Karina Portugal: learning to ask the right question
THE SO WHAT
Karina Portugal’s shift from ad festivals to helping global banks ask “can we trust this system?” captures a broader pivot from storytelling to verification. If you’re deploying AI in regulated domains, the scarce skill isn’t prompt writing—it’s framing the right assurance questions and building the evidence to answer them.
READ THE SOURCE
MORE FROM THE WIRE
Applied AIThe AI trust gap: No scaling without quality management
The ‘trust gap’ framing is code for this: AI rollouts are stalling not on model quality, but on missing testing, governance, and accountability. If your AI roadmap doesn’t have an owner for evals, red-teaming, and audit trails, you’re not blocked by technology — you’re blocked by basic quality management.
Applied AIAmazon Seeks to Raise at Least $25 Billion to Fund AI Spending
Amazon tapping bond markets for at least $25B to fund AI infra cements AI capex as a balance-sheet event, not just cloud opex. If you’re building on these platforms, assume pricing, capacity allocation, and partnership terms will increasingly be shaped by how you help amortize that spend — via committed usage, data, or ecosystem pull-through.
Applied AIApollo’s Slok Says S&P 493’s Stalled Margins Are a Big-Tech Risk
If the S&P 493 isn’t showing margin lift from AI spend, the market will eventually question the multiple premium baked into Big Tech’s AI narrative. For non-megacaps, this is a warning to tie AI initiatives to hard P&L outcomes quickly — ‘strategic experimentation’ won’t survive the next earnings reset.
Applied AIWhy Nvidia's NemoClaw signals the true enterprise agent era
When an infra provider talks about “enterprise-ready autonomy,” the center of gravity shifts from chat interfaces to long-running, governed agents wired into production systems. CIOs should start defining which workflows they’re willing to hand to agents, what guardrails they require, and how they’ll audit decisions over weeks-long executions.