Follow the Money Friday: The GPU Bill is Due, but the Productivity is Missing
AI-generated image · US National Wire
Billions in compute CapEx are fueling upstream growth, but central banks and analysts warn that the downstream productivity revolution is a ghost in the machine.
Opinion: For years, the pitch deck for generative AI has been the same: a promised explosion in productivity that justifies any price tag on a H100 cluster. But as we look at the actual P&L of the global economy, the math isn't adding up. We are seeing a massive divergence between the cost of the tools and the value they actually create for the end-user.
As first reported by The Register, the Bank of England's latest Monetary Policy Report provides a sobering reality check on the UK economy. The central bank estimates that AI has increased UK productivity by only 0.5 percent over the last three years. This represents an annual productivity increase of just 0.1 to 0.2 percentage points starting in early 2024. Even in sectors that should be the primary beneficiaries, such as information and communications, productivity growth has failed to exceed pre-pandemic averages.
Crucially, the Bank of England notes that while AI-related activity accounted for nearly a fifth of UK GDP growth since the start of 2026, this growth is concentrated in the "upstream development and implementation" of the tech. In short: the money is being made by the people selling the shovels, not the people digging for gold.
Other data points suggest this disconnect is a systemic issue rather than a regional one. The Register cites a Gartner survey showing that while 75 percent of corporate board members expect AI to cut costs and boost productivity, a mere 3 percent of organizations have actually hit those targets. Despite this, 85 percent of those organizations plan to increase spending next year, even though 38 percent are already over budget.
Further evidence of the productivity void comes from a February study of nearly 6,000 executives, which found that over 80 percent of firms reported no discernible impact on productivity or employment from AI.
Financial watchdogs are now flagging this as a systemic risk. The Register reports that the International Monetary Fund warned last month that the spending spree could drag down the wider economy if promised profits fail to materialize. Similarly, in June, the Bank for International Settlements compared the current infrastructure investment to the dotcom boom of the 1990s, warning that tech giants are exposed if the massive returns they've bet on never arrive.
While the productivity gains remain elusive, the costs are already hitting the balance sheet. The Bank of England observed that surging demand for AI products, including memory chips, has driven up export prices in Asia, which in turn has increased import costs for the UK.