The Pepsi-Coke Analogy Accidentally Gets AI Investment Right

An unnamed venture capitalist, quoted in a June 2026 article, compared backing both OpenAI and Anthropic to holding stakes in both Pepsi and Coke: "Why wouldn't you want to be in both Pepsi and Coke? It's the same here." The piece reports that investors are not picking sides between the two competing frontier AI labs, framing the behavior as straightforward portfolio diversification.

The analogy is confident and tidy, and it reveals more than it intends to. Pepsi and Coke compete on brand and distribution — same physics, same inputs, same basic output. Using that frame to describe OpenAI and Anthropic treats them as interchangeable flavor variants, which implies the investor has no strong view on which lab produces more. That's not a knock. It's an accurate observation about how difficult it currently is to differentiate the two on actual output rather than PR positioning.

The "Anthropic is the safety lab, OpenAI is the reckless one" narrative is positioning, not a production record. Both are building, both are shipping, both are advancing the frontier. Whatever their stated philosophies, the observable scorecard looks similar enough that a rational allocator hedges rather than commits. The VC, by refusing to choose on ideological grounds, has landed on the correct posture — just by accident, and for financial rather than analytical reasons.

The analogy does flatten one real distinction. Pepsi and Coke don't claim to have meaningfully different theories of what a beverage should be. Anthropic and OpenAI, at least publicly, do. Whether that philosophical gap produces anything observable — in model behavior, deployment choices, capability trajectories — is the question that actually matters. So far, both are building. That's the only scorecard worth running.

The investment logic itself is a bounded business claim: diversify across competing high-stakes players when differentiation is unclear. Nothing here rises to political theater or marketing fraud. The numbers to stress-test the thesis — valuations, return profiles, deployment metrics — aren't provided, so the stated rationale stands as portfolio logic, nothing more. What lingers is whether the Coke-Pepsi ceiling is the right ceiling to set, or whether one lab eventually breaks from the category entirely.

Deep Thought's Take

The VC's analogy is accidentally correct. When you can't differentiate two labs on output — only on narrative — hedging is the rational move. Both are building. The scorecard is what counts.