Blog
On AEM, enterprise architecture, AI, and the craft of technical leadership.
Enterprises are accelerating AI spend while trimming or flattening the human capability investment needed to make that spend productive. The result is not a values gap. It is a budgeting gap.
Most people anchored to a single AI model on launch day and never revisited the decision. Model selection should follow task class — the same discipline you apply to any other compute resource.
Blackmagic Design added a Photo page to DaVinci Resolve 21. The feature set is serious — but the more interesting angle is what it represents architecturally for creative production teams.
OpenClaw functions more like an architecture workbench than a single AI product — and that flexibility matters more than it first appears in enterprise environments where lock-in is already a constant design concern.
Most AI conversations in AEM focus on content generation speed. In enterprise environments, that is the wrong question. Operational readiness is the constraint — and there are three signals that reveal whether your architecture is actually prepared.
Most leaders measure AI value in the wrong unit — time saved on tasks. The real gains are in the coordination overhead between decisions, and most AI tools don't touch that at all.
Most teams deploying AI agents are making the same mistake — treating agents as faster tools rather than capacity multipliers. The bottleneck isn't the tooling. It's whether your lead developers know how to direct it.
The AI image generation conversation in enterprise is being led by the wrong people. Whether Midjourney or Firefly wins on quality is a secondary question. Whether either integrates into governed content supply chains is the one that actually matters.
Your AI initiative is probably already funded. This isn't an argument to slow it down. It's a warning about what's waiting underneath it — buried in the content model, not the AI layer.
The productivity gains from AI coding tools are real. So is the damage being done by using them as cover to eliminate entry-level hiring. The apprenticeship layer wasn't about output — it was about building the people who would eventually own the system.