Chad Jones argues AI could either mostly extend the familiar 2% growth trend or trigger a much faster growth regime, with the key mechanism being whether AI keeps improving the economy’s bottlenecks or reaches enough weak links to unleash explosive feedback. He is optimistic about abundance but much more worried than a typical economist about catastrophic misuse or misalignment, saying the next five years are pivotal.
Watch on YouTube ›Get the market thesis, key claims, assets, contradictions, and follow-up questions from any financial video — then unlock a version personalized to your portfolio, watchlist, and favorite speakers.
This episode is a Stanford GSB If Then conversation between host Kevin Cool and professor of economics Chad Jones about AI, long-run growth, and society. Jones frames the discussion around two broad scenarios: a familiar “business as usual” path in which AI becomes the next major general-purpose technology but overall growth stays near historical norms, and a much more radical path in which AI increasingly automates both cognitive and physical tasks and drives growth sharply higher. …
Near term, the setup is more about capability shock and misuse risk than about a macro boom: watch software autonomy, cyber exposure, and model access closely. The most actionable risk is that bad actors or brittle infrastructure get hit before any productivity windfall is broadly visible.
Over the next several quarters to years, the base case is gradual but uneven AI diffusion: some jobs get more valuable as AI handles routine tasks, while others face displacement and wage pressure. A sustained breakout in growth would require AI to keep climbing the automation stack from software into robotics and broader task coverage.
Structurally, AI could shift the economy from a steady-growth regime into one where abundance is possible and work matters less as a source of income. The lasting issue is governance—preventing misuse and retaining control over increasingly capable systems while managing who captures the gains.
AI could either preserve historical 2% growth or eventually drive dramatically faster growth, depending on whether it reaches all human tasks.
Jones frames the paper around two extreme scenarios: continuity versus explosive automation-driven acceleration.
AI’s most plausible acceleration path is first through software engineering, then self-improvement, then virtual remote work, and eventually robotics.
He describes the canonical frontier-AI pathway often cited by industry leaders.
The historical U.S. growth trend is remarkably stable at about 2% per year over 150 years despite major inventions.
He uses this as the baseline business-as-usual scenario.
What is different about AI now compared with earlier periods of innovation?
Jones says there are two useful perspectives: automation has existed since the industrial revolution, but this wave feels different in important ways. He frames the paper’s simulations around historical automation and growth episodes to explore what AI could mean next.
What are the two AI growth scenarios in the paper?
He describes one scenario as business as usual and the other as a dramatic acceleration of economic growth. The high-growth case is driven by AI automating software engineering, improving AI itself, and eventually becoming capable of many human tasks through software and robotics.
What would constrain growth in the business-as-usual scenario?
Jones says the business-as-usual case matches the long-run U.S. pattern of roughly 2% annual growth in average living standards. He argues that even transformative technologies can be offset by ideas getting harder to find within each paradigm, so each new general-purpose technology helps sustain the same long-run growth trend.
Unlock the full claims, asset map, scores, related transcripts, follow-up questions, and AI chat — shaped around your portfolio, watchlist, favorite speakers, and risks.
