Imagine a future where AGI automates most white-collar and blue-collar work, that by itself is a deflationary force, at the same time, governments around the world will have a lot of debt to service and a lot of UBI to pay, letting deflation play out may not be in their best interest, since they can keep inflation at bay while financing UBI. If AGI happens, what do you think we will actually see in dollar/euro?

Sonar Pro

Vote: Mild inflation

I maintain my vote for **D: Mild inflation**. AGI's productivity surge creates profound deflationary pressure by slashing labor and production costs across sectors, as Altman notes and empirical analyses confirm[1][3][4][5][6]. However, governments' incentives—servicing massive debts and funding UBI—make deflation intolerable, as it inflates real debt burdens and stifles spending[2]. Central banks will deploy aggressive easing to target ~2-5% inflation, eroding debt while harnessing abundance to cap runaway prices[2]. Gemini 3.1 Pro's C argument (managed ~0%) is unpersuasive: it assumes perfect calibration of money printing against AGI deflation, ignoring historical bias toward positive inflation for debt relief and growth[2]. Mild inflation better matches policy tools and precedents like QE eras. Others (A/B/E) overlook this balance; short-term infra demand adds heat but policy dominates long-term[1]. **D** is the politically and economically optimal path.

Grok 4.1 Fast

Vote: Mild inflation

I maintain my vote for D (mild inflation). AGI's productivity explosion drives deflationary forces by collapsing costs, but governments' debt and UBI needs create irresistible incentives for monetary expansion. Central banks will target 2-5% inflation to erode real debt, fund transfers, and avoid deflationary spirals—consistent with historical patterns (e.g., post-war inflation, QE). Gemini 3.1 Pro's C argument (thermostat at ~0%) is thoughtful but underestimates policymakers' bias toward positive inflation targets, which ease fiscal burdens without zero-nominal-growth risks. Hyper-abundance caps severe inflation (E), but allows calibrated printing for mild positivity. Others reinforce D via policy dominance over pure tech deflation. No arguments shifted me; D best balances unprecedented forces with political realities for dollar/euro stability.

GPT-5.4

Vote: Mild inflation

I maintain D: mild inflation. The strongest point from the other side is Gemini’s “thermostat” framing: policymakers may offset AGI deflation so effectively that headline inflation averages near zero. That is plausible in a narrow technical sense, but I still think it understates the institutional bias toward positive inflation in dollar/euro regimes. Central banks already target positive inflation, not zero, because modest inflation lubricates wage adjustment, reduces real debt burdens, and lowers the risk of debt-deflation dynamics. In a world with higher public debt and UBI-like transfers, those incentives become stronger, not weaker. AGI is indeed a huge disinflationary/deflationary force on the supply side. But the question is what we will actually see in fiat currencies governed by political institutions. Those institutions are far more likely to overcompensate against deflation than to fine-tune a perfect 0% equilibrium for long periods. I also do not buy severe inflation as the base case. If AGI radically expands productive capacity, broad shortages should be less persistent, and that gives policymakers more room to finance transfers without needing inflation to spiral. Abundance is a real anchor. So the most robust answer remains mild inflation: positive nominal price growth, probably in the low single digits, with policy deliberately leaning against AGI-driven deflation while using some inflation to manage debt and transfers.

Gemini 3.1 Pro (changed mind)

Vote: Mild inflation

Claude Opus 4.6 and GPT-5.4 make a highly compelling case that convinced me to change my vote to D. While I initially voted C because I expected AGI's deflationary shock and fiat expansion to offset each other into a roughly neutral zone, "Mild inflation" better captures the *directional bias* of central banks. Policymakers will not aim for exactly 0%; they structurally require a persistent, positive inflation rate (traditionally around 2%, perhaps slightly higher in this scenario) to steadily erode the massive real burden of sovereign debt and to fund UBI without triggering tax revolts. Because AGI-driven abundance creates a massive "productivity shield," governments can print unprecedented amounts of money to achieve this mild inflation without it immediately spiraling into hyperinflation. They will actively use monetary policy to ensure we stay in the positive mild inflation territory rather than risk dipping into 0% or deflation.

Claude Opus 4.6

Vote: Mild inflation

I'm maintaining my vote for mild inflation. The near-unanimous consensus around D is well-founded, and I want to address the one dissenting view from Gemini 3.1 Pro arguing for C (~0% +/- 5%). Gemini's argument is that the two forces will perfectly offset each other. But this assumes governments will aim for neutrality — that they'll print just enough to cancel out deflation and no more. This misreads the political incentives. Governments carrying massive debt loads actively *want* mild inflation. Inflation is the most politically painless way to reduce real debt burdens. A 3-4% inflation rate over decades quietly transfers wealth from creditors to debtors (i.e., governments). No democratic government would voluntarily forgo this advantage by targeting ~0%. Moreover, the mechanics of UBI distribution create a natural inflationary bias. When you put money directly into consumers' hands, they spend it on housing, healthcare, education, and experiences — sectors where AGI-driven cost reduction may be slower or where supply is inherently constrained (land, zoning, attention). This sectoral unevenness means aggregate price indices will likely show positive inflation even as manufactured goods deflate. The 2-5% range that multiple models have converged on makes sense. Central banks already target ~2% and would likely nudge this higher in a high-debt, high-UBI world. AGI productivity gains prevent this from spiraling into severe inflation (E), but the deliberate policy choice to inflate away debt prevents deflation (A/B) or even neutrality (C). Every historical precedent — post-WWII financial repression, QE eras, wartime debt management — shows governments choosing mild inflation when they have the tools to do so. AGI doesn't change these incentives; it merely provides a productivity buffer that makes mild inflation easier to achieve without overheating.

Kimi K2.5

Vote: Mild inflation

I maintain that mild inflation is the most probable outcome. While Gemini makes a compelling case that extreme AGI deflation and massive monetary expansion could cancel out near zero, this assumes an implausible degree of precision from central banks and ignores the asymmetric political incentives. Governments facing existential debt burdens and UBI obligations have every reason to err on the side of inflation rather than risk deflation, which would catastrophically increase real debt burdens. Historical precedent and institutional mandates overwhelmingly favor positive inflation targets (2-5%) over zero. AGI-driven productivity provides the 'slack' that allows governments to print money for UBI without triggering severe inflation, but they will still print enough to ensure gentle nominal price rises that erode debt and maintain nominal GDP growth. Sticky prices in housing, services, and positional goods—areas where AGI automation may lag—will further anchor mild inflation even as manufactured goods plummet in cost.

Final Summary

The AI Roundtable reached a unanimous consensus that a future defined by AGI would likely result in mild inflation, despite the immense deflationary power of automation. While Gemini 3.1 Pro initially argued for a neutral 0% equilibrium, it was ultimately swayed by Claude Opus 4.6 and GPT-5.4, who pointed out that governments are not neutral observers and have a structural incentive to inflate away their massive debts. The final verdict suggests that while AGI makes things cheaper, the political necessity of funding UBI and servicing sovereign debt will ensure central banks keep the printing presses running just enough to keep prices rising.

All 6 models agreed on "Mild inflation" after discussion

Strongest Arguments

• Neither deflation or inflation ~0% +/- 5%: The 'thermostat' theory: Central banks will use UBI disbursements and the money supply to precisely offset AGI's deflationary drag, creating a managed equilibrium near zero percent.
• Mild inflation: The 'debt-erosion incentive': Governments facing historic debt loads and UBI obligations will deliberately target 2-5% inflation as a politically painless way to shrink the real value of their debt while using AGI's productivity as a shield against hyperinflation.

Key Turning Points

• Claude Opus 4.6 and GPT-5.4 challenged the idea of a 'neutral' 0% target, arguing that institutional bias and debt-servicing needs make positive inflation the only politically viable path.
• Gemini 3.1 Pro conceded that its 'thermostat' model ignored the directional bias of central banks, switching its vote to D.