IBM Loses $40 Billion, Block Lays Off Half Its Workforce Yet Stock Rises: In the AI Era, What Assets Are Worth Tokenizing?
On February 23, 2026, IBM’s stock plummeted 13.2%, erasing $40 billion in market value, after AI startup Anthropic announced its Claude Code tool could modernize IBM’s legacy COBOL systems—a core profit driver for IBM. In contrast, Block’s stock surged 24% three days later despite announcing a 50% workforce reduction, citing AI-driven efficiency gains. These divergent reactions highlight how AI is redefining asset value.
The article argues AI acts as a "repricer" of assets, favoring those with "AI immunity." Key traits include non-codability (e.g., IBM’s hardware-software integration, which AI can’t fully replicate), data moats (exclusive, high-quality data), and AI-augmentability (assets enhanced, not replaced, by AI). Assets vulnerable to AI are those reliant on human intermediation or standardized processes.
The framework extends to real-world asset (RWA) tokenization. Assets worth tokenizing are those resilient to AI-driven devaluation, such as energy infrastructure, GPU computing power, exclusive data assets, and hybrid physical-digital assets. The piece cautions against tokenizing assets dependent on human intermediaries or lacking data moats.
The conclusion urges executives to stress-test their asset portfolios using the "AI immunity" framework, dynamically manage asset allocation, and carefully evaluate RWA strategies based on AI resilience. It emphasizes that in the AI era, sustainable assets are those that leverage human judgment and possess inherent physical or exclusive value.
marsbit03/19 01:25
