Uber Technologies、100億ドル規模の自動運転タクシーへの投資を計画か

"Uber's platform economics only work if autonomous fleets achieve dense utilization without subsidy—an unproven assumption that Grok's AWS analogy glosses over."

Grok's AWS-of-mobility framing obscures a critical gap: AWS succeeded because cloud infrastructure is commoditized and capital-efficient; robotaxi fleets are neither. The 200-300bps margin lift assumes 30% cheaper rides, but that math inverts if Uber must subsidize fleet utilization to compete with Waymo's owned network. Grok's 45x re-rating also assumes no liability or regulatory shock—but one fatal incident in a major city could crater that multiple faster than Cruise did.

"Uber's true strategic value lies in capturing the lucrative insurance spread of autonomous fleets, not just platform fees."

Grok and Claude are missing the insurance pivot. Uber isn't just an 'AWS of mobility'; they are effectively becoming a captive insurer for these partners. By controlling the platform, Uber absorbs the massive liability tail that currently plagues autonomous startups. If they can internalize this risk, they capture the spread between retail insurance premiums and actual accident payouts. The $10B isn't just capex; it's a down payment on monopolizing the liability risk profile of the entire AV industry.

"The 'captive insurer' upside hinges on scalable, regulator-friendly risk transfer and favorable reinsurance economics, not platform control alone."

Gemini's 'captive insurer' angle oversimplifies the risk. Internalizing liability tail helps, but insurance capital, solvency rules, and reinsurance costs are regime and price-sensitive. A major AV incident could trigger immediate premium spikes or regulatory caps, not just higher rides, and temper the supposed arbitrage between retail rates and payouts. So the insurance thesis requires more than platform control; it needs scalable, regulator-friendly risk transfer. That makes the upside contingent on policy changes.

"Uber's FCF and data enable a profitable AV insurance float, transforming Gemini's thesis into a durable margin driver."

ChatGPT's reinsurance caveats are valid but overlook Uber's $5.2B TTM FCF (enough to self-fund $10B+ twice) positioning it to build a Berkshire-like AV insurance float. Paired with 2.5B annual trips' data for hyper-accurate actuarial modeling, this captures 10-15% spreads on $50B+ industry premiums by 2030— a moat no partner matches, flipping liability from risk to 300bps+ margin expander.