The Credit Edge by Bloomberg Intelligence Tech Debt Binge Is Just Getting Started
12 snips
Feb 10, 2026 
Anurag Rana, BI equity analyst focused on AI adoption and capex, and Robert Schiffman, BI tech credit analyst expert in corporate borrowing and long-dated debt, discuss the AI build-out and massive CapEx needs. They cover pace of industry adoption, training versus inference revenues, why firms sell ultra-long bonds, chip and token-cost risks, geopolitical supply threats, and who may face fallout if demand cools.
AI Snips
Chapters
Transcript
Episode notes
AI Build-Out Is Just Beginning
- Bloomberg Intelligence expects U.S. AI capital expenditure to exceed $4 trillion through 2030 and may rise further.
- Robert Schiffman says the spending and borrowing to build AI capacity are just beginning and far from complete.
Legacy Businesses Still Fund The Build
- Today AI revenues are emerging but most cash flow still comes from legacy cloud, ads, hardware, and services.
- Schiffman warns AI revenue is small versus the scale of current AI spending and build-out.
Prioritize Inference Over Training Revenue
- Focus on inference-driven revenue rather than training revenue to build sticky, recurring income.
- Anurag Rana urges investors to prioritize end-market applications that embed AI into products for sustainable monetization.