AI is winning over consumers — but it's way behind with businesses, say Goldman Sachs analysts (www.businessinsider.com)

Goldman Sachs analysts say the AI story is bifurcated: consumer-facing tools like ChatGPT and Claude are wildly popular and demonstrate clear user value, but enterprise adoption is lagging well behind expectations. While investor enthusiasm has driven heavy spending on AI infrastructure and pushed indexes to record highs, analysts caution that business use cases aren’t yet delivering commensurate returns — prompting questions about whether the current spending spree is justified. Technically, the surge is driven by generative models (e.g., ChatGPT, Google Gemini) that are already outpacing available compute capacity, prompting what Goldman calls an “upside surprise” in infrastructure buildout. Nvidia projects $3–4 trillion in cumulative AI infrastructure spending by decade’s end, a figure some investors find hard to reconcile with realistic ROI unless AI becomes a dominant driver of economic output. McKinsey’s State of AI 2025 echoes the mismatch: 88% of firms use AI in at least one function, but only about a third have scaled it enterprise-wide and just 39% see bottom-line impact. The takeaway for the AI/ML community: consumer proof points are strong, but realizing durable enterprise value requires deeper workflow integration, measurable ROI metrics, and more efficient compute economics.