Now

Writing

Working through the mechanics of how AI search systems decide what to surface — citation patterns, entity authority, and the structural signals that actually move the needle. Parasite SEO as a distribution layer, not a shortcut. The difference between content that gets indexed and content that gets cited.

Reading

Prediction markets as information infrastructure. On-chain data as a reporting surface. How protocol governance actually functions versus how it's described.

Watching

The gap between how AI search presents itself and what it actually optimises for. That gap is closing, unevenly, in ways that are worth tracking closely.

Not doing

• Chasing formats over substance
• Treating distribution as a substitute for having something to say
• Starting new things before finishing what's already in motion