It’s not a storm — it’s the climate

It’s not a storm — it’s the climate

Sparkmind Invested EPS 2: Polish Grid, LQMs & SpaceX

5 February 2026

Summary:

Colin, Martti and Chris discuss the three conversations that matter this week. 

Optimising for value, not views.

  1. Polish Grid Attack
  2. Sandbox AQ and Large Quantitative Models (LQMs)
  3. SpaceX & xAI and 1 million satellites 

Remember, this show is for entertainment and is not investment advice.

TAKEAWAYS:

Polish Grid Attack:

  • A fire drill run by the arsonist
  • Designed to hide systems
  • Decentralised grids offer a huge attack surface
  • 165M euro investment by the Polish state
  • A warning to all European energy grids

SandboxAQ:

  • The rise of LQMs
  • Google spin-out
  • Maths is the language of the universe
  • Europe needs to build this capability, or the best we can do will be the “Loveable” to Sandbox’s ChatGPT

SpaceX:

  • IPO prep or real opportunity
  • Leapfrog Google in the AI race
  • Another Elon pipe dream or a stroke of genius
  • Blurring the line between robots and satellites
  • Robots will need a massive amount of compute
  • SpaceX’s “Act 3”
  • What will be Europe’s response to the cost of launching
  1.  

It’s not a storm — it’s the climate

It’s not a storm — it’s the climate

Sparkmind Invested EPS 1: Donut, Drones & Inference

22 January 2026

Summary:

Colin, Martti and Chris discuss the three conversations that matter this week. 
Optimising for value, not views.

  1. Donut Lab / Nordic Nano solid-state batteries
  2. Harmattan AI – 200M raise and everything Drone
  3. Fractile  – The importance of Sovereign Intelligence and the rise of inference! 

TAKEAWAYS:

Donut Lab / Nordic Nano:

  • The Donut Motor is world-class
  • Creative chaos signal
  • Founder motivation
  • Chemistry is binary
  • Impossible vs merely late
  • Vertical integration vs horizontal innovation

Harmattan AI:

  • High talent density young team
  • Prime buying capability
  • Multi-directional signal

Fractile AI:

  • Inference is so important
  • Europe needs to provide an upgrade  to its own manufacturing base
  • Without inference AI doesn’t work
  1.