OpenAI's Revenue Report Deepens AI Bubble Fears, Firmus Shelves IPO
THE SO WHAT
Soaring model revenue alongside a shelved data center IPO captures the split-screen reality of the AI trade—software multiples stretch while infra players rethink public listings. If you’re in the picks-and-shovels layer, assume more private capital and bespoke financing; if you’re an AI app, assume higher scrutiny on durability of that top-line growth.
READ THE SOURCE
MORE FROM THE WIRE
Tech & InnovationAI Angst Won't Pop the Bubble Yet: Markey Analysis
If public-market analysts see AI ‘angst’ as noise rather than a bubble trigger, capital may stay available longer than operators expect—sustaining high valuations and aggressive spend. Use the window to lock in strategic financing and long-dated contracts, but model a sharper repricing when sentiment finally turns.
Tech & InnovationAsia Data Center Boom Is Vulnerable to US Policies, Jardine Says
If billions in Asian data center capex hinge on a handful of US hyperscalers and Washington’s China stance, then political risk is now a first-order variable in your capacity planning. Operators relying on Asia for AI workloads should diversify across jurisdictions and ownership structures—not just regions in a cloud console.
What smart people are saying about the US suspending new PERM applications for Microsoft, Adobe, and Indian tech giants
Suspending new PERM applications for major tech employers tightens the skilled-labor funnel right as AI talent demand spikes. If your roadmap assumes steady inflows of foreign engineers, you need a contingency plan—nearshore centers, remote hubs, or more aggressive retention of existing staff.
Tech & InnovationSources: Apple told some suppliers to cut iPhone 18 Pro and Pro Max component production by 15%-20% as higher memory costs pushed up prices and hurt demand
A 15–20% production cut on Apple’s flagship phones tied to higher memory costs is a reminder that BOM inflation can break even premium demand. Hardware and device builders should stress-test pricing power against component volatility—especially for AI-heavy SKUs with large memory footprints.