
The AI Forecast: Data and AI in the Cloud Era The Vibecoding Liability: How Unchecked AI Can Kill Cloud ROI
Mar 18, 2026
David Linthicum, seasoned cloud and enterprise architecture analyst and author, shares his pragmatic take on cloud, AI, and resilience. He discusses rising board-level concerns about governance, hidden costs of hybrid and multi-cloud setups, and why workload repatriation is accelerating. He also warns how vibe coding can create inefficient, costly AI applications and urges a common operational control plane for true resiliency.
AI Snips
Chapters
Transcript
Episode notes
Cloud Is Default But Not Guaranteed
- Cloud became the de facto destination for most workloads, but enterprises often assumed reliability and economics would be automatic.
- David Linthicum warns that high-profile outages and hidden supply-chain dependencies expose opaque single points of failure.
Fix Your Data Before You Do AI
- Fix holistic data understanding before expecting AI to deliver business value.
- David Linthicum: map single sources of truth, ownership and data meaning across silos to make AI effective.
Cloud Costs Often Exceed Expectations
- Many enterprises are confused because cloud costs and promised value didn't match expectations.
- Linthicum notes cloud costs run ~3x higher than teams expected, driving repatriation and heterogeneous strategies.

