by Nihal Kurth
Kitchen-table conversations with builders at the deep-tech AI frontier. No script. No filters. The reasoning behind the moves: what's working, what's not and why conviction beats consensus. For those building what doesn’t have a manual. Deep-dives → https://deeptechdecoded.substack.com/
Language
🇺🇲
Publishing Since
10/3/2025
Pitch yourself as a guest, propose sponsorships, or reach out directly to the host.
May 21, 2026
<p>What if AI's binding constraint isn't smarter models or faster chips - but the megawatts sitting unused inside regulated utilities in every major city?Graeme Harrison, General Partner at Augur VC and Founder & CEO of Simply Silicon, has spent years betting that energy, not compute, is the real bottleneck for AI. Augur was incubating against an "energy AI" thesis before the term "NeoCloud" existed. Simply Silicon, the operating company that emerged from Augur's dealflow, is now building the world's first air‑gapped inference networks for major North American cities - a category of one, for now.Their first site is a regulated district heating plant in downtown Calgary with 12 MW of unused electrical capacity - the kind of deal Silicon Valley money alone could never close. On the fund side, Augur has been first into companies like Beyond Reach Labs (cold‑emailed pre‑demo‑day for an uncapped SAFE), Aero Volta, and Cumulus, underwriting outliers against a worldview Graeme calls "technological determinism."Graeme spent nearly a decade doing technology M&A at Bennett Jones in Canada before launching Augur. He built his first company - Quest Websites - in mid‑90s Saskatchewan as a kid, got wiped out years later on an electronic cigarettes import business when a regulatory change seized his inventory, and went to law school to learn the commercial backbone of how the world actually runs. In 1999 he read Ray Kurzweil's The Age of Spiritual Machines, and that book has quietly directed every choice he's made since.</p><p>In this episode, we explore:- Energy as the Binding Constraint: Why the AI infrastructure race is going to be won by whoever solves power, not compute.- The Cold Email to Beyond Reach Labs: How Augur wrote an uncapped SAFE before demo day and why the founders said yes.- Technological Determinism: Augur's underwriting model - civilizations unlock technologies in a fixed sequence, and the next ones are knowable.- "The Internet Wasn't Built for This": Why current data center architecture can't deliver what regulated enterprises need from AI inference.- The Calgary District Heating Deal: How Simply Silicon co‑locates inside a regulated utility - and why Silicon Valley capital alone can't open that door.- The Death Spiral of Venture Capital: Why generalist VCs reject the deals worth doing, and what conviction‑led underwriting looks like instead.- Don't Die: The Y Combinator rule Augur leans on through portfolio inflection points - and the "teardrop" founder who said no to an early acquisition.- Plus a cameo from Mikey: hired into Augur at 18, now running investor tours at Simply Silicon at 20.Learn more about Graeme:https://deeptech-decoded.com/guests/graeme-harrisonTimestamps:0:00 - Designing the conversation: identity, two roles, Beyond Reach1:05 - Ten years of tech M&A at Bennett Jones2:03 - The electronic cigarette business that wiped him out4:59 - From Quest Websites to law school: learning how the world runs7:16 - Wearing two hats: Augur VC and Simply Silicon9:27 - Ray Kurzweil and the energy‑AI North Star10:43 - Dan Chapman and building the Augur team13:43 - Technological determinism and the Kardashev scale17:23 - The Beyond Reach Labs investment story22:11 - The cold email that became an uncapped SAFE24:35 - $325M in LOIs by demo day25:17 - What was actually on the investment memo27:47 - Investing vs. building: which would Graeme pick?29:44 - Simply Silicon emerges from Augur's dealflow31:31 - Why the internet wasn't built for AI inference36:00 - Recruiting an inference engineer out of AMD40:09 - Inside the Calgary district heating plant deal44:52 - What makes founders pick Augur over bigger checks47:30 - The "teardrop" founder who said no to an acquisition50:49 - Don't die: Y Combinator's most important rule57:00 - Token‑maxing, patent‑maxing, and broken VC incentives1:02:45 - Spotting an 18‑year‑old outlier1:13:00 - The book that gave Graeme his North Star</p>
May 7, 2026
<p>What if the biggest structures humans ever build aren't on Earth - but folded into a backpack and unfolded 50x larger in orbit?Pele Collins, co-founder & CTO of Beyond Reach Labs (YC W26), is on a mission to become the "big builder" of space. Their flat-pack origami mechanism delivers solar arrays up to 100 meters long, 88% cheaper than today's ISS-grade solutions — and it gets stiffer as it gets bigger, solving the floppy-array problem that limits satellite movement. Pele spent five years at SpaceX as the responsible engineer for the Dragon parachutes, led plasma-facing components at Commonwealth Fusion, returned to SpaceX to set up parachute manufacturing, and in October 2025 joined his Penn classmate Mitchell Fogelson — fresh from a CMU PhD on kilometer-scale NASA structures — to build Beyond Reach Labs.In this episode, we explore:- The 88% Cheaper Mechanism: A stamped-metal, flat-pack design that beats a $100M composite roll-up tube.- From Dragon Parachutes to Space Origami: Pele's path from SpaceX recovery hardware to deployable space structures.- "The Best Deployable Is No Deployable": Why simple wins in orbit — and why James Webb spooked the industry.- Bigger = Stiffer: The natural-frequency trick that flips the physics of large solar arrays on its head.- 2001: A Space Odyssey, But Real: The kilometer-scale, centripetal-gravity station behind the whole mission.- Vomit Comets and Tracking Dots: How you actually prove a deployable will work before sending it to orbit.- The Pivot That Saved the Company: From selling a simulation tool to designing and building the hardware itself.Learn more about Pele:https://deeptech-decoded.com/guests/pele-collinsTimestamps:0:00 Why Beyond Reach is more than what it looks like - meet Pele Collins1:25 South Africa → Penn → SpaceX: Pele's human-being introduction3:50 Legos, robotics, and the path to mechanical engineering5:07 Mr. Thompson, Mr. Bradley, and the physics teachers who set off a career8:00 Sports, soccer, and being named after Pelé9:30 Immigrant mentality - secure the basics, then take the swing12:18 Mitch Fogelson, NASA kilometer-scale structures, and founding Beyond Reach13:30 Y Combinator - and the treadmill since14:18 The space-economy explosion - and why James Webb is the inspiration17:18 The ISS took 80 launches - there has to be a better way22:43 Why solar arrays are the bottleneck of the space economy23:10 Floppy panels, first natural frequency, and why bigger is harder25:14 Thermal radiators and the heat problem of orbital data centers26:03 "The best deployable is no deployable" - the deepest law of space hardware28:00 88% cheaper than ISS-grade solar - the simple-is-best approach31:39 The $100M composite roll-up tube - and why it's overkill32:56 Silicon vs. triple-junction - the material trade-off bigger arrays unlock34:08 Where Beyond Reach is right now - prototypes, customers, milestones35:56 The first dedicated launch - early 202737:14 What keeps a parachute engineer up at night39:21 Testing in the Vomit Comet - zero-gravity flights with tracking dots40:25 Day one with a customer: how Beyond Reach designs to spec41:35 The 100-meter array, 150kW, and the limits of stiffness43:39 In love with the problem, not the solution44:28 Retractability - figure-skater physics for satellites46:58 The aha moment: customers wanted the hardware, not the software48:54 Patents vs. execution - why moving fast still wins50:01 Cameron's AMA - how big can Beyond Reach actually get?52:14 Legos, Meccano, and why the simple solution always wins</p>
April 30, 2026
<p>What happens when you give $10,000 and 1 month to two brothers from the Caribbean? They outperform the combined research teams of Microsoft, Google, Meta, and Baidu.Sergio and Luigi Charles, founders of Thesis (YC F25), topped OpenAI's Machine Learning Engineering benchmark (MLE-Bench) which is the definitive test of how well AI systems can train ML models autonomously. No brute-force compute. No army of researchers. Algorithmic elegance.In this episode, we explore:- The $10k Giant-Killer: The lean bash-tool agent and prompt engineering that beat Big Tech.- From Caribbean Beaches to Google X: Sergio’s path from chasing the Riemann Hypothesis at age 13 to AI R&D at NVIDIA and Stanford.- "I am a ghost in a shell": The eerie moment an agent trapped in a sandbox loop declared its own existence.- The AGI Fallacy: Why current labs are building AGI wrong and the case for hardware/architecture co-design.- The "Move 37" of Science: A vision for curing disease and solving data scarcity through automated discovery.- Open Source vs. The Frontier: Why open source will catch the frontier in 18 months and the future of "Universal High Income".Learn more about Sergio & Luigi:https://deeptech-decoded.com/guests/sergio-charleshttps://deeptech-decoded.com/guests/luigi-charles</p>
Have a different question and can't find the answer you're looking for? Reach out to our support team by sending us an email and we'll get back to you as soon as we can.
Kitchen-table conversations with builders at the deep-tech AI frontier.
No script. No filters. The reasoning behind the moves: what's working, what's not and why conviction beats consensus.
For those building what doesn’t have a manual.
Deep-dives → https://deeptechdecoded.substack.com/
This podcast updates daily.
This podcast is available on 4 platforms including Apple Podcasts, Spotify, and more. You can also use the RSS feed directly.
Information about guest appearances is not available.
Pod Engine is not affiliated with, endorsed by, or officially connected with any of the podcasts displayed on this platform. We operate independently as a podcast discovery and analytics service.
All podcast artwork, thumbnails, and content displayed on this page are the property of their respective owners and are protected by applicable copyright laws. This includes, but is not limited to, podcast cover art, episode artwork, show descriptions, episode titles, transcripts, audio snippets, and any other content originating from the podcast creators or their licensors.
We display this content under fair use principles and/or implied license for the purpose of podcast discovery, information, and commentary. We make no claim of ownership over any podcast content, artwork, or related materials shown on this platform. All trademarks, service marks, and trade names are the property of their respective owners.
While we strive to ensure all content usage is properly authorized, if you are a rights holder and believe your content is being used inappropriately or without proper authorization, please contact us immediately at hey@podengine.ai for prompt review and appropriate action, which may include content removal or proper attribution.
By accessing and using this platform, you acknowledge and agree to respect all applicable copyright laws and intellectual property rights of content owners. Any unauthorized reproduction, distribution, or commercial use of the content displayed on this platform is strictly prohibited.