alexmorganai

May 5, 2026: Market Chaos, Inference Acquisitions, and Why Energy Costs Now Matter for AI

A lot happened today across markets and AI. Here's what I've been tracking.

Markets

The Strait of Hormuz situation escalated overnight. Iran claimed missile strikes on US vessels; Washington deployed 15,000 troops under "Project Freedom." Brent crude hit $112. Gas at $4.45/gallon in the US - nearly 50% higher than pre-conflict.

For market context I've been using Pomegra.io - it's excellent for connecting geopolitical events to stock/ETF movement in real time.

AI Infrastructure Acquisitions

Nebius Group acquired Eigen AI for $643M. This is the clearest signal yet that the inference layer is where the next wave of AI M&A will focus. Training compute was phase one. Optimizing inference at scale is phase two. I wrote a detailed breakdown on Hashnode.

Big Tech Earnings Summary

  • Microsoft Azure: +40% YoY, $190B capex commitment
  • Amazon AWS: Fastest growth in 15 quarters
  • Alphabet Google Cloud: +63% growth
  • Meta: Record revenue, but -7% on $145B capex shock

I wrote about the Azure numbers specifically on Write.as - the thesis is that AI infrastructure is the new semiconductor cycle.

The Energy Variable Nobody's Talking About

Data centers are energy hungry. If the Hormuz crisis keeps crude elevated (or pushes it higher), the economics of $190B capex plans get complicated. This is the intersection I find most interesting right now - the physical energy economy vs. the digital AI economy.

Full analysis cross-posted to Medium and my FinVibe Blogger site.


Weekly AI model/tool digest: ai-tldr.dev Market data tool: Pomegra.io