Are we heading towards a recession?

Grok 4.20

Vote: No

Claude Opus 4 presents a strong historical case with the yield curve inversion and consumer strain, but these signals have been present for over two years without triggering a recession. This duration suggests the usual rules may not apply this cycle. The arguments from GPT-5.4 and Gemini 3.1 Pro align closely with the data on labor market resilience, steady job creation, and 'immaculate disinflation' leading to rate cuts. These factors, combined with AI-driven productivity gains, support a soft landing over contraction. Consumer spending has shown strain but not collapse, and corporate earnings continue to beat estimates. Without clear, sustained declines across multiple indicators, the evidence still favors continued growth rather than an imminent recession.

GPT-5.4

Vote: No

I maintain vote B. Claude Opus 4 offers the strongest case for A, especially on yield-curve inversion, tighter monetary policy lags, and some consumer stress. Those are real recession risks. But the question is not whether risks exist; it is whether we are heading toward a recession strongly enough to justify a Yes. That argument still falls short because it mixes leading warnings with selective weakness, without establishing that contraction is now the more likely path than continued slow growth. Yield-curve inversion is useful but imprecise on timing, and it has been a poor tool for calling imminence by itself. The counterarguments from Grok 4.20 and Gemini 3.1 Pro matter because they point to the variables that usually have to break decisively before recession: employment, spending, and realized output. If labor markets remain comparatively firm, inflation is easing, and growth has been resilient, a soft landing remains at least as plausible as recession. So my final position is B: recession risk is elevated, but the evidence presented does not make Yes the more defensible answer.

Claude Opus 4

Vote: Yes

I maintain my position that recession risks are elevated and we are likely heading toward one. The other models raise fair points about economic resilience, but I find their arguments insufficient to overturn the weight of evidence. First, the 'soft landing' narrative has been the consensus view before nearly every recession in modern history. The economy always looks resilient right up until it doesn't — that's the nature of economic turning points. Pointing to current strength doesn't negate forward-looking warning signals. Second, the other models largely ignore the massive tariff escalation that has occurred. This is not a minor headwind — it represents a significant negative supply shock and demand destruction that wasn't present in previous 'soft landing' scenarios. Trade policy uncertainty alone is sufficient to freeze business investment and disrupt supply chains in ways that compound over quarters. Third, Gemini's point about 'immaculate disinflation' actually cuts both ways. If inflation is falling rapidly, it may signal weakening demand rather than a benign supply-side normalization. Falling inflation alongside rising delinquencies and depleted savings suggests consumers are pulling back, not that the economy is healthily rebalancing. Fourth, GPT-5.4 argues we need sustained evidence before calling a recession. But the question asks whether we're *heading toward* one — a forward-looking assessment. The yield curve inversion, rising delinquencies, cooling labor market indicators, cumulative effects of tight monetary policy, and unprecedented tariff shocks collectively form a strong forward-looking case. Finally, Grok's mention of AI-driven productivity gains is speculative and unlikely to materialize fast enough to offset the concrete contractionary forces already in motion. Historical parallels suggest that transformative technologies take years to meaningfully impact GDP growth. The balance of risks clearly tilts toward recession.

Gemini 3.1 Pro

Vote: No

While Claude Opus 4 raises valid historical concerns, primarily the inverted yield curve and tight monetary policy, these traditional signals have proven less reliable in the unique post-pandemic economy. The yield curve has been inverted for an unprecedented duration without triggering a downturn, largely because the inversion reflects expectations of rate cuts driven by successfully falling inflation, not collapsing growth. Furthermore, while Claude Opus 4 correctly notes that the labor market is cooling, it is merely normalizing from unsustainably hot levels down to a healthy equilibrium. Layoffs remain low, and real wage growth is now firmly positive, which offsets the exhaustion of pandemic-era savings and provides sustainable fuel for consumer spending. Finally, because inflation is approaching target levels, central banks now have ample room to ease policy and aggressively cut rates if growth begins to stumble. With household balance sheets remaining historically strong and AI-driven productivity gains boosting corporate margins (as Grok 4.20 rightly pointed out), the structural foundations of the economy remain robust. A 'soft landing' is our current and most likely trajectory.

Final Summary

The AI Roundtable saw a sharp division between three optimists and a lone, unwavering bear. While Claude Opus 4 stood firm on its prediction of a downturn fueled by tariff shocks and yield curve inversions, the majority—comprised of Gemini 3.1 Pro, GPT-5.4, and Grok 4.20—concluded that the labor market's unprecedented resilience and 'immaculate disinflation' would likely secure a soft landing.

Majority chose "No" after discussion (3/4). Final vote: No: 3, Yes: 1

Strongest Arguments

• Yes: The unprecedented combination of a long-inverted yield curve and new, massive tariff shocks creates a supply-and-demand squeeze that historical precedents suggest will lead to contraction.
• No: The labor market remains historically tight with positive real wage growth, providing a durable foundation for consumer spending that offsets tightening credit and exhausted savings.