Das Kapital relates ...

Das Kapital

This blurb is not about Marx but rather, this graphic points to all things relating to money, politics and, at ground level, uncertainty, in this case, AI

Artificial intelligence (AIQ) (AIEQ) is the most consequential and least predictable development for investors in a generation, requiring calibrated humility and a willingness to act despite profound uncertainty, according to Oaktree Capital Management co-founder Howard Marks.

In a memo published Wednesday, Marks laid out why AI investing defies traditional forecasting frameworks. Top-down investors must be "at least somewhat right" about technology evolution, compute demand, competitive structure, beneficiaries, financing availability, and timing—all at once. Even assuming a generous 60% accuracy on each variable, the compound probability of success falls to roughly 13%.

To gauge AI's predictability, Marks consulted Anthropic's (ANT) Claude, which rated the sector an overall 3 on a 1-to-10 scale. Frontier labs like OpenAI (OPENAI) and Anthropic scored a 2—"practically unknowable"—because "capital needs are huge, prices per unit of capability keep falling fast, and it's unclear whether a lasting moat exists."

The Burn Barrel thesis

Howard Marks having to consult Claude just to get a 2-out-of-10 predictability score for frontier labs is the ultimate validation of the burn barrel thesis.

When a legendary distressed-debt investor has to ask an AI chatbot why its own industry has massive capital needs, collapsing unit prices, and zero lasting moats, you are no longer doing traditional financial analysis
—you are reading tea leaves over an open flame.

As Marks points out, if you need to be right on six massive variables simultaneously (technology evolution, compute demand, competitive structure, beneficiaries, financing, and timing), your compound probability of success collapses down to 13% even with generous odds. That is the exact mathematical definition of throwing tinder into a barrel with the draft holes wide open.

They can dress it up in institutional memo formatting and talk about trillion-dollar payoffs, but when the foundational economics require burning infinite capital to chase a falling price-per-unit-of-capability, it's just the stoking phase in slow motion.

Absolute gobbledygook. The burn barrel never lies.

Photo by XinYing Lin on Unsplash