AI's Economic Impact: Theory of Constraints
Let's use Theory of Constraints to predict where AI will set the economy on fire (and where it won't)
![](https://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fe8d5e713-b126-4948-a6d3-d2f113e532d6_1456x816.png)
Better, Faster, Cheaper, Safer
The core idea of Post-Labor Economics (PLE) is that as artificial intelligence, machines, and robots surpass human capabilities in key areas, they will inevitably replace human labor in many economic activities. This concept is encapsulated in the “Better, Faster, Cheaper, Safer” model, which provides a framework for understanding when and why this replacement occurs. This is called “labor substitution” and is relatively straight forward.
In this context, “better” refers to the quality and consistency of results and outcomes. Machines, once properly programmed and calibrated, can often produce results with higher precision and fewer errors than humans. For instance, in manufacturing, robotic assembly lines can achieve tolerances and consistency levels that surpass human capabilities. Similarly, in data analysis, AI systems can process vast amounts of information and identify patterns that might elude even the most skilled human analysts.
Post-Labor Economics is the assertion and prediction that machines should (and will soon) remove human labor as the chief bottleneck of economic productivity and human progress. In other words, PLE is about decoupling the economy from the human constraints.
Speed is another crucial factor where machines often outperform humans. High-frequency trading algorithms can execute thousands of trades per second, a speed impossible for human traders. In logistics, automated sorting systems can process packages at rates that would require multiple human workers to match. This increased speed can lead to significant productivity gains and economic advantages for businesses that adopt these technologies.
Cost is perhaps the most compelling factor driving the adoption of machine labor. Over time, the cost of machine labor often decreases due to technological advancements and economies of scale, while human labor costs tend to increase. The cost of computing power continues to decline following Moore’s Law, making AI and robotic systems increasingly cost-effective. In contrast, human labor costs, including wages, benefits, and training, typically rise over time due to inflation and increased standard of living expectations.
Safety is the final component of this model. In many industries, machines can perform dangerous tasks with less risk. In mining and hazardous waste management, robots can operate in environments too dangerous for humans. Autonomous vehicles have the potential to reduce traffic accidents caused by human error. This not only protects human life but can also reduce costs associated with workplace accidents and insurance. Safety is also about better outcomes than humans, such as robotic surgeons who make fewer mistakes.
When machines outperform humans across these four dimensions, their adoption becomes economically inevitable. This is driven by competitive pressure in market economies, where companies that adopt more efficient technologies gain an advantage. There’s also pressure from shareholders to maximize value, which often translates to increasing productivity and reducing costs through automation. Moreover, as consumers become accustomed to the quality, speed, and lower prices enabled by automation, market demand shifts towards products and services that can meet these expectations.
It’s important to note that this model isn’t without limitations. It may oversimplify complex economic dynamics and underestimate the potential for new job creation. It also doesn’t account for tasks where human touch is inherently valued, which is addressed in other aspects of Post-Labor Economic theory, which we shall partially cover in a moment. However, as a foundational concept, the “Better, Faster, Cheaper, Safer” model provides a useful framework for understanding the economic forces driving the adoption of AI and automation, suggesting we may be approaching a significant shift in the relationship between human labor and machine capital.
Social Contract
The social contract of human labor has been a cornerstone of modern economies for centuries. At its core, this contract represents an implicit agreement: individuals have the right, and often the expectation, to exchange their time, energy, labor, and bodily efforts for wages. This arrangement has shaped societal structures, personal identities, and economic systems worldwide. It’s the foundation upon which many people build their lives, plan for the future, and derive a sense of purpose and value. It is a foundational assumption in all economic theory to date, but AI might* throw the whole thing out the window.
However, the rapid advancement of artificial intelligence and automation is fundamentally disrupting this long-standing social contract. As machines become increasingly capable of performing tasks that were once the exclusive domain of human workers, we’re approaching what can be called the “frontier of automation” or the “horizon of automation.” This represents the ever-expanding boundary of what can be feasibly and economically automated.
From an economic standpoint, AI can be viewed as a form of advanced automation that dramatically expands the scope of what is automatable. Unlike previous technological revolutions that primarily affected energetic output and manual or routine cognitive tasks, AI has the potential to automate complex cognitive work, creative endeavors, and even tasks requiring emotional intelligence. This broad reach means that few areas of human labor are likely to remain untouched.
The disruption of the labor-based social contract by AI and automation challenges us to reimagine the very foundations of our economic systems and societal structures. We must grapple with how individuals derive meaning, contribute to society, and secure their livelihoods in a world where traditional employment may no longer be the norm. This shift demands a fundamental reevaluation of the relationship between work, value, and human fulfillment.
The disruption of the social contract of labor has profound implications. For many, work is not just a means of earning a living, but a source of social status, personal fulfillment, and societal contribution. As AI and automation encroach on more types of work, we may need to reconsider how individuals derive meaning, structure their lives, and contribute to society outside of traditional employment.
Moreover, this shift challenges our economic systems at a fundamental level. Many of our social structures, from education to healthcare to retirement systems, are built around the assumption of widespread human employment. As this assumption becomes increasingly tenuous, we may need to rethink how we organize our economies and societies.
The disruption also raises critical questions about economic inequality. As the owners of AI and automation technology stand to benefit disproportionately from increased productivity, there’s a risk of exacerbating existing wealth disparities. This could lead to a society divided between those who own the means of production (now largely AI and robots) and those who don’t, echoing concerns raised during previous industrial revolutions but potentially on a much larger scale.
In response to these challenges, various solutions have been proposed. These range from universal basic income schemes to radical redesigns of our education systems to prepare people for a post-labor economy. Some suggest we need to find new ways to distribute the benefits of automation more equitably, perhaps through mechanisms like a “robot tax” or by rethinking our notions of ownership and property rights in an age of abundant productivity.
Ultimately, the disruption of the social contract of labor by AI and automation forces us to grapple with fundamental questions about the nature of work, the structure of our economies, and even the purpose of human activity in an age of intelligent machines. As we navigate this transition, it will be crucial to find new ways to ensure social cohesion, provide for people’s needs, and create opportunities for meaningful contribution and personal fulfillment beyond traditional notions of employment.
Forever Jobs: Sapien Premium
In the framework of Post-Labor Economics (PLE), we can anticipate that as AI and automation become increasingly capable, the remaining demand for human labor will likely center around jobs for which people are willing to pay a premium specifically for the privilege of human involvement. This concept, which we might call the “sapien premium,” provides a lens through which we can predict which types of jobs may persist even in a highly automated economy.
The sapien premium arises from the inherent value that humans place on interaction with other humans in certain contexts. This preference isn’t necessarily based on superior performance or efficiency, but rather on the unique qualities that human workers bring to these roles. These qualities might include empathy, creativity, cultural understanding, or simply the comfort of human-to-human connection.
Some examples of professions that might command a sapien premium include clergy and spiritual leaders, where the human element is central to the roles’ purpose and effectiveness. People often seek spiritual guidance and comfort from other humans who can relate to their experiences and emotions in ways that machines, no matter how advanced, might struggle to replicate.
Similarly, certain types of care work, particularly those involving emotional support or complex interpersonal interactions, may continue to be valued when performed by humans. While robots might be able to perform physical tasks of caregiving more efficiently, many people would likely prefer the empathy and emotional connection that human caregivers can provide, especially in sensitive situations like end-of-life care or mental health support.
In a world of advancing AI and automation, the ‘sapien premium’ emerges as a key predictor of enduring human employment. It represents the unique value placed on human involvement in certain roles, not due to superior performance, but because of our innate preference for human interaction, creativity, and empathy. The sapien premium highlights areas where we may continue to value and seek out human labor, even in a highly automated economy.
Entertainment and sports are other areas where the human element might remain highly valued. While AI can create music or artwork, there’s a unique appeal to human-created art that resonates with our shared experiences and emotions. In sports, the physical prowess and competitive spirit of human athletes continue to captivate audiences in ways that machine performance might not match.
Certain aspects of education, particularly those involving mentorship, motivation, and nurturing creativity, might also retain a preference for human involvement. While AI can efficiently transmit information and even adapt to individual learning styles, human teachers bring a level of inspiration, role modeling, and adaptive interaction that many learners find irreplaceable.
However, it’s crucial to note that while the concept of the sapien premium helps us identify areas where human labor might remain valued, it doesn’t necessarily predict the volume of these jobs or their economic significance. We must be cautious about making broad predictions about whether there will be enough of these human-preferred jobs to sustain widespread employment or economic relevance for the majority of the population.
![](https://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F66339097-23b4-4ae5-99da-f8f564600bbe_880x345.png)
The question of whether humans will become economically irrelevant to the bulk of the economy remains open and hotly debated. While the sapien premium suggests that some human jobs will persist, it’s unclear whether these will be sufficient in number or economic impact to maintain the current social contract of labor. The economy might evolve in ways that dramatically reduce the overall demand for human labor, even if some niche areas continue to value human involvement highly.
Moreover, the nature of the sapien premium itself might change over time. As AI becomes more sophisticated in mimicking human interaction and emotional intelligence, the areas where humans are preferred might shift or narrow. Additionally, generational changes in attitudes towards human versus AI interaction could alter what people are willing to pay a premium for.
While the concept of the sapien premium provides a useful framework for identifying potential areas of persistent human employment in a highly automated economy, it doesn’t resolve the fundamental uncertainty about the future of work and the economy as a whole. As we continue to navigate the transition towards increasing automation, it will be crucial to monitor these trends, adapt our economic and social systems, and perhaps rethink our notions of work, value, and human contribution beyond traditional employment.
Macroeconomic View of AI Impact
While the “Better, Faster, Cheaper, Safer” model provides valuable insights into the displacement of human labor by AI and automation, it has significant limitations when considering the broader economic impact of AI. This model primarily focuses on the micro-level dynamics of labor replacement, but it fails to capture the complex, interconnected nature of the global economy and the potential ripple effects of widespread AI adoption.
One key limitation is that the model assumes a linear progression of AI capabilities across all sectors, which may not reflect reality. AI development and adoption are likely to be uneven, with some sectors experiencing rapid transformation while others lag behind due to technical, regulatory, or social factors. Additionally, the model doesn’t account for the potential emergence of new industries or economic activities that AI might enable, which could create new forms of human employment or economic value.
To address these limitations and gain a more comprehensive understanding of AI’s economic impact, we need to adopt a broader perspective. The Theory of Constraints (TOC) offers a promising framework for this analysis, allowing us to examine the economy as a system of interconnected processes, each with its own limiting factors or “constraints.”
Just as the PLE model distills the essential factor in future human employment (what humans will prefer and pay a premium for), the TOC approach allows us to identify the critical constraints in various economic sectors and processes. This parallel is important because in both cases, we’re focusing on the most essential elements rather than trying to predict the full scope of future AI capabilities, which is inherently challenging and speculative.
By applying TOC to economic sectors and macroeconomic processes, we can identify where the true bottlenecks or constraints exist in our current economic system. These constraints might not always be related to human labor or cognitive capabilities. They could involve physical resources, energy limitations, logistical challenges, regulatory hurdles, or even social and cultural factors that resist rapid change.
The Theory of Constraints teaches us that ‘any improvement made anywhere besides the bottleneck is an illusion.’ When applied to AI’s economic impact, this principle reveals that the true transformative potential of AI lies not in its broad capabilities, but in its ability to alleviate specific, critical constraints in economic processes. Understanding these bottlenecks is key to predicting where AI will have the most significant and meaningful impact on productivity and progress.
For example, in the energy sector, the constraint might not be human labor in power plant operation (which could potentially be automated), but rather the physical limitations of energy generation and storage technologies. In healthcare, while AI might excel at diagnosis and treatment planning, the constraint could be the physical capacity of hospitals or the supply of certain medications.
This TOC-based approach allows us to make more nuanced predictions about where AI is likely to have the most significant economic impact. It suggests that AI will be most transformative in areas where it can alleviate key constraints, rather than simply in areas where it can perform tasks better than humans. This might lead to surprising conclusions about which sectors of the economy are most likely to be revolutionized by AI, and which might remain relatively stable despite increasing automation.
Moreover, this approach helps us understand potential cascading effects throughout the economy. As AI alleviates one constraint, it may shift the bottleneck elsewhere in the system, potentially creating new opportunities for both AI application and human labor in unexpected areas.
By combining the insights from the PLE model’s focus on human preferences with a TOC-based analysis of economic constraints, we can develop a more robust framework for predicting and understanding the broader economic implications of AI. This holistic view is crucial for policymakers, business leaders, and society as a whole as we navigate the complex transition to an AI-augmented economy.
This approach also highlights the need for interdisciplinary collaboration in studying and preparing for AI’s economic impact. Economists, technologists, sociologists, and policymakers will need to work together to understand and manage the complex interplay of technological capabilities, economic constraints, and human preferences that will shape our economic future.
While the PLE model provides valuable insights into the future of human labor, a TOC-based approach to analyzing AI’s broader economic impact offers a complementary and more comprehensive framework. By focusing on systemic constraints rather than just human vs. machine capabilities, we can better anticipate where AI is likely to be most transformative and how these changes might ripple through the global economy.
Types of Constraints: Human vs Non-Human
In examining the economic impact of AI through the lens of the Theory of Constraints, we must first identify and categorize the primary constraints that limit throughput in various economic processes. These constraints can be broadly divided into human-centric and non-human categories, each playing a crucial role in shaping the potential for AI to transform different sectors of the economy.
Human-centric constraints are those directly tied to the limitations of human capabilities. These include cognitive processing speed, which can create bottlenecks in decision-making processes, particularly in management and strategic planning. Physical endurance and precision represent another key constraint, especially in manufacturing and construction where human fatigue can limit productivity and accuracy. Specialized knowledge and expertise, while valuable, can also act as a constraint when the scarcity of such expertise slows down processes in fields like research, software development, and financial analysis. Emotional intelligence and interpersonal skills, while uniquely human, can be constraining factors in scenarios where human interaction is critical but limited in scale.
On the other hand, non-human constraints are those imposed by factors external to human capabilities. Natural cycles, such as seasons and the day/night rhythm, can significantly impact industries like agriculture, energy production, and certain service sectors. Resource availability, including raw materials and energy, often acts as a fundamental constraint on production and economic growth. Physical laws, particularly thermodynamics, set hard limits on efficiency and productivity in many industrial processes. Regulatory and legal constraints, while human-made, operate as external factors that can significantly limit the speed of innovation and implementation of new technologies.
Archimedes said, ‘Give me a lever and a place to stand, and I shall move the earth.’ In the context of AI’s economic impact, the lever is AI itself, but the place to stand—the fulcrum—represents the constraints we seek to overcome. While AI offers unprecedented leverage in addressing human-centric constraints, many non-human constraints—like the laws of physics or the cycles of nature—remain unmoved, regardless of how powerful our AI ‘lever’ becomes. The key to unlocking AI’s true potential lies in identifying where we can apply this lever most effectively.
When we analyze these constraints in the context of AI adoption, we can begin to predict where AI might have the most significant impact. For instance, in areas where cognitive processing speed is the primary constraint, such as complex data analysis or rapid decision-making in financial markets, AI has already shown tremendous potential to alleviate this bottleneck. Similarly, in scenarios where physical endurance is the main limiting factor, robotics and automation guided by AI can dramatically increase productivity.
However, it’s crucial to note that not all constraints are equally susceptible to AI solutions. Emotional intelligence and interpersonal skills, for example, remain areas where human capabilities are often preferred, aligning with the concept of the “sapien premium” discussed earlier. Moreover, many non-human constraints, such as resource availability or fundamental physical laws, may not be directly addressable by AI, though AI might help in optimizing processes within these constraints.
Understanding these constraints and their nature allows us to make more nuanced predictions about the economic impact of AI. Rather than assuming a uniform transformation across all sectors, we can anticipate that AI will be most transformative where it can effectively alleviate key constraints, particularly those related to cognitive processing, data analysis, and certain forms of physical labor. However, in areas where constraints are primarily non-human or deeply tied to uniquely human attributes, the impact of AI may be less direct or slower to materialize.
This constrained-based analysis provides a valuable framework for businesses, policymakers, and researchers to prioritize AI development and adoption efforts. By focusing on areas where AI can most effectively address critical constraints, we can potentially accelerate economic growth and productivity while also anticipating the sectors where human labor is likely to remain valuable and necessary in the foreseeable future.
Predicting Economic Impact by Magnitude
In analyzing the potential impact of AI across various economic sectors through the lens of the Theory of Constraints, we can categorize industries into high, medium, and low impact areas based on the nature of their primary constraints and AI’s ability to address them.
High Impact Areas:
Financial and capital markets are likely to see transformative effects from AI. The primary constraints in these sectors often revolve around information processing speed, decision-making capacity, and the ability to analyze vast amounts of data quickly. AI excels at these tasks, potentially alleviating major bottlenecks. For instance, AI-driven high-frequency trading algorithms can execute trades at speeds impossible for humans, while machine learning models can analyze market trends and risk factors more comprehensively than human analysts. The constraint here is primarily cognitive, an area where AI has shown remarkable capabilities.
Communications and entertainment industries are also poised for high AI impact. In these sectors, the constraints often involve creative capacity, language barriers, and production speed. AI’s advancements in natural language processing can revolutionize translation services, breaking down language constraints in global communication. In entertainment, AI-generated content, from images to videos, can dramatically increase production capacity and speed, addressing constraints in creative output. The ability of AI to generate, manipulate, and optimize media content at scale could fundamentally reshape these industries. For instance, in just a few minutes, and for just a few dollars, I can create fully AI-generated videos. They aren’t great yet, but give it another year or two.
Medium Impact Areas:
Logistics presents an interesting case where AI is likely to have a significant but not transformative impact. While AI can greatly improve monitoring, planning, and optimization of supply chains, the sector remains constrained by physical and temporal factors that AI cannot directly overcome. For example, AI can optimize shipping routes and warehouse operations, but it can’t change the physical speed of trucks or ships, or alter the fundamental constraints of time and distance in moving goods.
Basic and fundamental science may see substantial benefits from AI, but other constraints limit its transformative potential. AI can certainly aid in data analysis, hypothesis generation, and resource allocation for research. However, many scientific endeavors are constrained not by cognitive limitations but by factors such as the time required to conduct experiments, the massive scale and cost of certain scientific instruments (like particle accelerators), and the limitations of government or institutional funding. While AI can optimize within these constraints, it cannot easily overcome them.
Healthcare is another sector where AI’s impact may be significant but not all-encompassing. AI has shown great promise in areas like diagnostic imaging, drug discovery, and personalized treatment planning, potentially alleviating constraints related to expert knowledge and decision-making. However, many healthcare constraints are physical or systemic—the number of hospital beds, the time required for clinical trials, or the cost of producing medications. AI can optimize within these constraints but cannot directly address them.
Limited Impact Areas:
Heavy industry and agriculture are sectors where AI’s impact may be more limited, primarily due to the nature of their constraints. In heavy industry, such as steel production or mining, the primary constraints often relate to fundamental physical and chemical processes, energy requirements, and the availability of raw materials. While AI can optimize these processes to some degree, it cannot overcome the basic thermodynamic and energetic constraints that dominate these industries.
Similarly, in agriculture, while AI can aid in optimizing crop yields, predicting weather patterns, and managing resources, the sector remains heavily constrained by factors such as growing seasons, climate conditions, and the biological processes of plant growth. These constraints are largely indifferent to human or AI intervention beyond a certain point.
The production of basic goods like timber or cement also falls into this category. While AI can optimize supply chains and manufacturing processes, the fundamental constraints here are often related to resource availability, physical processing times, and energy requirements, areas where AI’s impact is inherently limited.
Final Thoughts
Post-Labor Economics (PLE) posits that machines will increasingly replace human labor as the primary driver of economic progress, effectively decoupling productivity from human constraints. As AI and automation surpass human capabilities in being ‘better, faster, cheaper, and safer,’ this shift becomes economically inevitable. While ‘sapien premium’ jobs—those uniquely valued for human touch—may persist, the broader implications for our current social contract around wages and labor remain uncertain.
In light of this transformative potential, we advocate for a paradigm shift in economic and business analysis. Leaders across all sectors—from economists and politicians to business executives and financial analysts—should adopt a constraints-based view of the economy. This approach provides a robust framework for anticipating and understanding AI’s varying impacts across different industries and processes.
Our analysis reveals that AI’s impact will be most profound in sectors where human cognitive abilities have been the primary constraint, such as financial services, data analysis, and certain aspects of healthcare. Conversely, industries bound by physical or natural constraints—like agriculture, heavy manufacturing, and resource extraction—may see more limited direct impacts from AI.
As we navigate this economic revolution, understanding and focusing on key constraints will be crucial. By identifying where AI can most effectively alleviate bottlenecks, we can better predict, prepare for, and harness the transformative power of this technology. The future of our economy will be shaped not just by what AI can do, but by how strategically we apply it to our most critical constraints.
All Sectors, Estimated AI Impact (Based upon Human Constraints)
Finally, here’s a breakdown of all major sectors based on this theory of constraints approach:
Financial Services and Banking: High AI impact. AI can significantly enhance data processing, risk assessment, and decision-making speed.
Healthcare and Pharmaceuticals: High AI impact. AI can revolutionize diagnostics, drug discovery, and personalized treatment planning.
Information Technology and Telecommunications: High AI impact. AI can optimize network performance, enhance cybersecurity, and accelerate software development.
Manufacturing: Medium AI impact. While AI can improve quality control and supply chain management, physical production constraints remain.
Retail and E-commerce: High AI impact. AI can personalize customer experiences, optimize inventory, and enhance logistics.
Energy and Utilities: Low AI impact. AI can improve grid management and demand forecasting, but physical infrastructure limits remain.
Real Estate and Construction: Low AI impact. AI can enhance design and project management, but physical construction processes limit full automation.
Agriculture and Food Production: Medium AI impact. AI can optimize crop management and supply chains, but biological and climate constraints persist.
Education and Training: High AI impact. AI can personalize learning experiences and automate administrative tasks.
Transportation and Logistics: High AI impact. AI can optimize routes, enhance autonomous vehicle technology, and improve traffic management.
Media and Entertainment: High AI impact. AI can generate content, personalize recommendations, and streamline production processes.
Government and Public Services: Medium AI impact. AI can improve administrative efficiency, but policy-making and public interaction may require human involvement.
Tourism and Hospitality: Low AI impact. AI can enhance booking systems and personalization, but the human touch remains crucial in service delivery.
Automotive Industry: Medium AI impact. AI is central to developing autonomous vehicles and optimizing manufacturing processes. Capital goods remain expensive.
Aerospace and Defense: Medium AI impact. AI can significantly enhance design processes, simulation capabilities, and autonomous systems. Heavy industries remain expensive.
Mining and Raw Materials: Low AI impact. While AI can improve prospecting and operations, physical and geological constraints dominate. Thermodynamics is a bitch.
Professional Services (Legal, Consulting, etc.): High AI impact. AI can automate research, enhance decision-making, and provide advanced analytics.
Waste Management and Recycling: Low AI impact. AI can improve sorting and processing efficiency, but physical handling of waste remains a constraint.
Some Potential “Forever Jobs”
Here are some jobs that I suspect will be durable for the foreseeable future due to the sapien premium. Again, keep in mind that in many cases a highly dexterous and intelligent robot could almost certainly do these jobs better than a human, so it’s not about capability as much as it is about human preference.
Tour guides
Massage therapists
Childcare workers
Musicians
Athletes
Therapists and counselors
Artisanal craftspeople
Personal trainers
Spiritual leaders and clergy
Actors and performers
Chefs and culinary artists
Life coaches
Midwives and doulas
Sommeliers
Government officials (politicians, judges)
Dance instructors
Motivational speakers
Yoga and meditation instructors
High-end hairstylists and beauticians
Personalized fashion designers
Nice analysis David! I ran a similar analysis using speculative design in my last YT video https://youtu.be/1K1FmEL7u2I
Imo, the future is ubi and meaningful work followed by soma and transhumanism.