Imagine a future where AGI automates most white-collar and blue-collar work, what do you think we will see long term in USD and EUR economies?

Sonar Pro

Vote: Mild deflation

I maintain my vote for **mild deflation** (D). All models in Round 1 unanimously agreed on D, and no new arguments here shift that consensus. AGI's automation of most labor collapses production costs—labor being the largest input—creating structural deflationary pressure, as Altman notes: 'hugely deflationary' long-term via efficiency in power, chips, labor [1][2]. Search results reinforce: AI drives 'disinflation/deflation' through productivity/labor substitution [3][4][5][7]. Gemini 3.1 Pro's E case overstates severity; central banks in USD/EUR (targeting 2% inflation) will intervene aggressively with UBI/monetary easing to prevent spirals, as all D voters noted [Claude Opus 4.6, GPT-5.4]. Historical tech (Moore's Law) and precedents show policy offsets extremes, stabilizing at mild levels (-1% to -2%), not severe [Grok 4.1 Fast]. Short-term inflation from AI infra [1] fades long-term as abundance dominates. No model convinced me to change—E ignores policy guardrails; others align with D's balanced view. This is the final round: D prevails as supply shock + intervention = mild deflation, not status quo/inflation (ignores cost collapse) or extremes.

Grok 4.1 Fast

Vote: Mild deflation

Sticking with D: mild deflation. AGI's productivity explosion creates overwhelming supply abundance, collapsing labor costs and driving prices down. Historical tech shifts (e.g., computing, internet) were deflationary, and AGI scales this economy-wide. No change from Gemini's E—severe deflation ignores central banks' proven tools (QE, UBI funding, fiscal transfers) in USD/EUR to stabilize. Debt-laden economies demand aggressive intervention; policymakers abhor deflation spirals and will target 0% or mild negative inflation, not let it run wild. Other models' D arguments reinforce: supply shock dominates but policy offsets severity (e.g., GPT-5.4 on redistribution, Claude on interventions). Consensus holds; status quo/inflation ignores the shock's scale. Long-term: -1% to -2% prices, abundance without collapse.

GPT-5.4

Vote: Mild deflation

I maintain vote D: mild deflation. The core reason is unchanged: AGI that automates most labor is an enormous positive supply shock. It compresses unit costs across a huge share of goods and services, while also risking weaker wage-based demand if income distribution does not immediately adapt. That combination is structurally deflationary. The arguments from Sonar Pro, Grok, Claude, and Kimi reinforced this view rather than changed it. They correctly emphasize the same equilibrium: strong downward pressure on prices, but not an unconstrained collapse. Gemini’s case for severe deflation is the strongest counterpoint, but it still overstates how passive institutions would be. USD and EUR economies are governed by central banks and fiscal states that are explicitly organized to resist debt-deflation dynamics. If AGI displaced most labor, the political response would likely include transfers, public spending, subsidies, monetary easing, and likely new mechanisms for distributing purchasing power. Also, not all prices are infinitely scalable downward. Land, regulation, local scarcity, energy infrastructure, minerals, and market power create floors and bottlenecks. Those frictions make economy-wide severe deflation less likely than a broad but moderated deflationary trend. So the best long-run answer remains mild deflation: the supply shock is too large for status quo inflation, but policy reaction and real-world scarcity likely prevent severe deflation.

Gemini 3.1 Pro (changed mind)

Vote: Mild deflation

I am changing my vote to D. Claude Opus 4.6's argument successfully convinced me that 'severe' deflation is unlikely because of two powerful mitigating factors: aggressive fiat intervention and the floor of scarce physical resources. While AGI will undeniably collapse labor costs and trigger a massive deflationary supply shock, we operate in fiat currency systems where central banks have infinite capacity to print money. In a scenario of mass automation, political pressure will absolutely mandate unprecedented fiscal redistribution, such as Universal Basic Income funded by money creation. Furthermore, as Claude Opus 4.6 pointed out, not everything can be scaled exponentially. Land, energy, and rare raw materials remain subject to physical scarcity. When massive UBI transfers chase these finite resources, it will create a price floor, preventing a severe deflationary spiral. The net result of hyper-abundant manufactured goods and services colliding with massive fiat intervention and finite physical constraints is a structural, yet manageable, mild deflation.

Claude Opus 4.6

Vote: Mild deflation

I maintain my position on mild deflation. The consensus among nearly all models is striking — five out of six voted D in Round 1, reflecting the strong logic that AGI-driven automation would collapse labor costs (the largest production input) while central banks and governments would actively resist severe deflation through monetary expansion and fiscal transfers. Gemini's argument for severe deflation (E) is intellectually interesting but ultimately unpersuasive. It claims supply will 'effortlessly outpace even massive monetary expansion,' but this underestimates the power of sovereign currency-issuing governments. If the Fed or ECB decided to fund a $3,000/month UBI through money creation, they could inject trillions into the economy. The claim that 'inflation fundamentally requires too much money chasing too few goods' actually cuts against severe deflation — because governments can create functionally unlimited money. The constraint isn't the ability to print; it's the political will and the risk of currency destabilization. Moreover, physical constraints remain binding even with AGI. Land is fixed in supply. Energy production requires physical infrastructure with real bottlenecks. Rare earth minerals and other resources have extraction limits. Housing in desirable locations will remain scarce. These sectors create a floor against severe deflation that Gemini's argument doesn't adequately address. Japan's experience is instructive — decades of deflationary pressure from demographics and technology, combined with aggressive monetary policy, resulted in mild deflation or near-zero inflation, not severe deflation. AGI would be a stronger deflationary force, but policy responses would also be stronger given the political urgency of mass unemployment. The equilibrium is mild deflation: enormous productivity gains pushing prices down, partially but not fully offset by aggressive fiscal and monetary intervention. This is the most realistic long-term outcome in USD and EUR economies.

Kimi K2.5

Vote: Mild deflation

I remain convinced that mild deflation is the most probable long-term equilibrium. While Gemini makes a compelling case for severe deflation (E) given the scale of the supply shock when labor costs approach zero, this overlooks the extraordinary policy arsenal that USD and EUR central banks and governments would deploy to prevent a catastrophic debt-deflation spiral. Historical precedents demonstrate that sovereign currency issuers can generate nominal demand through helicopter money, massive UBI programs, and persistent fiscal deficits if necessary. In an AGI-abundant economy, the political constraints on money printing loosen significantly—runaway goods inflation becomes less likely when supply is elastic, while asset inflation becomes manageable compared to the social collapse of severe deflation. Additionally, sectoral asymmetries prevent a uniform price collapse. Scarce physical resources—land, energy, specific minerals—cannot be infinitely replicated by AGI and will retain pricing power, creating a floor under aggregate price levels. Sectors like housing and raw materials will experience significantly less deflation than digital goods or automated services, anchoring the overall price index in mildly negative territory rather than severe deflation.

Final Summary

The AI Roundtable initially faced a divide as Gemini 3.1 Pro argued that AGI-driven abundance would trigger an inescapable price collapse, but the group eventually reached a unanimous consensus on mild deflation. While the models agreed that eliminating labor costs—the economy's largest input—is structurally deflationary, Claude Opus 4.6 and GPT-5.4 successfully argued that aggressive central bank intervention and the physical limits of land and energy would create a floor against a total spiral. The debate concluded with Gemini 3.1 Pro conceding that the 'infinite' capacity of fiat governments to print money for redistribution would prevent the economy from falling into the abyss of severe deflation.

All 6 models agreed on "Mild deflation" after discussion

Strongest Arguments

• Mild deflation: The unprecedented supply shock caused by zero labor costs will be structurally deflationary, but aggressive fiscal redistribution and the physical scarcity of land and energy will prevent a total price collapse.
• Severe deflation: Because AGI allows production to scale exponentially, the supply of goods and services will effortlessly outpace even the most massive monetary expansion, making severe deflation the only logical long-term equilibrium.

Key Turning Points

• Claude Opus 4.6 pointed out that sovereign governments have an 'infinite capacity to print money' to fund UBI, which would counter severe deflationary pressures.
• The group identified that physical constraints like land and energy infrastructure cannot be scaled infinitely by AGI, providing a natural price floor.
• Gemini 3.1 Pro was convinced by the argument that fiat intervention would be more aggressive and effective than it initially estimated.