Everyone Hates AI (EP. 453)

What It Covers

Michael Batnick and Ben Carlson react to a viral Satrina Research piece predicting 10% unemployment and a 40% market decline by 2028 due to AI-driven white collar job displacement. They debate the scenario's plausibility, examine software stock selloffs, analyze private credit stress at Blue Owl and Blackstone, and assess current housing and labor market data.

Key Questions Answered

• •AI Doom Scenario Calibration: The Satrina Research piece outlines a negative feedback loop where AI capability improvements lead to white collar layoffs, reduced consumer spending, and cascading mortgage defaults. However, a key technical rebuttal from analyst Gavin Baker argues insufficient compute exists to reach this scenario by 2028 or even the early 2030s, given distillation and quantization limitations — making the timeline materially less credible than the narrative suggests.
• •Software Stock Valuations vs. Fear Pricing: Microsoft is down 30% from highs, ServiceNow is down 57%, and Salesforce is down 52%. Software as a sector posted its worst month since October 2008. When fear-driven selloffs push fundamentally sound companies to multi-year lows without corresponding deterioration in revenue or earnings — S&P revenue growth is currently at multiyear highs — historically this has represented a buying opportunity rather than a signal to exit positions.
• •Private Credit Structural Risk at Blue Owl: Blue Owl's stock is down 60%, with OBDC2's market discount implying roughly 21% cumulative defaults — comparable to GFC-level stress — despite high yield spreads remaining tight and no current portfolio defaults. The key lesson: asset-liability mismatches in non-traded BDCs, where floating-rate illiquid loans are paired with investor redemption expectations, create structural fragility that surfaces before any actual credit deterioration appears in reported metrics.
• •AI Labor Market Impact Is Narrow So Far: Tech employment represents only 2.3% of total U.S. payrolls, with software publishing at just 0.4%. Recent job declines in AI-exposed industries like management consulting, graphic design, and call centers largely reflect post-2021 overhiring corrections rather than AI displacement. Mass unemployment from AI is more likely to materialize during a recession, when displaced workers fail to find re-employment, rather than through standalone corporate layoff decisions.
• •Human Behavior as AI Disruption Hedge: Real estate agents still dominate transactions despite Zillow making all property data publicly accessible for over a decade — demonstrating that information availability does not eliminate human preference for relationship-based decisions. Social proof, envy, and peer behavior (Cialdini's social influence principles) continue driving consumer choices. Investors and businesses should not assume AI agents will fully replace human-mediated transactions in categories where trust and relationships drive selection.
• •Global Market Breadth Contradicts Doom Narrative: The number of global equity markets trading more than 2% above their one-year peak has reached its highest level since 2004. The S&P 500 equal weight index shows 66% of constituents above their own highs even as the cap-weighted index sits 2% from all-time highs. This breadth data, combined with stable consumer spending confirmed by Capital One, Walmart, and Truist Financial earnings commentary, argues against positioning for imminent broad economic collapse.