Pieter Van Schalkwyk

CEO at XMPRO

How the Farmer's Journey from Ox to Orchestrator Reveals the True Pattern of Technological Transformation

The question I encounter most frequently when discussing AI and Multi-Agent Generative Systems is not a technical one. It is existential: "Will AI replace jobs and leave people without work?"

This question deserves a serious answer. But before I provide one, I want to challenge the framing itself. The fear of technological unemployment assumes a zero-sum relationship between machines and human labor, that every task automated represents a human displaced. History reveals something far more interesting: technological transformation does not eliminate work. It disaggregates work into specialized functions distributed across an expanding value chain.

To understand what MAGS will do to human work, we need to look not at what AI can automate, but at what previous transformations actually created. The story of the farmer provides our clearest lens.

The Quality of Life Paradox: What GDP Fails to Measure

In The Second Machine Age, Erik Brynjolfsson and Andrew McAfee make a crucial observation that reframes how we should think about technological impact. They argue that conventional measures like GDP systematically underestimate the improvements in quality of life created by each industrial revolution.

The reason is simple: GDP measures transactions, not value. When a farmer transitioned from ox-drawn plowing to tractor cultivation, GDP captured the purchase of the tractor and the increased crop yield. It failed to capture:

• The reduction in physical toil
• The liberation from dawn-to-dusk manual labor
• The years added to working lives
• The mental bandwidth freed for strategic thinking
• The new possibilities that emerged when human effort could be directed toward higher-order activities

Brynjolfsson and McAfee call this the "bounty" of technological change: broad gains in living standards that manifest in consumer surplus, expanded leisure, improved health, and longer life expectancy. Each major technological wave has delivered this bounty. The question is not whether it arrives, but how it is distributed.

This distinction matters for understanding MAGS. If we measure only the tasks that agents automate, we will conclude that work is disappearing. If we measure the value created across the entire ecosystem that MAGS enables, we will see something very different: the emergence of entirely new categories of human contribution.

The Farmer's Journey: A Parable of Transformation

Consider the farmer 150 years ago, working behind an ox with a single-blade plow. This farmer possessed an integration of skills that we rarely see today:

• Physical strength to guide the plow
• Intimate knowledge of soil conditions
• Intuitive understanding of weather patterns
• Mechanical ability to maintain simple tools
• Animal husbandry to work effectively with the ox
• Strategic judgment about when and where to plant

The farmer was a one-person operation encompassing every function from planning to execution. This farmer fed their family from a small homestead.

The First Disaggregation: Mechanization

The combustion engine transformed this picture entirely. The tractor replaced the ox, multiplying the farmer's productive capacity by an order of magnitude. But something else happened that receives less attention: the integrated skillset of the ox-farmer began to disaggregate into specialized functions.

The farmer could no longer repair their own equipment. The mechanical complexity of tractors required dedicated mechanics. Those mechanics needed parts, which required manufacturing facilities. Those facilities needed engineers, logistics networks, distribution channels, and financial systems to fund capital equipment purchases.

The farmer's immediate skill requirements shifted from physical stamina and animal management to mechanical operation and basic maintenance. But the total ecosystem of skills required to enable that farmer's work expanded dramatically. Where one farmer once needed only their own hands and the cooperation of an ox, the tractor-era farmer now depended on a complex value chain of specialists.

Did the tractor eliminate work? In one sense, yes: it eliminated much of the backbreaking physical labor. But in a more important sense, it created work by enabling specialization that would have been impossible in the ox-era economy. The U.S. farm workforce shrank by over 70% from 1950 to 1990, yet total employment grew as value shifted toward upstream activities (manufacturing, engineering, finance) and downstream operations (logistics, processing, retail, food services).

This is the pattern: technology replaces integrated labor with disaggregated specialized labor distributed across an expanding value chain.

The Second Disaggregation: Remote Operation

Today's farmer operates from a control room. GPS-guided tractors with auto-steering, sensor networks, drone surveillance, and cloud-based analytics have transferred yet another layer of cognitive work from the farmer to the system.

• The farmer no longer needs to navigate the field. Satellites handle that.
• The farmer no longer needs to estimate soil moisture by touch. Sensors provide continuous measurements.
• The farmer no longer needs to scout for pest infestations by walking the rows. Multispectral imagery identifies problems before they become visible to the human eye.

But look at what this further disaggregation has created:

• Satellite systems require aerospace engineers, ground station operators, software developers, and telecommunications specialists.
• Sensor networks require IoT engineers, embedded systems designers, and precision calibration technicians.
• Data analytics platforms require data scientists, machine learning engineers, and UX designers.
• Precision agriculture software requires agronomists who can translate farming knowledge into algorithmic logic.

The farmer's immediate role has shifted again: from machine operator to system monitor and strategic decision-maker. But the total ecosystem of human expertise required to enable that farmer's work has expanded by another order of magnitude. The single farmer feeding a village has become a corporate data architect feeding cities across the world, supported by specialists who never set foot on a farm.

The Third Disaggregation: Autonomous Orchestration

MAGS represents the next phase of this progression. In the autonomous farm of the near future, fleets of AI-coordinated robots handle planting, spraying, harvesting, and real-time optimization. The farmer issues high-level directives ("maximize yield while maintaining soil health metrics within these parameters") and the multi-agent system decomposes this objective into thousands of coordinated actions executed by specialized agents.

This is already happening. John Deere has unveiled fully autonomous tractors with 360-degree cameras and AI navigation. Companies like Carbon Robotics and Monarch Tractor are deploying autonomous systems with pay-per-hour models that remove capital barriers. Multi-agent platforms now coordinate teams of robots in complex field operations.

What cognitive work remains for the farmer? The answer reveals the true pattern of MAGS-enabled transformation: the farmer becomes the architect of the utility function itself.

The Utility Function: Where Human Value Concentrates

AI systems optimize toward whatever objective function they are given. They do so with precision, speed, and tireless consistency that humans cannot match. But they cannot choose what to optimize for. They cannot balance the incommensurable values that define any real-world enterprise: profit versus sustainability, efficiency versus resilience, short-term yield versus long-term soil health, financial returns versus community impact.

This is where human cognitive labor concentrates in the MAGS era: the design, validation, and continuous refinement of the objectives that guide autonomous systems. The farmer becomes the philosopher of the farm, responsible for answering questions that no algorithm can resolve:

• What does "optimal" mean for this operation?
• How do we weight competing objectives?
• What tradeoffs are acceptable, and which cross ethical boundaries?
• How do we adapt our goals as conditions change?

This is not a diminished role. It is an elevated one. The farmer who once walked behind an ox now shapes the intelligence that directs an autonomous fleet. The cognitive labor has not disappeared. It has migrated to a higher level of abstraction where human judgment matters most.

The Emergence of New Specialized Roles

The farmer-as-orchestrator is only one node in an expanding ecosystem. The disaggregation pattern that characterized mechanization and remote operation accelerates in the MAGS era. Consider the specialized roles already emerging:

Platform Orchestration Specialists manage the alignment and integration of diverse technological systems, from drones and sensors to autonomous fleets and cloud analytics. The complexity of connecting these tools requires dedicated expertise in selecting, configuring, and optimizing the technical stack.

Data Scientists and AI Agronomists interpret the flood of data generated by autonomous systems, building predictive models and translating AI insights into strategic recommendations. The farm's primary input is now a continuous stream of real-time data. Specialists are needed to make sense of it.

Precision Calibration Technicians ensure that sensors and actuators operate within required tolerances. The accuracy of autonomous systems depends entirely on the precision of their inputs. This specialized skill becomes critical infrastructure.

IoT Network Engineers establish and secure the wireless communications that enable machine-to-cloud coordination. Every autonomous system depends on reliable connectivity. Dedicated expertise ensures it functions.

Specialized Field Technicians (often employed by equipment manufacturers or Robot-as-a-Service providers rather than individual farms) maintain the complex machinery that makes autonomous operation possible. Their skills combine mechanical expertise with electronics, embedded systems, and predictive maintenance analytics.

Risk Management Analysts help farmers navigate the financial complexity of modern agriculture: commodity price hedging, futures and options, insurance products, and capital allocation. As farm operations scale, financial sophistication becomes as important as agronomic knowledge.

System Agronomists work within cross-functional teams to ensure that AI systems incorporate essential agronomic knowledge. They bridge the gap between what data scientists can build and what farmers actually need.

Each of these roles represents cognitive labor that did not exist in the ox-and-plow era. They exist because MAGS enables operations at a scale and complexity that demands specialization. The single integrated skillset of the traditional farmer has disaggregated into an ecosystem of expertise, each specialization creating value that contributes to the whole.

The Pattern Applied: A Universal Truth

The farmer analogy is not just illustrative. It reveals a universal pattern that applies across every domain where MAGS will operate.

Manufacturing

The craftsman of the pre-industrial era possessed integrated skills: design, material selection, fabrication, quality control, and customer relationships resided in a single person. Industrialization disaggregated these functions into specialized roles distributed across factory floors, engineering offices, and sales departments.

Automation further disaggregated cognitive work: production planning, process control, quality assurance, and maintenance became specialized functions requiring dedicated expertise. MAGS accelerates this again, creating new roles:

• M-shaped supervisors who combine breadth across multiple domains with depth in agent orchestration (I wrote about these roles in "The Industrial Agentic Organization"
• T-shaped domain experts who provide deep specialized knowledge and teach agents how to analyze situations
• Agent training specialists who develop and refine the knowledge bases and learning mechanisms that enable agent teams to improve
• Integration architects who design connections between agent systems, existing automation, and human workflows
• Governance specialists who establish oversight frameworks and compliance mechanisms

Scientific Research

Perhaps the most striking example of disaggregation in action is the emergence of what Peter H. Diamandis and his Moonshots podcast team call "lights-out labs": fully autonomous research facilities where AI formulates hypotheses, designs experiments, and directs robotic systems to execute research 24/7 without human intervention.

Google DeepMind is building an autonomous materials science laboratory in the UK. A company called Laya, spun out of MIT and Harvard, is applying similar approaches to biological sciences. The hosts describe this as "the biggest breakthrough in scientific progress since the scientific method was invented."

But notice what is happening: the integrated role of the scientist (hypothesis generation, experiment design, lab work, analysis, publication) is disaggregating into specialized functions.

• AI handles hypothesis generation and experiment design
• Robots execute the physical work
• Humans provide oversight, direction-setting, and ethical boundaries
• Specialists validate results and interpret meaning
• Others connect discoveries to practical applications

The lights-out lab does not eliminate scientific work. It disaggregates it, creating new specialized roles that did not exist when a single scientist performed every function from conception to publication. These labs are essentially "data mining nature," generating new knowledge through experiments at 1,000 to 10,000 times the speed of human-led research. This acceleration does not reduce the need for human scientific contribution. It transforms where that contribution occurs.

The Universal Pattern

Whether we examine farming, manufacturing, or frontier scientific research, the pattern holds: technology does not eliminate work. It disaggregates integrated labor into specialized functions distributed across an expanding value chain. Each phase of disaggregation creates new categories of human contribution that were inconceivable in the previous era.

The Acceleration: Why This Time Feels Different

Understanding the pattern is essential. But we must also acknowledge that the current transformation is unfolding faster than any before it.

OpenAI's GDP-val benchmark, designed to measure AI's ability to automate knowledge work across 44 occupations and 1,320 specialized tasks, tells the story clearly. GPT 5.2 achieved a 70.9% score, meaning that in 71% of comparisons between a human performing knowledge work and the machine, the AI did a better job at more than 11 times the speed and less than 1% of the cost.

This is not a distant future projection. This is now.

Dr. Alex Wissner-Gross , one of the The Moonshots podcast panel, predicts that 2026 will mark "the biggest collapse of the corporate world in the history of business." What he is describing is not the end of work but the rapid disaggregation of integrated corporate functions into new specialized configurations. The companies that fail will be those that cannot adapt their organizational structure to the new pattern. The companies that thrive will be those that understand disaggregation as opportunity rather than threat.

Here is what the benchmarks do not measure: the new work that emerges when knowledge tasks are automated. When the farmer stopped walking behind the ox, there was no benchmark measuring the demand for tractor mechanics. When the factory installed assembly lines, there was no metric for industrial engineers. The GDP-val benchmark captures what is being automated. It says nothing about what is being created.

The disaggregation pattern tells us that new specialized roles will emerge. The acceleration tells us they will emerge faster than in any previous transformation. This combination creates the opportunity and the challenge we must address.

The Real Risk: Who Benefits from Disaggregation?

If the pattern holds, and MAGS creates more work through disaggregation rather than eliminating work through automation, what should we actually be concerned about?

The risk is not that jobs will disappear. The risk is that the benefits of disaggregation will be captured by those who can access augmentation tools, while others are left behind. The pattern creates new specialized roles, but those roles require new capabilities. Workers who develop those capabilities thrive. Workers who cannot access the tools or training fall further behind with each cycle.

Michael Carroll illustrates this with a control room scenario on Mimi Brooks' Bold Agendas podcast. Consider two operators facing the same situation: a dashboard flashing red, indicating a process anomaly.

The non-augmented operator:

• Drills down through multiple screens to diagnose the problem
• Figures out the solution through troubleshooting
• Puts it through work processes and supervisors
• Creates value slowly, limited by the latency of human cognition and organizational coordination

The agentic augmented operator:

• Receives an instant notification with context: "Oil temperature elevated, bearing will fail in 8 hours based on thermal degradation model"
• The agent asks: "Would you like me to increase coolant flow now and schedule preventive maintenance for the next planned downtime?"
• Creates value immediately, enabled by the agent's continuous analysis and predictive capability

Both operators have jobs. But they create value at dramatically different rates. The augmented worker compounds advantages with every interaction. The non-augmented worker falls further behind.

As Mike observes: "We don't replace people. We make them more important. We make them more important because they create value at a rate that is so much faster... because we get rid of the latency in decision making."

This is the dynamic that demands attention: not job elimination, but capability divergence. The disaggregation pattern creates new work, but unequal access to augmentation determines who can perform it. Workers with AI tools become exponentially more valuable.

Workers without them become comparatively less competitive, not because they lack ability, but because they lack the tools that multiply their ability.

The Human Challenge: Culture, Skills, and Access

If the real risk is unequal access to augmentation rather than job elimination, then our response must address three related challenges: organizational culture, individual skills, and broad access to tools.

The Cultural Barrier

The Moonshots analysis offers a direct insight: the barrier to transformation is not primarily technical or even economic. It is cultural.

As the hosts observe: "I don't think this is a skills issue. This is a cultural problem... you need a mindset shift at scale."

Organizations are not failing to adopt MAGS because they lack technical capability. They are failing because their organizational structures, incentive systems, career paths, and mental models were designed for integrated labor, not disaggregated specialization. The corporate hierarchy that made sense when managers coordinated human workers performing integrated tasks becomes friction when multi-agent systems can handle coordination automatically.

The transformation requires change at every level:

• Executive leadership must understand what MAGS enables and commit to organizational redesign
• Middle management must adapt to orchestrating agent teams rather than human teams
• Individual contributors must develop skills in human-agent collaboration
• Boards must support the investment required for transformation

The companies that will thrive are those that recognize this as an organizational design challenge, not just a technology adoption challenge. MAGS does not fit into existing structures. It requires new structures designed around the disaggregation pattern.

The Skills Transition

The question is not whether MAGS will create new work (it will) but whether workers will have the skills to perform it. The skills required for the ox-era farmer were different from those required for the tractor-era farmer, which were different from those required for the remote-operation-era farmer, which are different from those required for the MAGS-era orchestrator.

Each transition required learning. Each transition created winners and losers based on who adapted fastest. The MAGS transition will be no different, except that the pace of change is faster and the scope is broader.

One of the Moonshots hosts described a client who faced laying off 1,000 people after an AI transformation sprint. Their solution:

• Provided one year of income support to displaced workers
• Gave them time to find new work aligned with the transformed economy
• Offered to rehire those who could not transition

The program was "incredibly successful." It points toward a model where organizations take responsibility for bridging the gap between old roles and new ones.

The Access Imperative

Investment in reskilling is not charity. It is economic necessity. An economy with vast unmet demand for MAGS orchestrators and agent trainers, combined with vast unemployment among workers with obsolete integrated-labor skills, is an economy that has failed to manage the transition.

The solution is not to slow the adoption of MAGS. The solution is to democratize access to augmentation, to ensure that the tools and training that enable human-agent collaboration are available broadly rather than concentrated in elite operations. The goal must be to move workers from old roles to new roles faster than the old roles disappear.

The Bounty Revisited: What We Stand to Gain

Return to Brynjolfsson and McAfee's insight: GDP fails to capture the quality of life improvements that technology creates. The same limitation applies to employment statistics. Counting jobs tells us nothing about the nature of work, the satisfaction it provides, or the human flourishing it enables.

The farmer who once spent twelve-hour days walking behind an ox now monitors dashboards and makes strategic decisions. The factory worker who once performed repetitive manual tasks now supervises autonomous systems and focuses on exception handling and continuous improvement. The scientist who once spent years on a single experiment now directs AI systems that explore thousands of hypotheses simultaneously.

The cognitive burden shifts from execution to orchestration, from routine to judgment, from physical toil to strategic thinking. This is the bounty that MAGS offers: not just efficiency gains measured in output, but human flourishing measured in the quality of work itself.

The drudgery that characterized traditional labor (the physical danger, the mind-numbing repetition, the exhaustion that left no energy for anything beyond survival) can be delegated to systems that do not suffer from it. What remains for humans is the work that humans actually find meaningful: solving problems, making decisions, creating value, building relationships, exercising judgment in ambiguous situations.

MAGS does not eliminate this work. It clears away the routine tasks that prevented humans from focusing on it.

Shaping the Disaggregation

The transformation underway follows a pattern that has repeated through every major technological transition: integrated labor disaggregates into specialized functions distributed across an expanding value chain.

• The ox-farmer contained every function.
• The tractor-farmer depended on mechanics, manufacturers, and financial systems.
• The remote-operator farmer relies on satellite engineers, IoT specialists, and data scientists.
• The MAGS-era farmer will orchestrate autonomous systems while an ecosystem of experts maintains, improves, and governs the technology.

The same pattern applies to manufacturing, to scientific research, to every domain where MAGS operates. Total work expands even as individual task categories are automated. Human cognitive labor migrates toward higher-order activities where human judgment matters most.

The fear that AI will replace jobs misframes the question. The real questions are:

• How fast can workers develop the skills that MAGS-era work demands?
• How broadly will augmentation tools be distributed?
• How can organizations ensure that the benefits of disaggregation reach everyone rather than concentrating among those who already have advantages?

These are questions of policy, investment, and intention, not questions of technological inevitability. MAGS will create more work, not less. But only deliberate effort will ensure that this work is accessible to all who are willing to adapt.

The farmer's journey from ox to orchestrator shows us the pattern. Whether we shape it to benefit everyone depends on the choices we make now.

Pieter van Schalkwyk is the CEO of XMPro, specializing in industrial AI agent orchestration and governance. Drawing on 30+ years of experience in industrial automation, he helps organizations implement practical AI solutions that deliver measurable business outcomes while ensuring responsible AI deployment at scale.

About XMPro: We help industrial companies automate complex operational decisions. Our cognitive agents learn from your experts and keep improving, ensuring consistent operations even as your workforce changes.

Our GitHub Repo has more technical information. You can also contact me or Gavin Green for more information.

Read more on MAGS at The Digital Engineer