The Substitution Effect
- The Theory: In economics, the Substitution Effect suggests that as the price of AI “Capital” falls, firms will naturally replace expensive human “Labour”.
- The Reality: While the cost of using AI—via API (Application Programming Interface) tokens—is crashing, the cost of building the infrastructure is skyrocketing. This Sustainability Gap means today’s cheap AI might be a temporary “loss-leader” subsidised by tech giants.
- The Pivot: High-skilled developers aren’t being replaced; they are being upgraded. Because the demand for human Labour is a Derived Demand (based on the client’s need for original, accountable strategy), your value actually increases when you use AI to handle the “grunt work.”
- The Goal: Move from being a “service provider” (coding) to a “value architect” (strategy and accountability).
In our previous deep dive, we explored how AI (Artificial Intelligence) might finally fix the UK’s “flatline” economy. But as the cost of implementation continues to plummet, the conversation is shifting from a national level to an individual one. For business owners and developers alike, the question is no longer “Can AI help?” but rather “Who does AI replace?”
In economics, this is known as the Substitution Effect. When the price of one Factor of Production (like Capital) falls, firms are incentivised to use more of it. If AI is the “Capital,” does that mean the “Labour” (you) is on the chopping block?
The Econ Tie-in: Factors of Production & The Substitution Effect
In a traditional economy, businesses balance two main inputs: Labour (human effort) and Capital (machinery, software, and tools). The Substitution Effect occurs when the relative price of one changes. Right now, the “price” of AI Capital—measured in API tokens—is crashing. Logically, a firm should substitute expensive human labour for cheap AI capital.
However, this ignores Derived Demand. In economics, the demand for labour is “derived” from the demand for the final product. If a client demands a website that isn’t just “functional” but is “strategically brilliant,” the demand for the human who can guarantee those traits actually increases as the tools get cheaper.
The Tech/Agency Tie-in: Creative Strategy vs. Generated Content
At a surface level, AI is a master of the “Commodity Task.” It can write a boilerplate script or a 500-word SEO (Search Engine Optimisation) description in seconds. If your entire value proposition is “I write code,” you are competing directly with the falling price of Capital. However, an agency’s value isn’t just the output; it’s the outcome.
- AI-generated code lacks “Strategic Accountability.” If a database leaks, the LLM (Large Language Model) isn’t in the boardroom explaining the fix.
- AI-generated content lacks “Originality.” It is a statistical average of what has already been written. It cannot innovate; it can only iterate.
The Sustainability Gap: Is the “Cheap AI” Era a Bubble?
There is a catch. While the price of using AI (Inference) is falling, the cost of building it is skyrocketing. We are currently witnessing the AI Cost Paradox. In 2026, the world’s tech giants are spending hundreds of billions on data centres and GPUs (Graphics Processing Units)—the specialised chips required to “train” these models. Furthermore, the energy required to power these centres is becoming a massive financial and environmental burden. This creates aSustainability Gap. Currently, we enjoy “cheap” AI because of:
- Loss-Leader Pricing: Tech giants are subsidising the cost to win market share, essentially selling the “Capital” at a loss.
- The GPU Bottleneck: Supply chain constraints mean that while demand is high, the physical hardware is incredibly expensive to produce and maintain.
If the massive investment in infrastructure doesn’t pay off soon, we could see a “Market Correction.” AI might stop being a cheap commodity and return to being a “Luxury Capital” good.
When Capital Makes Labour More Valuable
When the price of a “complementary good” falls, demand for the main product rises. Think of AI as the fuel and the developer as the engine. As fuel becomes cheaper, people don’t stop buying engines—they drive further. Because AI handles the “grunt work,” the high-skilled developer is freed up to focus on architecture, user psychology, and complex integrations.
The result? The developer becomes more productive. In economic terms, their MRP (Marginal Revenue Product)—the additional revenue a business earns from employing one more unit of labour—goes up.
The Structural Unemployment Risk
We must be realistic: Structural Unemployment occurs when there is a mismatch between the skills workers have and the skills firms need. Those who resist AI “Capital” will find themselves too expensive to compete. The goal for the UK workforce is to become the “Strategic Pilot” that the machine requires.
The Angle: Why “Originality” is the New Gold Standard
If everyone uses the same AI models to build the same applications, “sameness” becomes a market floor. The firms that win in 2026 will be those that use the falling cost of AI (while it lasts) to subsidise human-led originality. 1. AI is the floor: It handles the standard, the repetitive, and the “good enough.” 2. Humanity is the ceiling: High-skilled developers provide the edge, the accountability, and the “never-been-done-before” logic.
Final Thought
The “Substitution Effect” is only a threat if your output is indistinguishable from the software’s output. By pairing AI Capital with human Strategic Accountability, UK agencies can move from being “service providers” to “value architects.”
The machine isn’t taking your seat at the table; it’s just clearing the plates so you can focus on the main course.
