Lemonade CEO Points Out Structural Issues in AI-Driven Deflation
Lemonade CEO Daniel Schreiber stated that AI-driven deflation will not uniformly cover the entire economy; prices in scalable sectors like software may drop significantly, while bottleneck sectors like land remain largely unaffected. Simply printing money as a demand stimulus tool is too blunt, as it can raise prices in non-deflationary areas, creating mismatched inflation.
He further pointed out that this non-inflationary monetary creation space only exists during periods of sustained productivity gains from AI and deflation. Once prices trend toward a new equilibrium, the room for monetary expansion shrinks, and at this point, labor substitution may be most severe while demand for Universal High Income (UHI) is highest. Schreiber mentioned the MOSAIC model proposed with his collaborators aims to address the composition issues, limitations of monetary policy, and UHI funding needs simultaneously.
Source: Public Information
ABAB AI Insight
Schreiber's analysis reveals the asymmetric distribution of AI productivity shocks at the sector level. Areas with near-zero marginal costs, such as software and digital services, are prone to rapid deflation, while bottleneck sectors like housing, energy, and healthcare, which have low supply elasticity, exhibit strong price rigidity, leading to overall inflation metrics masking structural imbalances. This composition issue amplifies the bluntness of monetary policy: accommodative policies elevate all demand but translate into price pressures in bottleneck areas rather than uniformly stimulating output.
This observation embeds long-term mechanisms of wealth distribution and class mobility. AI accelerates the accumulation of excess profits in capital and technology-intensive sectors while exacerbating labor market substitution, shrinking traditional wage income channels. The proposal of redistribution schemes like UHI attempts to capture part of the productivity surplus from frontier firms to the societal level through dynamic VAT or AI dividend mechanisms, preventing excessive concentration of wealth among a few capital owners.
Historically, this reflects the recurring institutional constraints during technological substitution cycles. In past industrial revolutions, productivity gains were accompanied by new demand creation and job migration; in the current AI phase, the disconnection between digital bottlenecks and the physical world makes adjustments more challenging. Schreiber's emphasis on the "peak funding demand after price stabilization" highlights the risk of temporal mismatch: as deflationary dividends narrow, social stability pressures peak, testing the adaptability of fiscal and monetary frameworks under decoupling of productivity and distribution, driving a slow reassessment of capital from traditional labor-intensive paths to mixed redistribution models.