AI Disruption, Infrastructure Bottlenecks, and the New Laws of Software
An analysis of how AI is dismantling traditional software moats, creating massive infrastructure bottlenecks in power and memory, and shifting the paradigm of venture capital. The discussion explores the transition from 'SaaS-pocalypse' to a world where human identity verification becomes critical.
The Shift in Software Physics
In the current AI revolution, the 'laws of physics' for building technology companies have fundamentally changed. Historically, software moats were built on customer lock-in, proprietary data, and high switching costs. However, AI is eroding these defenses. It is now possible to replicate code and move data with unprecedented speed, and the role of the user interface is becoming secondary as AI agents begin to interact with software directly. For incumbents, the 'SaaS-pocalypse' represents a crisis of terminal value, where traditional defenses no longer protect long-term profitability.
Infrastructure: The Critical Bottleneck
While the focus is often on GPU availability, the true bottlenecks are shifting toward physical infrastructure. The US is facing a severe shortage of rare earth minerals, electricity, and memory (DRAM). The demand for AI tokens is growing vertically, but the capacity to build power transformers and electricity grids cannot scale at the same pace. This creates a massive opportunity for investment in the 'hard' side of technology—the physical components and manufacturing capacity required to sustain the AI era.
The Convergence of AI and Cryptography
As AI makes personalized communication (emails, phone calls) indistinguishable from human interaction, the risk of fraud and social engineering increases. This necessitates a shift toward hardware-based roots of trust and cryptographic keys to verify identity. Blockchain and game-theoretic properties will likely become the essential infrastructure for verifying if an entity is human, signing content to prove its origin, and allowing AI agents to act as independent economic actors through 'internet money'.
The Future of Entrepreneurship
Despite the disruptive nature of these changes, the long-term trajectory of technology historically leads to improvement. The 'gate' for entrepreneurship has lowered; 8 billion people can now materialize ideas into products, music, or movies. While the transition is frightening, it mimics previous industrial shifts—such as the move from farming to industrial work—where new, currently unimaginable needs and industries will emerge to replace those lost to automation.
Key insights
-
AI is dissolving traditional software moats like customer lock-in, proprietary data, and switching costs, making it easier to replicate code and shift data.
Impact: Incumbent software companies must pivot their value proposition from feature-set and lock-in to distinct, high-value outputs to avoid obsolescence.
-
The primary constraints on AI scaling are no longer just chips, but physical infrastructure: electricity, memory (DRAM), and power transformers.
Impact: Investment will likely shift toward power grid modernization and manufacturing capacity for rare earth minerals and memory components.
-
AI-driven personalization of communication makes traditional identity verification (CAPTCHAs) obsolete, necessitating cryptographically signed content and identity.
Impact: A surge in demand for decentralized identity solutions and cryptographic verification tools to prevent fraud and systemic trust collapse.
-
AI agents will require a native internet currency (crypto) to function as autonomous economic actors, as traditional banking systems are not built for non-humans.
Impact: This could drive a resurgence and practical application of cryptocurrency as a bearer instrument for AI-to-AI transactions.
-
The 'mythical man-month' is partially invalidated by AI; with sufficient capital and data, companies can now throw money at the problem to solve software challenges faster than previously possible.
Impact: Development cycles are compressed from years to weeks, accelerating the pace of disruption and reducing the time-to-market for new features.
Action items
-
CEOs of legacy software companies should conduct a 'moat audit' to identify if their value is tied to lock-in or if it provides a distinct, irreplaceable value that AI cannot easily replicate.
Impact: Prevents catastrophic valuation drops by forcing a pivot to high-value, specialized services before the market erodes the existing product.
-
Infrastructure investors should target 'unsexy' bottlenecks in the power supply chain, such as power transformer manufacturing and rare earth mineral procurement.
Impact: Alleviates systemic bottlenecks that prevent the deployment of more GPUs and data centers, creating high-yield investment opportunities.
-
Enterprises should implement hardware-based root-of-trust and cryptographic signing for high-stakes corporate communication to mitigate AI-driven social engineering.
Impact: Reduces the risk of fraudulent wire transfers and identity theft caused by hyper-realistic AI deepfakes of executives.
Quotes
“The same forces that let startups move faster are dissolving the moats that protected incumbents.”
“I think we'll probably have enough chips long before we have enough electricity.”
“The best companies have hostages, not customers.”