Upside #61 - Defence Special PLUS what IS the AI bubble really?

Voices: Nicholas Nelson (Archangel) • Dan • Lomax • Mads TL;DRDefence-first wins on capability and returns; primes are partners and channels.Helsing: buys platforms/revenue for access; layers AI—different from Anduril’s buy-TRL-tech + scale model.Beyond drones: biggest gap/opportunity is tactical EW.Procurement: more fast lanes (SOF, pilots); primes getting easier to work with.AI: real profits exist (esp. NVIDIA), but value chain is fragile; expect a correction, not a collapse. Picking winners > timing.02:40 — Why defence-first Beats dual-use on outcomes and returns; lifelong focus.04:32 — Definitions Customer = MoDs + primes; aim: lethality/readiness and societal resilience. Beware “defence-washing”.06:37 — What’s hot Avoid herd to drones only; counter-UAS, EW, human performance, deception, survivability.08:23 — Helsing buys Grob Neo-prime play: new co buys legacy manufacturing for platform access.10:42 — The two Defence M&A playbooks Anduril: buys mid-TRL tech (Area-I, Dive LD/Ghost Shark, Adranos) → scales via brand/distribution. Helsing: buys finished products/revenue (Mittelstand) → immediate customers; then add AI.14:25 — Prime status & capital Distribution + capital to AI-enable platforms.17:47 — Roll-up vs build Narrative “build”; execution “roll-up + build”.19:47 — Drones & ‘drone wall’ Layered answer: blunt with drones, hold with conventional forces.21:49 — The big one: Electronic Warfare (EW) NATO underinvested; tactical EW is the unmet need; legacy kit is ’80s/’90s.24:54 — Startup wedge Put EW at the edge (drones/aircraft/fixed) → near-term wins.26:33 — Baltic realism History, 2007–09 Estonia cyber, current incursions; likely Kaliningrad corridor.28:19 — Founder mistakes Tech ≠ win by itself; experience + gov engagement matters; US analogue: top funds have IC/SOF DNA.30:43 —  Are there really only a “Few buyers?” Many real buyers inside a MoD/DoD (services, sub-units, innovation orgs).36:23 — Sovereignty & US primes US strategics will buy abroad; Europe balancing autonomy with jobs/exits.41:07 — Starlink vs IRIS² Starlink’s lead + cadence; IRIS² slower—watch timelines vs evolving threats.47:18 — AI bubble? Warnings vs fundamentals; self-funded capex; real profits.49:37 — NVIDIA ramp $4.4B (2023) → $73B this year; growth tempers multiples.51:48 — AI Circular money & margins Cursor → Anthropic → hyperscalers → NVIDIA; only NVIDIA mints big margins; margin pressure coming (new semis, China, SLMs).53:12 — Picking beats timing Dot-com lesson: Cisco losses vs Amazon wins.54:19 — Capacity vs efficiency Capex likely useful long-run, but open source squeezes costs.55:52 — Platform risk Frontier labs moving up-stack; vertical AI + trust + data = moat.58:58 — Base case Likely correction (30–50%) at some point; timing is unknowable (not investment advice).