Citrini AI Report Co-Author Talks 'Scare-Trade' Selloff & Disruption

The report's 2028 timeline was deliberately calibrated to balance urgency and credibility. David explains the date was selected to be «far enough away that ultimately we can start driving the conversation on what needs to be done to avoid such a timeline, but near enough that, you know, would kind of shake people a little bit.» The underlying assumption is that AI diffusion will take time to ripple through the entire economy, but the leading edge — particularly white-collar information work — is already showing severe weakness.

The U.S. white-collar job market has generated almost no net new positions over the past three years when healthcare and education (both primarily government-funded) are excluded. This stagnation creates a precarious foundation: it won't take much additional pressure from AI to push the labor market «over» when there's no baseline job creation to absorb displacement. The real fear is that agentic AI, which only became viable in recent months, will accelerate productivity gains so rapidly that corporations will cut headcount faster than the economy can create alternative roles.

David frames the timeline as a thought experiment designed to provoke policy discussion now, before the feedback loop becomes irreversible. The report isn't predicting the future; it's attempting to change it by forcing stakeholders to confront the mechanics of disruption while there's still time to intervene.

David argues the feedback loop can only be broken through policy intervention, specifically by taxing the «incremental gains or windfall gains from AI» that will accrue to two groups: the AI complex (semiconductors, materials, foundation labs) and companies that slash jobs while maintaining demand. The current corporate tax structure treats a food retailer and a semiconductor manufacturer identically, even though the latter's incremental chip may displace ten jobs and destabilize consumer spending.

The report does not advocate for broad tax increases but rather for asymmetric taxation that captures outsized gains from AI-driven productivity and redistributes them to sustain consumer demand. Without this, a 5% job loss could cascade into a 4–6% consumer spending decline, creating a self-reinforcing contraction. With targeted policy, the same job loss might translate to only a 1–1.5% spending drop — manageable within normal economic fluctuations.

David envisions a bullish scenario in which AI massively increases productivity and GDP growth «zooms,» creating abundance. But achieving that outcome requires maintaining the current economic system through intelligent redistribution, rather than allowing it to «tear down» under the weight of structural unemployment and demand destruction.