Perspectives

AI doesn't move through one market — it moves through lenses: the materials it needs, the capital that funds it, the applications that pay for it, the jobs it reshapes, and the skills it rewards. Power & grid is the lens with real depth today — our data-centre buildout deep dive. The other five stay honestly marked in research, with the questions we're working on, until the numbers are real.

1 of 6 perspectives published. New deep dives land here as the research does — watch the changelog.

• Power & gridPublished

  Published as our data-centre buildout deep dive

  Read the deep dive
• Materials & miningIn research

  Questions we're researching

  • How much of copper's move is AI data centers vs EVs and grid?
  • Which materials are true bottlenecks vs noisy correlated prices?
  • Where does mining geography create policy risk for the buildout?
• VC & capitalIn research

  Questions we're researching

  • When does debt financing of AI infra change the risk profile?
  • What would a 2000-style telecom outcome look like for each layer?
  • Which weekly signals do operators and investors actually watch?
• Application layerIn research

  Questions we're researching

  • Which app categories create durable inference load vs one-off training spikes?
  • Where are wrappers being repriced, and what survives?
  • How do vertical apps (health, legal, industrial) couple back to power demand?
• Jobs & laborIn research

  Questions we're researching

  • What jobs does the buildout create locally in the next 24 months?
  • Which roles are displaced vs augmented on a five- and ten-year clock?
  • How should we show distribution, not just net job counts?
• Students & skillsIn research

  Questions we're researching

  • Which skills compound across a decade-scale AI wave?
  • What should a student understand about power, chips, and capital — not just apps?
  • Where do apprenticeships beat credentials for buildout-adjacent work?