster because of national security. But speed cannot replace consent.</p><p>For investors, the lesson is also simple: critical minerals can no longer be read only through EV demand charts. You have to look at who is financing the project, who receives equity, who holds warrants, who controls future offtake, and whether a defense buyer is standing behind the “green” rhetoric.</p><p>Canada can build a clean-tech economy. It can be a raw-material partner for the U.S. It can strengthen the North American supply chain. But then this story has to be called what it is — not only climate policy, but also military-industrial policy.</p><p>Because a battery and a missile sometimes begin at the same mine.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0d32d4ef1321" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AI Started Eating Everything Around the Chip]]></title>
            <link>https://medium.com/@vicente_scoutt/ai-started-eating-everything-around-the-chip-d446d3217e1f?source=rss-24567b27da74------2</link>
            <guid isPermaLink="false">https://medium.com/p/d446d3217e1f</guid>
            <category><![CDATA[business]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[stocks]]></category>
            <category><![CDATA[energy]]></category>
            <dc:creator><![CDATA[Vicente Scoutt]]></dc:creator>
            <pubDate>Thu, 18 Jun 2026 14:43:33 GMT</pubDate>
            <atom:updated>2026-06-18T14:43:33.931Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YPBZTQ4DD0jHFsZtwf7a9A.png" /></figure><p><strong>116% growth in data center components looks like a victory for the semiconductor market. But behind that number, something else is visible: the AI boom has already moved beyond GPUs and started pulling memory, networking, storage, electricity, cooling and metal along with it.</strong></p><p>When the market talks about AI infrastructure, it almost automatically talks about NVIDIA. That makes sense. GPUs have become the symbol of this cycle. But new data from Dell’Oro Group shows that the story is already much broader. In 1Q 2026, worldwide revenue for Data Center IT Semiconductors and Components for servers and storage systems increased 116% year-over-year. This is no longer just “more GPUs.” It is the expansion of the entire machine around AI.</p><p><a href="https://www.lightreading.com/data-centers/data-center-it-semiconductor-and-component-revenue-increased-116-in-1q-2026-dell-oro">Light Reading writes that the growth was driven by ongoing AI infrastructure expansion and rising memory prices. Dell’Oro also expects this market to remain on track for triple-digit growth for full-year 2026. For me, the key point is not only the 116% number. The key point is that AI is no longer just a software story. It is becoming a component story.</a></p><p><strong>AI does not scale with one chip. It scales with everything that keeps that chip working without pause.</strong></p><p>There is something boring about the word “components,” but that is exactly where the main shift is hidden. AI clusters need more than accelerators. They need HBM, DRAM, NAND, high-speed back-end NICs, CPUs, networking silicon, storage systems and an entire architecture that allows thousands of servers to move data without delays.</p><p>Dell’Oro specifically notes that the ramp of merchant and custom accelerators increased demand for adjacent components, including HBM and high-speed back-end NICs. Storage has also become part of this story, because pre-training and post-training workloads require large capacity. AI does not simply “think.” It reads, copies, moves, stores and constantly feeds compute systems with data.</p><p><strong>A model looks immaterial only on the screen. Inside a data center, its appetite has weight, temperature and an electricity bill.</strong></p><p><a href="https://www.delloro.com/news/data-center-it-semiconductor-and-component-revenue-increased-116-percent-in-1q-2026/">That is why the list of the largest vendors is so telling. NVIDIA remained the largest vendor by total revenue, but Samsung and SK Hynix followed behind it.</a> This is no longer just a GPU cycle. It is a memory cycle, a storage cycle and a networking cycle at the same time. HBM, DRAM and NAND have become not background elements, but bottlenecks and revenue engines.</p><p>But even that is not the full picture. When revenue for semiconductors and components grows, the load on the entire physical shell of the data center grows with it. More AI servers mean more power delivery, more cooling, more switchgear, more cabling, more racks, more substations, more backup systems and more grid connection work. The denser AI hardware becomes, the stronger the building around it has to become.</p><p>That is why the data center boom is gradually becoming more than a technology story. It is becoming an infrastructure story. Semiconductors are the top layer. Beneath them are transformers, busbars, copper cables, fiber, liquid cooling loops, power distribution units and substations. Without that, even the best accelerator remains an expensive component with nowhere to work.</p><p><strong>A data center is not a cloud. It is a factory where electricity is turned into computation.</strong></p><p>This is where the copper layer appears. Not as a forced insertion, but as the physical logic of the entire system. AI clusters require more electrical connections, more power distribution, more cooling equipment and more transmission capacity. Copper is not in the Dell’Oro headline, but it is in the cables, busbars, transformers, motors, cooling systems and grid infrastructure without which this semiconductor growth never reaches production.</p><p>In the broader copper-gold theme, that is why it is worth looking not only at large producing mines. Before every cable, transformer or substation, there is a longer road back — to mineral supply, exploration and early maps that still need to be tested. <a href="https://novaredmining.com/"><strong>NovaRed Mining</strong></a> is working on the Wilmac Copper-Gold Project in British Columbia — 16,078 hectares in the Quesnel porphyry belt, roughly 10 km west of Copper Mountain Mine. Its planned 2026 field program includes expanded soil sampling, 4 IP/AMT geophysical surveys across the North Lamont, West Lamont, Wilmac and Plume grids, as well as an initial drill program being considered for fall 2026, subject to an approved permit.</p><p>This is not yet a production story and not a final answer on supply. But this is often what the future material base of an AI-heavy economy looks like at an early stage — not a finished cable inside a data center, but soil data, geophysics, drill targets and an attempt to understand what lies beneath the surface.</p><p>The final point is simple: 116% growth in data center components is not just a strong quarter for the semiconductor market. It is a sign that the AI boom has entered a phase of physical expansion. Now it is not only models and GPU headlines that are growing. Everything between electricity, memory, network and computation is growing too. And that is where the digital economy stops being a light “cloud” and becomes heavy infrastructure.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d446d3217e1f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Ore Does Not Read Strategy]]></title>
            <link>https://medium.com/@vicente_scoutt/ore-does-not-read-strategy-020faf7e5472?source=rss-24567b27da74------2</link>
            <guid isPermaLink="false">https://medium.com/p/020faf7e5472</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[investing]]></category>
            <category><![CDATA[materials]]></category>
            <category><![CDATA[copper]]></category>
            <category><![CDATA[stock-market]]></category>
            <dc:creator><![CDATA[Vicente Scoutt]]></dc:creator>
            <pubDate>Wed, 17 Jun 2026 14:44:45 GMT</pubDate>
            <atom:updated>2026-06-17T14:44:45.504Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/813/1*sNohUMeG9OW4f6lqh3wPdw.png" /></figure><p>Europe can write any critical minerals targets it wants. But first it will have to understand the rock itself — what is inside it, how it breaks, and whether anything useful can actually be taken from it without too much loss.</p><p><a href="https://arxiv.org/abs/2606.05798">What caught my attention in this arXiv paper was not the name Digital Rock Physics itself. It sounds dry, almost like something from a lab where a normal reader would quickly close the tab</a>. What caught me was something simpler: Europe is talking more and more about critical minerals as a security issue, but without a precise understanding of the ore itself, that conversation can easily hang in the air.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/799/1*IpyTPLOhO2gDH9y6tgYA3w.png" /></figure><p>A country can set a target: more extraction by 2030. More processing. More recycling. Less dependence on outside suppliers. But the rock does not become simpler because of that. If the ore is complex, if the metal sits inside it in an awkward way, if old waste still has to be learned and processed properly, then a political phrase does not help much.</p><p>That is what the paper is really about. Digital Rock Physics is a way to look deeper into the material: not just “there is copper here” or “there is lithium here,” but how the minerals sit inside the rock, what pores, cracks and textures are there, and how the material reacts during processing. It sounds technical, but the meaning is very grounded. Before putting money into a supply chain, it is better to understand whether the material can actually be worked with.</p><p>And this is where the question becomes more important than the technology itself: what is more expensive — studying the ore carefully at the beginning, or realizing too late that it is difficult to process? Because with critical minerals, the problem is not only where to find the resource. The problem is also how to turn that resource into something useful for batteries, grids, electronics, defense or AI infrastructure.</p><p>The idea of a Digital Ore Passport is especially interesting. Put simply, it is not a passport in the usual sense, but a digital description of the ore: what it contains, how it behaves, how it can be processed, and what risks or waste may appear. In a world where everyone talks about supply-chain transparency, this no longer sounds strange. If a battery needs a story of origin, why should ore not have one too?</p><p>I like this idea because it cuts through some of the grand language around “new industrial policy.” Countries can make agreements, sign declarations, map supply chains and allocate money. But underneath all of that, there is still a material. It can be convenient or difficult. Rich or poor. Clean or problematic. And that can matter just as much as a major political statement.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/775/1*4olfce55ROqIbzAPuLOYWg.png" /></figure><p>One early junior example in this data-driven logic is <a href="https://novaredmining.com/">NRED. NovaRed has MetalCore</a>, an AI-assisted mineral evaluation platform that works with geological, geochemical and geophysical data and has U.S. patent application №19/680,101. This is not the same thing as the Digital Rock Physics described in the paper, but the broader direction is similar: the future copper chain does not begin only with drilling. It begins with the ability to read weak signals earlier and avoid fooling yourself with a nice-looking map.</p><p>This paper does not promise a quick solution to mineral shortages. And honestly, that is exactly why it feels useful. It does not sell a simple answer. It reminds us that a secure supply chain is not built only from political will. It has to be assembled from rock, data, chemistry, processing, logistics and time.</p><p>Critical minerals have already become part of geopolitics. But before geopolitics, there is a smaller level — a mineral grain, a crack, a pore, a reaction. And if Europe wants to depend less on someone else’s chains, it will have to begin with a very simple, almost uncomfortable question: what is actually inside this ore?</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=020faf7e5472" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Server Room Heated the Ground]]></title>
            <link>https://medium.com/@vicente_scoutt/the-server-room-heated-the-ground-f637f07f87f6?source=rss-24567b27da74------2</link>
            <guid isPermaLink="false">https://medium.com/p/f637f07f87f6</guid>
            <category><![CDATA[business]]></category>
            <category><![CDATA[infrastructure]]></category>
            <category><![CDATA[financial-markets]]></category>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[money]]></category>
            <dc:creator><![CDATA[Vicente Scoutt]]></dc:creator>
            <pubDate>Thu, 11 Jun 2026 14:48:31 GMT</pubDate>
            <atom:updated>2026-06-11T14:48:31.284Z</atom:updated>
            <content:encoded><![CDATA[<p>Stargate looks like a story about AI power. But behind it, something else becomes visible — electricity, water, heat, and metal, without which the “cloud” quickly stops feeling weightless.</p><p>At first, I also looked at Stargate as a story about compute. Oracle, OpenAI, large data centers, more power for AI — all of this sounds like the usual language of the technology market. More models, more chips, more speed. But then one detail in this story became hard to ignore: AI does not just consume electricity. It changes the space around it.</p><p>Al Jazeera writes about data centers not as clean digital infrastructure, but as physical objects that heat the ground, use water, and put pressure on power grids. In 2024, data centers consumed about 415 TWh of electricity, around 1.5% of global consumption. By 2030, that figure could rise to nearly 945 TWh. This is no longer just background noise for technology. It is an energy story of its own.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/845/1*KDDAtOlwk-pezB-lpvEEvA.png" /><figcaption><a href="https://www.aljazeera.com/news/2026/6/11/how-much-heat-does-an-ai-data-centre-produce-and-where-are-they-located">https://www.aljazeera.com/news/2026/6/11/how-much-heat-does-an-ai-data-centre-produce-and-where-are-they-located</a></figcaption></figure><p>What caught my attention most was not even that. The study cited by Al Jazeera shows that after AI data centers began operating, the land surface temperature around them rose by about 2°C on average, and in some cases by as much as 9.1°C. The effect could be seen up to 10 km away. In other words, a data center does not end at the fence. It can change the area around it.</p><p>That is where Stargate in Abilene, Texas, becomes especially telling. The article describes it as an AI supercluster for OpenAI with roughly 1.2–2 GW of power capacity. For comparison, a typical hyperscale data center may require 100–300 MW. And one 100 MW facility, according to estimates, may consume around 2.5 billion litres of water per year. After numbers like that, the word “cloud” starts to sound a little strange. What kind of cloud is it if it needs gigawatts, cooling, land, and grid connection?</p><p>This is where the AI story stops being only about software. If every large cluster needs cables, transformers, substations, cooling systems, backup power, and new transmission lines, then the question is simple: who will supply the physical side of this race? Chips get the main headlines. But electricity does not reach a server room without copper.</p><p>At this point, early copper stories no longer feel like distant background. One junior example is <a href="https://novaredmining.com/">NRED</a>. Its Wilmac Copper-Gold Project in British Columbia should not be read as a ready-made answer for AI infrastructure. It is an early exploration story where signals still need to be tested. But projects like this remind us of something simple: before an AI cluster receives 100 MW or 1 GW, there has to be a long process involving metal, land, geology, and future supply chains.</p><p>I think Stargate carries an important lesson for the whole market. We are used to discussing AI as if it is almost weightless: a model, a prompt, an answer, performance. But what stands behind that answer? How much heat needs to be removed? How much electricity needs to be delivered? How much copper has to be inside the system so that all of this does not remain just a polished presentation?</p><p>AI can respond in seconds. Land, grids, and materials do not move that fast. And maybe that is becoming the real bottleneck now: not ideas, but the physical body of the digital world.</p><p>Stargate shows the scale of the ambition. But even more, it shows the cost of scale. When a server room starts heating the ground around it, the future stops being an abstraction. It becomes a question of electricity, water, copper, and the place where all of it has to actually work.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f637f07f87f6" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[France Is Ready to Spend Millions on Electrification. But What Might It Fail to Buy in Time?]]></title>
            <link>https://medium.com/@vicente_scoutt/france-is-ready-to-spend-millions-on-electrification-but-what-might-it-fail-to-buy-in-time-abcf422f5496?source=rss-24567b27da74------2</link>
            <guid isPermaLink="false">https://medium.com/p/abcf422f5496</guid>
            <category><![CDATA[canada]]></category>
            <category><![CDATA[stocks]]></category>
            <category><![CDATA[copper]]></category>
            <category><![CDATA[mining]]></category>
            <category><![CDATA[investing]]></category>
            <dc:creator><![CDATA[Vicente Scoutt]]></dc:creator>
            <pubDate>Fri, 29 May 2026 18:51:44 GMT</pubDate>
            <atom:updated>2026-05-29T18:51:44.471Z</atom:updated>
            <content:encoded><![CDATA[<p><em>Transformers can be ordered. Charging stations can be built. Even new grids can be funded. But one part of the energy transition still cannot be purchased as quickly as governments would like.</em></p><p>France is accelerating electrification. <a href="https://www.reuters.com/business/energy/edf-invest-240-million-french-electrification-plan-2026-05-26/">Reuters reports that EDF is investing €240 million in a program to support heat pumps, public EV charging and energy-intensive industries</a>. The country also wants electricity to reach 60% of domestic energy use by 2030.</p><p>It sounds like an energy story.</p><p>But the more interesting part sits underneath.</p><p>The more countries talk about electrification, the less they talk about what electrification is made from. Every new charging station, every heat pump, every transformer and every kilometer of grid adds demand for materials that cannot be created at the push of a button.</p><p>Electricity moves fast.</p><p>New copper sources do not.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/821/1*WqUvUNly7ugybHfICPIGNw.png" /></figure><p>That is why the market is increasingly watching not only energy companies, but also future copper supply. While governments count megawatts, the mining sector counts years: geology, data analysis, drilling, permits and infrastructure.</p><p>NovaRed Mining is approaching this problem through MetalCore, an AI-assisted mineral evaluation platform. If future copper demand accelerates alongside electrification, faster understanding of geological data may become just as important as building new grids.</p><p>Maybe the biggest challenge for France’s electric future is not producing more electricity.</p><p>Maybe it is finding enough metal in time for a world that suddenly wants to electrify everything at once.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=abcf422f5496" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>