Is AI the Next Microprocessor or the Next Shipping Container?
For anyone investing in or betting on the next big thing, the central question is where generative AI sits in the history of technological revolutions. Will it create a new wave of wealth for startups and their backers, or will its value be captured elsewhere? In his article, “AI Will Not Make You Rich,” Jerry Neumann presents a compelling framework for this analysis, arguing that AI more closely resembles the economic impact of shipping containerisation than the microprocessor.
This framework, based on the work of economist Carlota Perez, helps analyse how technological revolutions create and distribute value. It suggests that where we are in a broader technological cycle is the primary determinant of who wins.
The Four Phases of a Technological Wave
Carlota identifies four phases in every major technological revolution. These cycles describe a predictable human and financial response to new technology.
- Irruption: The initial period of experimentation and uncertainty, where a new technology emerges. Investment is risky and driven by true believers.
- Frenzy: A period of speculative mania where capital floods into the new paradigm, leading to a bubble as hype outpaces reality.
- Synergy: After the bubble pops, a “golden age” begins. The technology is mature enough for practical application, and companies focus on building real businesses.
- Maturity: The technology becomes fully integrated. Growth slows, markets consolidate, and innovation becomes incremental, focused on extending the current paradigm.
Two Models of Innovation
Jerry contrasts two historical innovations to illustrate how these phases play out. The key difference is the element of surprise.
It’s the element of surprise that should strike us most forcefully when we compare the early days of the computer revolution to today. No one took note of personal computers in the 1970s. In 2025, AI is all we seem to talk about.
This lack of surprise shapes the entire competitive landscape.
| Feature | The Microprocessor (Irruption Phase) | Shipping Containerisation (Maturity Phase) |
|---|---|---|
| Initial Perception | A niche solution; its potential was a surprise. | The benefits were obvious to everyone from the start. |
| Competition | Incumbents were slow to react, creating a moat for startups. | Incumbents and competitors reacted immediately, leading to intense competition. |
| Innovation Style | Permissionless and distributed tinkering by enthusiasts. | Centralised and capital-intensive; required huge investment. |
| Value Capture | A new class of entrepreneurs and investors became wealthy. | Value flowed to customers and a few early pioneers. |
Containerisation arrived late in its technological wave. Its benefits were so clear that it sparked an immediate competitive free-for-all, squeezing profits. The real winners were downstream businesses that used cheap, reliable logistics to reinvent their own industries.
The business ended up being dominated primarily by the previous incumbents, and the margins went to the companies shipping goods, not the ones they shipped through.
My Analysis: AI Is Not a Monolith
Jerry’s argument that AI is a maturity-phase innovation of the ICT wave is convincing. It is an extension of a 60-year quest for machine intelligence, built on existing infrastructure. Every major company has an AI strategy.
To understand where value might accrue, we must break down the AI stack:
Foundation Models: These are the “container ships” of our era. They are enormously expensive to build and operate, requiring vast capital for chips and data centres. This part of the market fits the containerisation analogy perfectly. Competition is fierce among the large, well-funded incumbents behind models like OpenAI’s GPT series, Google’s Gemini, and Anthropic’s Claude.
The Commoditisation Engine: The containerisation analogy is strengthened by the rise of highly efficient, semi-open-source models, particularly from Chinese companies like Alibaba (Qwen) and others. These models are often cheaper to run and focus on software optimisation, triggering price wars and accelerating the commoditisation of model performance. This directly mirrors the rate wars in container shipping that squeezed profits and suggests high margins will be difficult to sustain at the model layer.
AI-Native Applications: These are companies building products directly on top of foundation models. As Jerry argues, they risk being squeezed by the model providers through pricing or vertical integration. If an application becomes successful, the underlying model provider has every incentive to capture that value by raising prices or launching a competing feature.
AI-Enabled Businesses (Downstream): This is where the “IKEA and Walmart” opportunities lie. These are not tech companies, but businesses in sectors like law, finance, and healthcare that use AI to fundamentally reduce costs and improve services. The value is not in building the AI but in using it to create a new business model. A firm that uses AI to offer legal services at a fraction of the traditional cost is a modern equivalent of IKEA using containerisation to sell flat-pack furniture globally.
The lesson from Jerry’s analysis is clear: the most durable fortunes will likely be made not by building the next model, but by being the first to figure out how to use it to transform an old industry.[1]
The main wildcard in this analysis is the open-source movement. If cheap, powerful, and truly open models become widely available, they could enable the kind of permissionless tinkering that characterised the microprocessor era. This could potentially trigger a new “Irruption” phase, but for now, the market is dominated by the capital-intensive, centralised approach. ↩︎
