AI Productivity Boom Shapes Macro Outlook for Digital Assets

Large language models have taken centre stage across professional sectors since the public introduction of ChatGPT. The platform is currently projected to host 800mn weekly users while Microsoft estimates that roughly 25% of the working age population in the global north actively utilizes artificial intelligence models.

The economic evidence of these productivity improvements remains fiercely debated despite widespread adoption over the last few years. Technology transformations historically require extended periods to materialize in macro data. The integration of electromagnetic induction took nearly a century to meaningfully boost electricity productivity, Federal Reserve Bank of San Francisco President Mary Daly recently noted.

Empirical evidence of productivity gains

Microeconomic studies show clear productivity improvements, but general support in macroeconomic data has been lacking until recent revisions from the Congressional Budget Office. Business sector productivity in the fourth quarter of 2025 registered 2.2% above pre-pandemic forecasts. The economic cycle since late-2019 is now showing the largest productivity increase since 1973 outside the dot-com bubble, Harvard University professor Jason Furman observed.

United States productivity increased by roughly 2.7% in 2025 compared to the 1.4% annual average of the previous decade. General purpose technologies often fail immediately to produce measurable gains due to periods of massive intangible investment that temporarily suppress output, Stanford University Digital Economy Lab director Erik Brynjolfsson detailed in a recent paper.

Microeconomic analyses consistently demonstrate double-digit productivity improvements. Boston Consulting Group consultants finished tasks roughly 25% faster when using the technology while randomized control trials at Microsoft and Accenture revealed a 26% increase in completed code pull requests, University of Chicago professor Alex Imas found.

Macroeconomic evidence remains significantly scarcer and generally points to productivity gains below 1%. Predicted gains over a ten-year period fall below 0.55% while self-reported time savings translate to a 1.1% total productivity increase, according to respective studies from MIT economist Daron Acemoglu and the St. Louis Federal Reserve. The Economist similarly estimates that the technology boosted global productivity by a maximum of 0.5 percentage points last year.

The stark difference between these two data sets stems from the narrow focus of micro studies where participants receive dedicated training for highly specific tasks. Broad employment roles consist of multiple integrated tasks where artificial intelligence might optimize one function while severe human bottlenecks remain in others, such as developers needing to manually review machine-generated code, Imas concluded.

Inflationary pressures and the monetary policy debate

A fierce ongoing debate centres on whether this massive technological investment and subsequent productivity growth will prove inflationary or disinflationary. The technology is structurally disinflationary and shares deep economic parallels with the early internet, Federal Reserve chair nominee Kevin Warsh recently said. Treasury Secretary Scott Bessent supported this specific view during a recent CNBC interview by noting the economy is in the nascent stages of a productivity boom that will cause inflation to drop substantially this year.

The structural developments could lower production costs and put downward pressure on prices but rapidly scaling the technology could simultaneously force upward pressure on certain price categories. The ultimate effect on inflation is not exclusively downward, Federal Reserve Vice Chair Philip Jefferson warned.

Realized productivity gains would inherently mean higher interest rates as the developments directly impact the neutral reference rate used to set monetary policy. The neutral rate would face a stronger upward bias if the technology delivers a material productivity impact, Cleveland Federal Reserve President Beth Hammack noted in a Wall Street Journal interview.

Market analysts remain significantly more agnostic regarding the deflationary narrative. Humans have been massively innovative since the Industrial Revolution yet inflation consistently surged during those same historical periods. Fiat currency regimes ultimately drive inflation rather than pure technological innovation, Deutsche Bank strategist Jim Reid observed.

Artificial intelligence serves as a general purpose technology that boosts overall productivity growth, but the exact impact on inflation remains highly conditional on household and corporate income expectations. Inflation will initially decline before rising if the income anticipation is imperfect while fully anticipated growth causes inflation to pick up in the short term and remain elevated, the Bank for International Settlements research concluded.

Historical policy parallels and risk assets

Policymakers frequently draw direct analogies to how Alan Greenspan responded to the technological shift in the mid-1990s. The internet evolution was in full force but failed to appear in official productivity measures while the labour market heated up and inflation remained elevated. The internet revolution allowed the economy to increase its potential output and grow at a higher pace without putting pressure on inflation, Greenspan successfully argued while keeping rates relatively flat.

Proponents of the current productivity narrative argue macroeconomic conditions allow for this exact policy accommodation to be repeated today.

Other institutional voices remain highly sceptical of the current productivity story. The recent gains are strictly cyclical and represent the direct aftermath of companies going on massive hiring sprees during the pandemic. Firms are now actively pushing for worker efficiency gains due to economic uncertainty caused by tariff policies and severe margin squeezes, TS Lombard Managing Director Dario Perkins countered.

Greenspan also changed his dovish views late in that decade as inflation began unfolding. The former chair ultimately realized the massive technological investment was pushing up neutral interest rates and forcing the central bank to hold higher base rates to keep inflation tightly checked, Perkins added.

The exact impact of this technological adoption on risk assets and the cryptocurrency complex remains incredibly difficult to gauge from a pure policy perspective. Hard macroeconomic evidence proving a sustained productivity boost has yet to fully materialize. The technology will likely weigh on both sides of the inflationary scale and severely complicate the mandate for central banks once it does arrive. Technological adoption will ultimately serve as just one of many structural factors shaping future monetary decisions alongside labour market health and broad economic conditions.