SVRGN Weekly Digest #49 💫
Events, diagnostics, hardhat GoPros, and of course, some DeepSeek hype
Next week you can catch me at the Nordic Defence Tech Summit where I’ll be moderating a panel with ICEYE, Palantir and SandboxAQ! If you’re in Stockholm and want to meet up, let me know.
The week after that, we’re hosting 300+ people for the European Defense Tech Hackathon at the Munich Security Conference (non-affiliated, just coinciding but good for you if you’re going). It’s a 4-day event for hackers, defense investors, companies, veterans and startups. We’re starting with workshops on everything from computer vision, data for defense, drone TTPs and additive manufacturing. Then there is 2 full days of hacking on challenges defined by top startups/scaleups in the space and the Ukrainian MoD.
On Sunday Feb 16th we’re hosting a massive demo day at the an awesome location in central Munich, open to externals. Sign up on Luma as a visitor!
🤼 People
Tor-Oskar Karlberg - Investor at Stephen industries
Stephen industries is a family-owned investment firm out of Helsinki, Finland. Tor-Oskar and I chatted about the impact investing can have, on local and global scale. We agree that as smaller funds, we need to spend time on things that others can’t or won’t. We also explored in-depth a potential co-invest. More coming.
💼 Portfolio jobs board
If you’re sad that I’m not listing any jobs here this week, send me a message:
🚀 Companies
Nine Diagnostics - Very fast diagnostics for cancer
Status: Pre-Seed
Source: CAVI
Founders: Daniel Heller, Freddy Nguyen
Why it’s cool:
Closing the “Live-Chain”, aka speeding up the feedback loops between identifying what is wrong and giving treatment is an important endeavor. Nine Diagnostics have some proprietary carbon nanotube sensing tech which allows them to create multidimensional profiles which can inform the treatment decision faster than existing analysis tools.
Ironsite - The future of workers
Status: Incubation
Source: Southpark Commons
Founders: Max Mona, Randy Perillo, Andrew Jebasingh
Why it’s cool:
Construction is fairly un-digitalized for various reasons (speculation: material to labor cost ratio, little competition, lack of new talent inflow, capital/political environment non-conducive to construction booms). Ironsite is putting GoPro’s on hardhats like it’s an extreme sport, and extracts insights in real-time on construction processes from the data.
Podium Automation - Custom control panels
Status: Pre-seed
Source: CAVI
Founders: Jacob Buser, Jamie Niu
Why it’s cool:
Can you imagine something less sexy than those grey cabinets every building has with wiring inside? It controls the power and connectivity of the space they’re in, be it industrial robots, HVAC systems, residential apartment buildings or a green house. If you know what UL 508A is, this is for you.
💡 Ideas & Science
So you want to build your own data center via Future of Computing newsletter
Data centers seem like a promising beta play. If you want to know how to build one, taught by someone who already did so, this is a nice post! There are too many posts on data center economics and physics (cooling is hard and cool) to mention here, but it’s worth getting deep into. Fun fact: racks are getting so heavy because of compute concentration that floors need to be re-thought.
Deepseek: The Quiet Giant Leading China’s AI Race
Chinatalk did a great interview with the Deepseek CEO and founder of one of the top quant hedge funds in the world, High-Flyer (幻方). They are outsiders, non-finance people, starting from first-principles only and tackling one hard problem after the other. They hire almost exclusively based on capability instead of experience. Hypothesis: high capability and capacity to learn transfers better than any experience. Even specific experience turns stale quickly in an unconventional org.
Taiwan's Future, Nuclear Proliferation, and Trump 2.0 with Lily Ottinger of ChinaTalk
Another ChinaTalk episode, this time a cross-over with Econ 102, Noah Smith and Eric Torenberg’s podcast. Why Xi Jinping is the ultimate Boomer (”Software isn’t real technology”) and why Japan and South Korea should get nuclear weapons immediately. It’s a great listen honestly. Many spicy takes.
This is not about “the end-THE END” but just about the end of satellites. Space debris is a problem as we know, it clutters the valuable orbits, and grows exponentially (breaking satellites travelling at 28,000 km/h into tiny pieces will break other satellites which hits others and so on). The worst is that once we have this band of bullets around the earth we can’t use space very well anymore. So we’d need to go to the stratosphere, or clean it up. This blog makes an argument that it’s not so bad, since stuff naturally falls out of the sky, even at altitudes beyond what we thought.