Inference budgets are overrunning by "orders of magnitude" - what now?

59 · Thestack · April 20, 2026, 4:54 p.m.
Summary
The blog post discusses how the financial industry, specifically Goldman Sachs, is coping with rising inference budgets through the redistribution of value via open-source models and focusing on proprietary systems. It highlights the challenges and adjustments developers face in this rapidly changing landscape.