"I don't think you'll see a lot of that in 2026. It's really in 2027 and 2028 that you'll see more meaningful progress from AI." [2027 and 2028]
Solomon explicitly dates the inflection point for measurable AI benefits, attributing the delay to the difficulty of embedding process change—a structural constraint on adoption velocity.
Reasoning from this article
Solomon frames this as a universal constraint ('it's hard to create process change'), not a Goldman Sachs-specific issue. The article shows Goldman Sachs has already invested 6-12 months in six process redesigns with minimal 2026 impact expected. This pattern—long lead times between AI investment and productivity realization—likely applies across financial services and other process-heavy industries, suggesting market expectations for near-term AI ROI may be systematically overoptimistic.