Between Worlds Charles Sansbury on data, AI, and the rise of the Exascale Organziation
Mar 3, 2025
Charles Sansbury, CEO of Cloudera and veteran data leader, talks about data infrastructure, private AI, and exascale operations. He covers the shift from AI theory to ROI and data quality. He explores private models, on-prem vs cloud tradeoffs, agentic interfaces, data lineage, and how organizations must reskill and reorganize for AI-driven scale.
AI Snips
Chapters
Books
Transcript
Episode notes
Proprietary Data Trumps Model Hype
- Foundation models are commoditizing while proprietary data becomes the differentiator.
- Private AI (targeted models trained on internal data) offers better accuracy, cost-efficiency, and control.
Build Targeted Models On Owned Data
- Train smaller, targeted models on your proprietary data to improve performance and reduce compute costs.
- Keep proprietary data under your control rather than exposing it to public cloud repositories.
On-Premises Returns For Heavy AI Workloads
- Rising compute and energy costs are driving a shift back toward on-premises for always-on, large-scale AI.
- Cloud remains valuable for variable, fast-scaling needs, but isn't a universal cost saver.
