š§² Magnets, Metals, and Moats Reshore; š°ļø Propellantless Thruster š Robotic Harvesters Near Market; š Volcano Homes Tap Geothermal & more | Deep Tech Briefing n.88
Weekly Intelligence on Deep Tech Private Markets.
Welcome back to Deep Tech Briefing ā the weekly space by The Scenarionist where we analyze and discuss the key events of the week shaping Deep Tech Private Markets.
āØIn Todayās Briefing:
Editorial
Interesting Reading
The Big Idea:
The Blue-Collar Bot Wars. Figure beats Musk to the finish line. Now comes the hard part: proving the unit economics really work.
Deep Tech Key Moves:
Quantum rails, hydrogen solids, fusion pulses, propellantless thrustersāand the industrial spine behind deep tech
A week when national power, compute demand, and private deep tech all pulled in the same direction
Signal in the Data:
AIās Appetite for Power Meets the Limits of the Grid
Data-center capacity charts double as an AI revenue stress test
Deep Tech Power Play:
From Wellington to Brussels to Washington ā IP reform unlocking university spin-outs, a water-innovation engine for drought, flood and blue-economy tech, an AI competitiveness push, and funding for critical-mineral security.
Greetings!
Late November 2025 is turning into a stress test for what āreindustrializationā actually means in practice. This weekās deep tech signals point less to grand visions and more to the unglamorous mechanics of keeping factories staffed, data centers powered, and supply chains sovereign.
On the factory floor, humanoid robots have quietly crossed a psychological line. One general-purpose machine has just finished almost a year of shift work on a mainstream auto production line, clocking real hours, following real takt times, and collecting real scuff marks. That is a very different proof point from a glossy demo. It suggests humanoids are starting to live inside industrial rhythms rather than performing one-off stunts. At the same time, more conventional autonomy in mobility is running into hard edges: lawsuits, customer disputes, and liquidity pressure are forcing a rethink of what āfull self-drivingā economics actually look like.
In parallel, AIās appetite for power is colliding with the limits of the grid. U.S. data-center pipelines now show a visible layer of projects effectively stuckāwaiting on interconnection studies, transformers, and long-dated power contracts. On paper, capacity looks abundant; in reality, the constraint is firm megawatts in the right zip codes. That turns certain substations, transmission corridors, and behind-the-meter generation projects into de facto toll booths on the AI economy. The more AI training and inference become dependent on scarce grid upgrades, the more returns will concentrate in assets that secured power early and can prove it.
Energy and materials are responding in kind. Fusion experiments are pushing into regimes that used to belong only to national labs, a new generation of fission machines is edging from design to criticality, and hydrogen is being repackaged into forms that are easier to ship, store, or synthesize into drop-in fuels. Battery chemistries are bifurcating: safer, abundant ingredients for grid and maritime storage on one side; higher-performance blends for drones, defense, and AI-heavy sites on the other. Underneath that, rare-earth magnets, critical minerals, and specialty processors are being treated less like commodities and more like long-lived infrastructure.
Policy is becoming the connective tissue. A new European platform for water innovation, a fresh push on AI competitiveness, sizable U.S. funding for critical materials, and revamped IP rules in a smaller tech-forward nation all point in the same direction: governments are not trying to pick individual startups, but they are clearly tilting the field toward certain stacksāwater systems, secure compute, domestic extraction, university spin-outs.
Soā¦the frontier is no longer defined by a single breakthrough in isolation, but by the ability to line up labor, robots, power, minerals, and intellectual property into durable industrial systems. The rest of this weekās briefing digs into where that alignment is already visibleāand where the gaps still look mispriced.
Enjoy the read!
- Giulia
āØ For more, see Membership | VC Guides | Insights | Rumors | Exit
šø Interesting Reading:
Why the Time Has Finally Come for Geothermal Energy The New Yorker ā A long-view narrative on how policy, drilling tech and culture finally line up to make geothermal feel inevitable.
Inside the Deep-Tech Revolution: Why Private Chipmakers Hold the Key BISinfotech ā Why control of niche silicon may decide who actually captures value in AI, sensing and autonomy.
Customer Adoption: The Silent Killer of Deeptech Startups Financial Express ā India as a case study in how procurement, integration risk, and slow sales cycles cap deeptech growth.
Europeās ā¬398B Spinout Machine Startups Magazine ā Useful map of which campuses, tech verticals, and exit routes are actually working across the continent.
From plough to platform: The WEFās deep-tech play for Indian agriculture ET Edge Insights ā How WEF wants to turn Indiaās farms into a deep-tech platformāfrom CRISPR to farm data rails and marketplaces.
US Green Hydrogen Startups Are Moving On To Greener Pastures CleanTechnica ā Green hydrogen founders are voting with their feetāuseful signal on which markets are truly open for business.
The Australian Lesson Focusing Startup Investments On A Few Strategic Sectors Crunchbase News ā Australiaās choice to back a handful of deeptech wedges instead of āeverything, everywhereā venture.
Whoās Still Betting on Transport? Crunchbase News ā Breaks down where capital has pulled back in mobility and where patient money is still stepping in.
How Generative AI Is Reshaping Venture Capital Harvard Business Review ā GenAI as firm infrastructureāsourcing, screening and portfolio supportārather than yet another āVC killerā headline.
āOur funds are 20 years oldā: Limited partners confront VCsā liquidity crisis TechCrunch ā Inside the LPāGP standoff around 18ā20 year fund livesāand what that means for illiquid deeptech bets.