<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Bet The Decade: episodes</title><link>https://betthedecade.com/episodes/</link><atom:link href="https://betthedecade.com/feed.xml" rel="self" type="application/rss+xml"/><description>New Bet The Decade episodes with transcript and sources. Not financial advice.</description><language>en</language><lastBuildDate>Fri, 13 Nov 2026 01:29:09 -0800</lastBuildDate><item><title>Tempus AI Grew EBITDA 17-Fold. It Still Posted a GAAP Loss</title><link>https://betthedecade.com/episodes/tempus-ai-grew-ebitda-17-fold-it-still-posted-a-gaap-loss/</link><guid isPermaLink="true">https://betthedecade.com/episodes/tempus-ai-grew-ebitda-17-fold-it-still-posted-a-gaap-loss/</guid><pubDate>Fri, 13 Nov 2026 01:29:09 -0800</pubDate><description>T1 Energy nearly doubled quarterly net sales in a year: $250.1 million in Q2 2026, up from $132.8 million in Q2 2025, with 935 MW of modules produced in Dallas and a gross margin of about 20%. Yet the stock fell from $5.47 the day before that report to $2.29 on November 9. In this episode, recorded before the Q3 report on November 12, we look at what T1 already reported, the G2_Austin solar cell fab (2.1 GW Phase 1, first cells targeted for Q1 2027), the Section 232 polysilicon tariffs, the Clearway offtake, how exponential AI demand for power feeds U.S.-made solar, the strongest case against </description><category>TEM</category></item><item><title>SpaceX&#x27;s AI Plan Needs 1,000,000 Tons a Year. Earth Launched 2,877</title><link>https://betthedecade.com/episodes/spacex-s-ai-plan-needs-1-000-000-tons-a-year-earth-launched-2-877/</link><guid isPermaLink="true">https://betthedecade.com/episodes/spacex-s-ai-plan-needs-1-000-000-tons-a-year-earth-launched-2-877/</guid><pubDate>Thu, 12 Nov 2026 01:05:52 -0800</pubDate><description>Every rocket on Earth put 2,877 tons into orbit in 2025, most of it on SpaceX rockets. SpaceX&#x27;s own &quot;basic math&quot; for AI compute in orbit needs 1,000,000 tons a year, more than 300 times that. And the rocket meant to carry it, Starship, has reached orbit once. Recorded before Starship Flight 15 (no earlier than Nov 13, 2026). In this episode: the Falcon 9 record (about 119 launches in 2026 so far, $74M list price, about $3,250 per kg), why Falcon 9 can&#x27;t carry Starlink V3 in useful numbers, Raptor 3 and Starship V3, Flight 14 and Flight 15, the ~$100/kg Starship cost target, AI computers in orb</description><category>SPCX</category></item><item><title>What Is Ondas Really Worth Before Its Biggest Quarter Yet?</title><link>https://betthedecade.com/episodes/what-is-ondas-really-worth-before-its-biggest-quarter-yet/</link><guid isPermaLink="true">https://betthedecade.com/episodes/what-is-ondas-really-worth-before-its-biggest-quarter-yet/</guid><pubDate>Thu, 12 Nov 2026 00:19:13 -0800</pubDate><description>Ondas grew quarterly revenue more than 13-fold in a year: $83.8 million in Q2 2026, up 67% from Q1, with about 85% same-portfolio growth. Pro forma backlog jumped from $68 million at the end of 2025 to about $757 million on June 30, and the company guided Q3 revenue to $140–155 million. Yet the stock fell from $9.77 before that report to $5.97 on November 9. In this episode — recorded before the Q3 report on November 12 — we look at what Ondas already reported, the U.S. Army strike-drone program (a $982 million contract ceiling, more than $240 million ordered so far), the November 5 deal to bu</description><category>ONDS</category></item><item><title>What Is Grab Really Worth After Fintech Turned Profitable?</title><link>https://betthedecade.com/episodes/what-is-grab-really-worth-after-fintech-turned-profitable/</link><guid isPermaLink="true">https://betthedecade.com/episodes/what-is-grab-really-worth-after-fintech-turned-profitable/</guid><pubDate>Wed, 11 Nov 2026 06:21:36 -0800</pubDate><description>Grab just reported third-quarter 2026 results: revenue of $1.112 billion, up 27% year over year, and record adjusted EBITDA of $237 million, up 73%. Financial services turned adjusted-EBITDA-positive for the first time, at $12 million, as the loan book reached $2.7 billion. The company raised full-year guidance and bought back $400 million of shares in September and October. In this episode: what Grab reported, why the fintech milestone matters, what it is doing with the cash, how AI speeds this business up, the strongest case against, a 2036 vision (clearly labelled as opinion), and The Math:</description><category>GRAB</category></item><item><title>What Is IREN Really Worth After a $20 Billion Backlog?</title><link>https://betthedecade.com/episodes/what-is-iren-really-worth-after-a-20-billion-backlog/</link><guid isPermaLink="true">https://betthedecade.com/episodes/what-is-iren-really-worth-after-a-20-billion-backlog/</guid><pubDate>Wed, 11 Nov 2026 00:24:44 -0800</pubDate><description>IREN just reported fiscal first-quarter 2027 results: $228 million of revenue, up 66% from the prior quarter, with AI Cloud nearly tripling to $200 million. The revenue backlog rose 22% to $20.3 billion, about two-thirds with investment-grade counterparties. A target of more than $4 billion of annualized run-rate revenue is, the company says, fully contracted and fully funded — but the timing moved from year-end 2026 to January 2027, and maybe February, after a customer-requested network change. On the earnings call, the co-CEO said that customer was Microsoft. After the quarter, a customer pr</description><category>IREN</category></item><item><title>What Is CoreWeave Really Worth After Q3?</title><link>https://betthedecade.com/episodes/what-is-coreweave-really-worth-after-q3/</link><guid isPermaLink="true">https://betthedecade.com/episodes/what-is-coreweave-really-worth-after-q3/</guid><pubDate>Tue, 10 Nov 2026 14:29:23 -0800</pubDate><description>CoreWeave just reported a record third quarter: $3.66 billion of revenue, up 168% year over year, above its own $3.45–3.60 billion guidance. Signed backlog reached about $132.5 billion. And yet its bonds were recently priced at a yield above 16%. This episode asks what the company could be worth if that machine keeps working — and what happens if it does not.</description><category>CRWV</category></item><item><title>What Is Rocket Lab Really Worth After a $3.6B Backlog Quarter?</title><link>https://betthedecade.com/episodes/what-is-rocket-lab-really-worth-after-a-3-6b-backlog-quarter/</link><guid isPermaLink="true">https://betthedecade.com/episodes/what-is-rocket-lab-really-worth-after-a-3-6b-backlog-quarter/</guid><pubDate>Tue, 10 Nov 2026 09:00:01 -0800</pubDate><description>Rocket Lab just posted a monster third quarter: record revenue of $268M (+73% year on year), a record backlog of $3.57B (+52% in one quarter), positive GAAP operating cash flow, and 16 Neutron missions sold before the rocket has flown once, with the base price raised from $55M to $70M. In this episode: the Q3 2026 numbers, the backlog, why Neutron got more expensive before its first flight, how AI demand (including orbital data centers) is pulling on launch and satellites, the Q4 guidance, the risks, a 2036 vision (clearly labelled as a vision, not a fact), and what our portfolio model shows f</description><category>RKLB</category></item><item><title>What Is Hims &amp; Hers Really Worth After +49% Growth? (Q3 2026)</title><link>https://betthedecade.com/episodes/what-is-hims-hers-really-worth-after-49-growth-q3-2026/</link><guid isPermaLink="true">https://betthedecade.com/episodes/what-is-hims-hers-really-worth-after-49-growth-q3-2026/</guid><pubDate>Tue, 10 Nov 2026 08:00:39 -0800</pubDate><description>Hims &amp; Hers just grew revenue 49% in Q3 2026, the second quarter in a row of acceleration, yet the stock closed about 12% lower on report day.</description><category>HIMS</category></item><item><title>AST SpaceMobile Revenue Nearly Tripled. Why Build a Rocket Now?</title><link>https://betthedecade.com/episodes/ast-spacemobile-revenue-nearly-tripled-why-build-a-rocket-now/</link><guid isPermaLink="true">https://betthedecade.com/episodes/ast-spacemobile-revenue-nearly-tripled-why-build-a-rocket-now/</guid><pubDate>Tue, 10 Nov 2026 07:00:22 -0800</pubDate><description>The next day, Japan&#x27;s Rakuten joint venture added up to about $1B (¥148B) in non-dilutive funding.</description><category>ASTS</category></item><item><title>Why Your Phone Could Soon Talk Directly to Satellites</title><link>https://betthedecade.com/episodes/why-your-phone-could-soon-talk-directly-to-satellites/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-your-phone-could-soon-talk-directly-to-satellites/</guid><pubDate>Tue, 10 Nov 2026 02:45:42 -0800</pubDate><description>Your phone goes quiet when there is no cell tower nearby. What if the tower were in orbit and your ordinary phone could talk to it directly? That is the bet behind AST SpaceMobile (ASTS), a company from Midland, Texas, building some of the largest antennas ever flown in low Earth orbit. A full film: the history, how phone-to-satellite works, what could protect AST, what SpaceX changes, the latest numbers, the bear case, and my personal model.</description><category>ASTS</category><category>SPCX</category></item><item><title>What Is 10x Genomics Really Worth After Atera Ships?</title><link>https://betthedecade.com/episodes/what-is-10x-genomics-really-worth-after-atera-ships/</link><guid isPermaLink="true">https://betthedecade.com/episodes/what-is-10x-genomics-really-worth-after-atera-ships/</guid><pubDate>Tue, 10 Nov 2026 00:38:26 -0800</pubDate><description>10x Genomics expects to ship about 50 of its new Atera machines by year-end, and management says orders are not the bottleneck. Production is. Revenue grew just 1% in Q3 2026, yet the company&#x27;s market value roughly quadrupled this year.</description><category>TXG</category></item><item><title>What Is Butterfly Network Really Worth Now That AI Wants Its Chip?</title><link>https://betthedecade.com/episodes/what-is-butterfly-network-really-worth-now-that-ai-wants-its-chip/</link><guid isPermaLink="true">https://betthedecade.com/episodes/what-is-butterfly-network-really-worth-now-that-ai-wants-its-chip/</guid><pubDate>Mon, 09 Nov 2026 23:21:12 +0000</pubDate><description>Butterfly Network has lost about $918M since it was founded, much of it on one stubborn idea: put an entire ultrasound machine on a single silicon chip. For a probe business growing about 5%, that looked like a costly mistake. Now AI companies are paying for that chip. Midjourney built a body-scanner prototype around 40 Butterfly modules, an OpenAI-backed brain-interface startup signed a multi-year license, and partner revenue jumped more than 1,000% in Q3 2026. In this episode: why ultrasound on a chip matters, how AI turns a cheap probe into a useful one, why AI and neurotech companies want </description><category>BFLY</category></item><item><title>Why Can&#x27;t We Read Our Own Proteins Yet?</title><link>https://betthedecade.com/episodes/why-can-t-we-read-our-own-proteins-yet/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-can-t-we-read-our-own-proteins-yet/</guid><pubDate>Sun, 08 Nov 2026 08:00:10 +0000</pubDate><description>Your genome has about 20,000 protein-coding genes, but the proteins they make may come in millions of slightly different versions, called proteoforms. Most lab tools cannot read them one molecule at a time. Nautilus Biotechnology (NAUT) wants to. How its Iterative Mapping method works, what its Nature Methods tau paper proved, why its launch keeps slipping, the numbers, the bear case, and what I am watching. Timing note: Nautilus reported its Q3 2026 results on Tuesday, October 27, 2026, before the U.S. market open. The numbers in this video are from that report, the latest available at the pr</description><category>NAUT</category></item><item><title>How TeraWulf Turns Power Into AI Rent in 2026</title><link>https://betthedecade.com/episodes/how-terawulf-turns-power-into-ai-rent-in-2026/</link><guid isPermaLink="true">https://betthedecade.com/episodes/how-terawulf-turns-power-into-ai-rent-in-2026/</guid><pubDate>Sat, 07 Nov 2026 17:00:08 +0000</pubDate><description>My TeraWulf thesis: power becomes a rent-producing AI infrastructure platform — if construction and financing preserve the economics. Q3 2026 changes the revenue mix: approximately $63M of HPC lease revenue inside $76.2M total revenue. This developed episode follows the chain from secured power to contracted customers, project financing, delivered capacity and rent. We separate 102 MW online from 336 MW under construction at Lake Mariner, explain the Anthropic/Justified commitment and Muskie approval condition, and distinguish tenant revenue from warrant remeasurement and asset-sale gains.</description><category>WULF</category></item><item><title>What Is IonQ Really Worth After the Quantum Decade Starts?</title><link>https://betthedecade.com/episodes/what-is-ionq-really-worth-after-the-quantum-decade-starts/</link><guid isPermaLink="true">https://betthedecade.com/episodes/what-is-ionq-really-worth-after-the-quantum-decade-starts/</guid><pubDate>Sat, 07 Nov 2026 08:00:08 +0000</pubDate><description>IonQ raised $2.0B in October 2026 — before GAAP operating profit. Q3 revenue hit $173.8M (+336% YoY); FY2026 guide raised to $505–515M; RPO ~$1.2B; cash ~$3.7B pro forma after the raise. The company builds trapped-ion quantum machines and closed the ~$1.8B SkyWater foundry deal. This episode: where I see IonQ in a decade, why AI accelerates its scaling, the strongest case against, and The Math — bear / base / bull scenarios to 2035 (NFA). Distinct from our earlier IonQ videos (useful quantum mid-decade; manufacturable foundry).</description><category>IONQ</category></item><item><title>Why Does Tesla Sell Batteries to SpaceX? The Musk Web</title><link>https://betthedecade.com/episodes/why-does-tesla-sell-batteries-to-spacex-the-musk-web/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-does-tesla-sell-batteries-to-spacex-the-musk-web/</guid><pubDate>Fri, 06 Nov 2026 17:55:08 +0000</pubDate><description>In the first nine months of 2026, Tesla booked $626M of revenue from one customer buying its Megapack batteries: SpaceX. Tesla also owns a SpaceX stake valued at $2.87B, SpaceX now owns xAI (the company behind Grok), Grok runs in Tesla cars, and both companies plan a joint chip factory, Terafab. In its latest quarterly report, Tesla said its board formed a special committee of disinterested directors to explore deeper ties with SpaceX, possibly including a business combination. No terms have been published and no deal is assured. SpaceX has also formed its own special committee. How the compan</description><category>TSLA</category><category>SPCX</category></item><item><title>Why Hasn&#x27;t AI Delivered a Blockbuster Drug Yet?</title><link>https://betthedecade.com/episodes/why-hasn-t-ai-delivered-a-blockbuster-drug-yet/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-hasn-t-ai-delivered-a-blockbuster-drug-yet/</guid><pubDate>Fri, 06 Nov 2026 17:52:45 +0000</pubDate><description>Recursion runs BioHive-2, a supercomputer with 504 NVIDIA H100 chips that ranked No. 35 in the world at launch. Yet no AI-discovered drug has been approved by the FDA, from anyone. Why AI drug discovery is so hard to prove, what Recursion (RXRX) is testing in people, the numbers, the bear case, and what I am watching. Updated with Recursion&#x27;s Q3 2026 results (reported November 2, 2026) and the new REC-4881 data presented the same day.</description><category>RXRX</category></item><item><title>Why Is America Short of Solar Cells When AI Needs Power?</title><link>https://betthedecade.com/episodes/why-is-america-short-of-solar-cells-when-ai-needs-power/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-is-america-short-of-solar-cells-when-ai-needs-power/</guid><pubDate>Thu, 05 Nov 2026 17:00:00 +0000</pubDate><description>AI data centres could use more than twice as much electricity by 2030 (IEA), and solar is one of the fastest ways to add power. Yet America can assemble about 65 GW of solar panels a year and make only about 3 GW of the cells inside them. Why cells are the bottleneck, how the Section 232 tariffs change the math, T1 Energy (TE), the numbers, the bear case, and what I am watching. Next date to watch: T1 plans to report Q3 2026 results on Thursday, November 12, 2026, before the U.S. market opens.</description><category>TE</category></item><item><title>Why the AI Shovel Seller Looks Cheaper Than the AI Labs</title><link>https://betthedecade.com/episodes/why-the-ai-shovel-seller-looks-cheaper-than-the-ai-labs/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-the-ai-shovel-seller-looks-cheaper-than-the-ai-labs/</guid><pubDate>Thu, 05 Nov 2026 00:00:06 -0800</pubDate><description>Why does the AI shovel seller look cheaper than the AI labs buying its shovels? Funding rounds vs public prices, value vs revenue, simple payback math, why Nvidia may deserve a discount, what Cisco taught us in 2000, and the full bull case for the labs.</description><category>NVDA</category></item><item><title>Why Did the FDA Expand This Bladder Cancer Drug Early?</title><link>https://betthedecade.com/episodes/why-did-the-fda-expand-this-bladder-cancer-drug-early/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-did-the-fda-expand-this-bladder-cancer-drug-early/</guid><pubDate>Wed, 04 Nov 2026 09:00:09 -0800</pubDate><description>For some bladder cancer patients, the next option is surgery to remove the bladder. On October 14, 2026, the FDA expanded ImmunityBio&#x27;s Anktiva (with BCG) to BCG-unresponsive, papillary-only non-muscle-invasive bladder cancer, before its January 6, 2027 target date. How it works, what the published figures say, the road through the FDA, the numbers, the bear case, and my personal model.</description><category>IBRX</category></item><item><title>Why Unfinished Data Centers Look So Profitable on Paper</title><link>https://betthedecade.com/episodes/why-unfinished-data-centers-look-so-profitable-on-paper/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-unfinished-data-centers-look-so-profitable-on-paper/</guid><pubDate>Wed, 04 Nov 2026 00:00:07 -0800</pubDate><description>Amazon ended 2025 with about $72 billion of buildings and equipment still under construction, and none of it was being depreciated. Why can a half-built AI data center look so cheap on paper? Three ordinary, legal, disclosed accounting rules: construction in progress is not depreciated until it is ready for use, part of the interest is capitalized while building, and companies choose how many years their servers will last. Explained with public filings, the 1990s fiber boom, the case for long server lives, and four things to check in any filing.</description></item><item><title>Why Is Healthcare Turning Into a Subscription?</title><link>https://betthedecade.com/episodes/why-is-healthcare-turning-into-a-subscription/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-is-healthcare-turning-into-a-subscription/</guid><pubDate>Tue, 03 Nov 2026 09:00:08 -0800</pubDate><description>Millions of people now pay for their health care the way they pay for streaming: monthly. Hims &amp; Hers (HIMS), founded in San Francisco in 2017, turned online visits, licensed providers and medicine shipped to the door into a subscription. Then came the GLP-1 decade: compounded weight-loss copies, a fight with Novo Nordisk, FDA pushback, and a surprise partnership. A full film: the history, how the model works, the GLP-1 saga, the latest numbers, the bear case, and my personal model.</description><category>HIMS</category></item><item><title>Medicine Made in Space: The 2035 Vision</title><link>https://betthedecade.com/episodes/medicine-made-in-space-the-2035-vision/</link><guid isPermaLink="true">https://betthedecade.com/episodes/medicine-made-in-space-the-2035-vision/</guid><pubDate>Tue, 03 Nov 2026 01:38:42 -0800</pubDate><description>It&#x27;s 2035. The space station is retired, Starships fly often, and some carry capsules of drug experiments that return weeks later. Not a prediction: it&#x27;s what several companies say they&#x27;re building toward, and SpaceX wants to be the road. What orbit already taught drugmakers, why 2030 is a turning point, who is booked on Starfall, which medicines could benefit, how big this could be for SpaceX, the risks, and three 2035 scenarios. Most important first: today, no approved drug is made in space, and SpaceX does not make drugs or run drug factories. It builds the rockets and the Starfall return c</description><category>SPCX</category></item><item><title>Everyone Wants to Escape Nvidia. They All Use the Same Factory.</title><link>https://betthedecade.com/episodes/everyone-wants-to-escape-nvidia-they-all-use-the-same-factory/</link><guid isPermaLink="true">https://betthedecade.com/episodes/everyone-wants-to-escape-nvidia-they-all-use-the-same-factory/</guid><pubDate>Tue, 03 Nov 2026 00:00:06 -0800</pubDate><description>Amazon says its own chips business (Trainium, Graviton) is running at more than $25B a year (Q2 2026 run rate). Google has its TPUs, Nvidia sells graphics processors, and Cerebras built a chip the size of a whole wafer. Four rivals, one factory: their flagship AI chips (Nvidia Blackwell, Google Ironwood, Amazon Trainium3, the Cerebras WSE-3) are all made by TSMC. Why do the biggest tech companies design their own AI chips, who actually builds them, what TSMC gets out of the chip war, what the PC clone era (IBM PC, Compaq, &quot;Intel Inside&quot;) can teach us, the strongest case against, and my persona</description><category>TSM</category><category>CBRS</category></item><item><title>How Neuralink Wires a Computer to a Brain</title><link>https://betthedecade.com/episodes/how-neuralink-wires-a-computer-to-a-brain/</link><guid isPermaLink="true">https://betthedecade.com/episodes/how-neuralink-wires-a-computer-to-a-brain/</guid><pubDate>Sun, 01 Nov 2026 09:00:10 -0800</pubDate><description>Seventy-five words per minute: Neuralink says one person with ALS turned brain signals into text that fast in their first week with the implant (company-reported, not peer-reviewed). So how do you actually wire a computer to a human brain? Inside the N1 implant, why a robot does the surgery, what participants can do today, where the FDA stands, the company&#x27;s 2029-2031 plan, how investors get exposure to a private company, and the bear case.</description></item><item><title>How an App Became a Bank for 17 Million People</title><link>https://betthedecade.com/episodes/how-an-app-became-a-bank-for-17-million-people/</link><guid isPermaLink="true">https://betthedecade.com/episodes/how-an-app-became-a-bank-for-17-million-people/</guid><pubDate>Sun, 01 Nov 2026 01:00:09 -0700</pubDate><description>In 2018 SoFi was a single-product student loan refinancing business. Today it is a nationally chartered bank with 17.1 million members, $51.3B in deposits and 26.7 million products, all inside one app. How did a Stanford student-loan startup become a bank, and is it a tech company or a lender at heart? A full film: the history, the SoFi loop, the bank charter, lending, the latest Q3 2026 numbers, the bear case, and my thesis.</description><category>SOFI</category></item><item><title>Towns Stalled $198B of Big Tech&#x27;s AI Data Centers</title><link>https://betthedecade.com/episodes/towns-stalled-198b-of-big-tech-s-ai-data-centers/</link><guid isPermaLink="true">https://betthedecade.com/episodes/towns-stalled-198b-of-big-tech-s-ai-data-centers/</guid><pubDate>Sat, 31 Oct 2026 10:00:11 -0700</pubDate><description>Local opposition blocked or delayed about $68B of US data center projects in Q2 2026 alone, and almost $200B in the first half of 2026, according to Data Center Watch. Then Texas paused data center permits. Here&#x27;s the bet: the backlash doesn&#x27;t kill the demand for power, it changes who gets paid for it. We cover why towns say no, ERCOT and PJM, transformers, who pays the bill, who gets paid behind the fence, and a bull and bear case to 2035.</description></item><item><title>Why the $1.8 Trillion Space Economy Needs a Second SpaceX</title><link>https://betthedecade.com/episodes/why-the-1-8-trillion-space-economy-needs-a-second-spacex/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-the-1-8-trillion-space-economy-needs-a-second-spacex/</guid><pubDate>Sat, 31 Oct 2026 03:41:23 -0700</pubDate><description>Rocket Lab has launched its Electron rocket 98 times, its big reusable Neutron rocket just slipped to 2027, and it is buying Iridium. Is this the company that becomes the &quot;second SpaceX&quot; of a space economy that could reach ~$1.8 trillion by 2035? A full deep-dive: Electron, Neutron, Space Systems, the $2.36B backlog, the bear case, and my thesis.</description><category>RKLB</category></item><item><title>Why Washington Is Betting Up to $5B on Old Nuclear Reactors</title><link>https://betthedecade.com/episodes/why-washington-is-betting-up-to-5b-on-old-nuclear-reactors/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-washington-is-betting-up-to-5b-on-old-nuclear-reactors/</guid><pubDate>Sat, 31 Oct 2026 01:00:11 -0700</pubDate><description>Washington just offered America&#x27;s largest nuclear operator up to $5 billion, and most of the extra power already has a buyer: Google, Amazon or Meta. On Oct 27, 2026 the US Department of Energy announced a CONDITIONAL loan commitment of up to $5B to Constellation for nuclear &quot;uprates&quot; at 8 existing plants, targeting nearly 1.1 GW of new capacity. What did Constellation actually get, what is an uprate, how does Big Tech fit in, and what could break the bet? Bull case to 2035, plus the other side of the trade.</description><category>CEG</category></item><item><title>Nvidia&#x27;s Real Bottleneck Isn&#x27;t the Chip</title><link>https://betthedecade.com/episodes/nvidia-s-real-bottleneck-isn-t-the-chip/</link><guid isPermaLink="true">https://betthedecade.com/episodes/nvidia-s-real-bottleneck-isn-t-the-chip/</guid><pubDate>Fri, 30 Oct 2026 10:00:08 -0700</pubDate><description>About 90% of the silicon in a top AI chip package is memory, not processor. NVIDIA&#x27;s Vera Rubin GPU carries 288 GB of HBM4, and SK hynix says its DRAM and HBM volumes are effectively sold out through 2029. Here&#x27;s the bet: in the AI decade, memory sets the pace. We look at how HBM is built (TSVs, base die, CoWoS), who makes it (SK hynix, Samsung, Micron), who pays, what comes next (HBM4E, HBF), and an honest bear case to 2035.</description><category>NVDA</category></item><item><title>Terafab: Why Musk Needs His Own Chip Factory</title><link>https://betthedecade.com/episodes/terafab-why-musk-needs-his-own-chip-factory/</link><guid isPermaLink="true">https://betthedecade.com/episodes/terafab-why-musk-needs-his-own-chip-factory/</guid><pubDate>Thu, 29 Oct 2026 12:52:40 -0700</pubDate><description>Terafab is Elon Musk&#x27;s bet that chip supply, not chip design, becomes the bottleneck of the AI decade. Tesla, SpaceX and Intel plan to build a chip fab in Texas: a ~$3B research fab in Austin, then a main campus at Gibbons Creek with filings that add up to about $119B by the mid-2030s. In this video: what Terafab is, how a fab and EUV lithography actually work, the plan and the numbers, TSMC/Samsung/Intel, and an honest bull and bear case. This is analysis, not financial advice.</description><category>TSLA</category><category>SPCX</category></item><item><title>Why Catching Rockets Changes Everything</title><link>https://betthedecade.com/episodes/why-catching-rockets-changes-everything/</link><guid isPermaLink="true">https://betthedecade.com/episodes/why-catching-rockets-changes-everything/</guid><pubDate>Thu, 29 Oct 2026 12:00:38 -0700</pubDate><description>Next month SpaceX wants to catch BOTH halves of Starship with the launch tower for the first time. Here&#x27;s how a rocket catch works, the tech behind it (Mechazilla&#x27;s chopsticks, the tile heat shield, Raptor), the reuse economics, the bear case, and why it matters for the decade.</description><category>SPCX</category></item><item><title>Who Survives the Neocloud Shakeout? The CoreWeave Bet</title><link>https://betthedecade.com/episodes/who-survives-the-neocloud-shakeout-the-coreweave-bet/</link><guid isPermaLink="true">https://betthedecade.com/episodes/who-survives-the-neocloud-shakeout-the-coreweave-bet/</guid><pubDate>Thu, 29 Oct 2026 06:48:23 -0700</pubDate><description>In October, CoreWeave, Nebius and IREN each fell more than 30%. Elon Musk says most of the 200+ neoclouds won&#x27;t make it; Gartner expects ~80% to be absorbed or shut within several years. So who survives the neocloud shakeout? This video tests one thesis: the survivors will have signed backlog, secured power, on-time delivery — and capital they can afford. CoreWeave (NASDAQ: CRWV) is the biggest test case: ~$104B backlog, 1.5 GW active power and a goal of at least 8 GW by 2030… against ~$51B of debt-like claims, 15%+ bond yields and credit default swaps above distress levels. Plus this week: De</description><category>CRWV</category></item><item><title>Quantum, Built Like a Chip: What Makes IonQ Different</title><link>https://betthedecade.com/episodes/quantum-built-like-a-chip-what-makes-ionq-different/</link><guid isPermaLink="true">https://betthedecade.com/episodes/quantum-built-like-a-chip-what-makes-ionq-different/</guid><pubDate>Thu, 29 Oct 2026 00:11:42 -0700</pubDate><description>Quantum only becomes a decade technology if it can be built like chips. This Bet The Decade episode breaks down IonQ’s manufacturable-quantum bet: trapped ions, electronic qubit control, the SkyWater foundry close, Superion 256’s 2027 delivery window, DARPA Quantum Benchmarking Initiative Stage C, and how Quantinuum’s Helios sets a peer bar — with a short, data-backed bear case. 📌 Chapters 0:00 Built like a chip 0:58 What IonQ sells 2:45 Scale levers · Stage C 4:56 Helios peer bar 5:56 The bear case 7:44 2030 and 2035 scenarios 8:50 Verdict</description><category>IONQ</category></item><item><title>TeraWulf Isn&#x27;t a Neocloud — It&#x27;s the AI Power Landlord (WULF Thesis)</title><link>https://betthedecade.com/episodes/terawulf-isn-t-a-neocloud-it-s-the-ai-power-landlord-wulf-thesis/</link><guid isPermaLink="true">https://betthedecade.com/episodes/terawulf-isn-t-a-neocloud-it-s-the-ai-power-landlord-wulf-thesis/</guid><pubDate>Thu, 29 Oct 2026 00:00:00 +0000</pubDate><description>Is TeraWulf (WULF) a neocloud — or the landlord neoclouds rent from? After October&#x27;s AI-infrastructure sell-off and Elon Musk&#x27;s &quot;guppy-level death&quot; warning, we break down how TeraWulf&#x27;s power-landlord model works, what the Q2 2026 numbers say, the Kentucky pipeline with Anthropic, and a six-point bear case with real numbers.</description><category>WULF</category></item><item><title>Southeast Asia&#x27;s Superapp Is Building a Fintech Decade</title><link>https://betthedecade.com/episodes/southeast-asia-s-superapp-is-building-a-fintech-decade/</link><guid isPermaLink="true">https://betthedecade.com/episodes/southeast-asia-s-superapp-is-building-a-fintech-decade/</guid><pubDate>Wed, 28 Oct 2026 13:39:37 -0700</pubDate><description>Southeast Asia&#x27;s superapp is building a fintech decade. Grab (NASDAQ: GRAB) — rides, deliveries, ads, lending, and digibanks across eight countries. On Oct 27, 2026 the stock sat near $2.70–$2.80 (~$11B market value) after a brutal drawdown from its ~$6.24 52-week high. Meanwhile Q2 2026 revenue hit $997M (+22%), on-demand GMV $6.5B, MTUs 53.9M, and adjusted EBITDA $168M (+54%). Management raised FY2026 guidance to $4.10–$4.15B revenue and $720–$740M adjusted EBITDA, with ~$7.4B gross cash liquidity and a new $750M buyback authorization. This episode covers the superapp flywheel, why fintech m</description><category>GRAB</category></item><item><title>AI That Keeps Working When You Close the Laptop — Grok Bot (xAI / Cursor)</title><link>https://betthedecade.com/episodes/ai-that-keeps-working-when-you-close-the-laptop-grok-bot-xai-cursor/</link><guid isPermaLink="true">https://betthedecade.com/episodes/ai-that-keeps-working-when-you-close-the-laptop-grok-bot-xai-cursor/</guid><pubDate>Wed, 28 Oct 2026 13:33:48 -0700</pubDate><description>Decade vision: always-on AI teammates become the default way knowledge work gets done. This episode explains Grok Bot from public docs only — persistent cloud computers, multi-Bot teams, skills/routines, paid access (Cursor / SuperGrok / The X Subscription Premium+), weekly usage limits, and the bull vs bear case for agent platforms to 2030–2035.</description></item><item><title>This Tiny Wireless Stock Became a Defense Drone Machine</title><link>https://betthedecade.com/episodes/this-tiny-wireless-stock-became-a-defense-drone-machine/</link><guid isPermaLink="true">https://betthedecade.com/episodes/this-tiny-wireless-stock-became-a-defense-drone-machine/</guid><pubDate>Wed, 28 Oct 2026 13:11:34 -0700</pubDate><description>Ondas Inc. (NASDAQ: ONDS) — from private rail wireless and industrial drones to a multi-domain autonomous defense roll-up. Bull case, bear case, valuation, and decade scenarios. Timing note (Oct 27, 2026 Europe/Warsaw): next earnings Nov 12, 2026; DZYNE/Cyberhawk integration still early.</description><category>ONDS</category></item><item><title>Hydro AI Campuses vs Gas Peaker Neoclouds: IREN</title><link>https://betthedecade.com/episodes/hydro-ai-campuses-vs-gas-peaker-neoclouds-iren/</link><guid isPermaLink="true">https://betthedecade.com/episodes/hydro-ai-campuses-vs-gas-peaker-neoclouds-iren/</guid><pubDate>Wed, 28 Oct 2026 02:49:21 -0700</pubDate><description>Some scenes are AI illustrations. Wikimedia Commons images credited below / in credits.md.</description><category>IREN</category></item><item><title>GreenX vs Poland: The Arbitration Story</title><link>https://betthedecade.com/episodes/greenx-vs-poland-the-arbitration-story/</link><guid isPermaLink="true">https://betthedecade.com/episodes/greenx-vs-poland-the-arbitration-story/</guid><pubDate>Tue, 27 Oct 2026 14:22:37 -0700</pubDate><description>An international tribunal ordered Poland to pay GreenX Metals about £252 million — more than the company&#x27;s entire stock-market value of roughly A$290 million (~£154 million) on October 27, 2026. Singapore already rejected Poland&#x27;s first challenge; the Court of Appeal decision could land any week. Meanwhile GreenX is about to drill copper at Tannenberg in Germany (~1,900 km², 90% owned) with an Exploration Target of 1.3–3.9 Mt copper.</description><category>GRX</category></item><item><title>This Cancer Lab Is Becoming an AI Giant</title><link>https://betthedecade.com/episodes/this-cancer-lab-is-becoming-an-ai-giant/</link><guid isPermaLink="true">https://betthedecade.com/episodes/this-cancer-lab-is-becoming-an-ai-giant/</guid><pubDate>Tue, 27 Oct 2026 09:31:21 -0700</pubDate><description>Tempus AI (NASDAQ: TEM) — a cancer diagnostics flywheel becoming a multimodal AI data platform. Revenue guide ~$1.6B for 2026, Personalis MRD deal, valuation, and bull/bear scenarios to 2030/2035. Timing note (Oct 27, 2026): Tempus Q3 earnings Nov 2, 2026 (before open); Personalis shareholder vote Nov 30, 2026 (deal not closed).</description><category>TEM</category></item><item><title>AI&#x27;s Power Bill: Who Pays, and Who Gets Paid?</title><link>https://betthedecade.com/episodes/ai-s-power-bill-who-pays-and-who-gets-paid/</link><guid isPermaLink="true">https://betthedecade.com/episodes/ai-s-power-bill-who-pays-and-who-gets-paid/</guid><pubDate>Mon, 26 Oct 2026 00:11:41 -0700</pubDate><description>AI data centers need enormous amounts of power. In PJM, America&#x27;s largest grid, data centers drove $6.3B of the $16.4B cost of the latest capacity auction, according to the grid&#x27;s market monitor, and those costs land on electricity bills in 13 states + D.C. Who gets paid? Constellation Energy (CEG), the largest US nuclear operator, now has long-term deals with Microsoft (Crane restart), Meta (Clinton), Amazon (Calvert Cliffs) and Google (~3.6 GW, Oct 6, 2026). Yet the stock closed ~$253 on Oct 23, 2026 (market value ~$90B), down more than a quarter in a year. We cover the contracts, Q2 2026 re</description><category>CEG</category></item><item><title>Tesla: Why Wall Street Can&#x27;t Agree</title><link>https://betthedecade.com/episodes/tesla-why-wall-street-can-t-agree/</link><guid isPermaLink="true">https://betthedecade.com/episodes/tesla-why-wall-street-can-t-agree/</guid><pubDate>Sun, 25 Oct 2026 07:11:49 -0700</pubDate><description>RBC says $480. Same company, same Q3 2026 report. Tesla closed around $372 on Oct 23, 2026 (market value ~$1.47 trillion), while Q3 revenue was flat at $28.05B and GAAP net income fell 43% to $787M. In this Tesla stock analysis we look forward: robotaxi and Full Self-Driving (1.74M paid customers), Optimus and the Terafab chip plan, the energy storage business (Megapacks almost sold out for 2027, per Musk), the SpaceX special committees, the bear case (valuation, flat car business, rising costs) and my illustrative 2030 and 2035 scenarios.</description><category>TSLA</category></item><item><title>CRISPR Stock Is Half Cash. Is Wall Street Ignoring the Pipeline?</title><link>https://betthedecade.com/episodes/crispr-stock-is-half-cash-is-wall-street-ignoring-the-pipeline/</link><guid isPermaLink="true">https://betthedecade.com/episodes/crispr-stock-is-half-cash-is-wall-street-ignoring-the-pipeline/</guid><pubDate>Sat, 24 Oct 2026 02:52:32 -0700</pubDate><description>CRISPR Therapeutics (CRSP) is worth about $4.8 billion, and about $2.4 billion of that is cash and securities. The rest buys Casgevy, the first approved CRISPR therapy (Q2 sales +151% year over year), and a pipeline led by CTX310, a one-time gene-editing infusion that cut LDL cholesterol about 53% on average after one year in a small Phase 1 trial. In this CRISPR stock analysis we look forward: the Casgevy ramp, CTX310 and the in vivo pipeline (CTX340, CTX460), the CTX611 Factor XI readout due by year end, zugo-cel in autoimmune disease, the balance sheet, the bear case (collaboration losses, </description><category>CRSP</category></item><item><title>SpaceX Stock at $2 TRILLION: Too Late, or Just Getting Started?</title><link>https://betthedecade.com/episodes/spacex-stock-at-2-trillion-too-late-or-just-getting-started/</link><guid isPermaLink="true">https://betthedecade.com/episodes/spacex-stock-at-2-trillion-too-late-or-just-getting-started/</guid><pubDate>Fri, 23 Oct 2026 06:39:16 -0700</pubDate><description>Last quarter 52% of SpaceX&#x27;s revenue came from AI. Revenue grew 161% to $14.6 billion, yet SpaceX stock (SPCX) fell 9% the next day and trades around $145, a market value of roughly $2 trillion. In this SpaceX stock analysis we look forward: Starship Flight 15 and the cost to orbit, Starlink subscribers and Starlink Mobile, the xAI / Grok compute build-out (1.7 GW to 13 GW?), the $63B backlog, Q1 2027 guidance, the bear case (capex, lockups, control) and my 2030 and 2035 scenarios.</description><category>SPCX</category></item><item><title>Cerebras Stock: The $25 Billion Bet Against Nvidia&#x27;s Moat</title><link>https://betthedecade.com/episodes/cerebras-stock-the-25-billion-bet-against-nvidia-s-moat/</link><guid isPermaLink="true">https://betthedecade.com/episodes/cerebras-stock-the-25-billion-bet-against-nvidia-s-moat/</guid><pubDate>Thu, 22 Oct 2026 07:00:59 -0700</pubDate><description>Cerebras stock (CBRS) has fallen from a $311 first-day close to about $118, yet the company has $25.4 billion of contracted future revenue. In this Cerebras stock analysis we look forward: the OpenAI 750-megawatt contract, AWS Bedrock, independent inference speed benchmarks, the AI inference market forecast, 2027 targets, and the 2030 and 2035 bull and bear case. Can anyone break Nvidia&#x27;s CUDA moat?</description><category>CBRS</category><category>NVDA</category></item><item><title>How Nvidia Won AI: The 20-Year Bet Wall Street Ignored</title><link>https://betthedecade.com/episodes/how-nvidia-won-ai-the-20-year-bet-wall-street-ignored/</link><guid isPermaLink="true">https://betthedecade.com/episodes/how-nvidia-won-ai-the-20-year-bet-wall-street-ignored/</guid><pubDate>Thu, 22 Oct 2026 06:28:27 -0700</pubDate><description>In the mid-1990s Nvidia was weeks from running out of money. Today it is worth more than $5 trillion. This isn&#x27;t a fan story. It&#x27;s an investor&#x27;s breakdown of the moat, the key bets, the valuation, and the bull and bear case for 2035.</description><category>NVDA</category></item></channel></rss>
