Are you capturing the full ROI of your AI investments — or are hidden costs quietly eroding it?
In this episode of AI Insights, Dell’s Global CTO and Chief AI Officer John Roese tackles one of the most talked-about challenges in enterprise AI: tokenomics. The decisions teams make about which models to use, where to place AI workloads and how to architect applications aren’t just technical — they’re economic. Get them wrong, and you can find yourself spending $100 to produce $1 of value.
John shares why hybrid AI architectures, model diversity and matching the right tool to the right task are essential to navigating the path to value — allowing you to scale AI without letting costs scale out of control.
Chapters:
00:00 — Why Token Costs Could Undermine Your AI ROI
03:16 — Innovate First, Optimize Second
05:07 — Model Diversity Is No Longer Optional
07:30 — The Application Is Not the Model
09:45 — You Don’t Need State of the Art for Every Task
11:23 — ROI Is About the Investment, Not Just the Return
