The global corporate landscape is currently navigating a fundamental shift in how value is defined, generated, and sustained through the integration of artificial intelligence. While the initial wave of AI adoption was characterized by simple automation and cost-cutting measures, a more complex framework is emerging that distinguishes between three distinct types of organizational value. The first and most prevalent is subtractive value, which focuses on removing costs and inefficiencies from core business processes. The second is additive value, a strategy that utilizes AI to generate new revenue streams, revive dormant client relationships, and unlock the latent potential of a company’s workforce. The third, and arguably most significant, is transformative or non-linear value—a concept explored in the recent book Autonomous—which involves using AI to fundamentally alter the nature of entire industries and redefine the competitive landscape.
As organizations progress through their AI transformation journeys, the ability to capture these varying types of value depends heavily on their maturity level. Data from recent industry reports, including McKinsey & Company’s "The State of AI in 2024," suggests that while nearly 72% of organizations have adopted AI in at least one business function, the majority remain stalled in the subtractive phase. These companies often view AI as a traditional technological implementation rather than a comprehensive organizational metamorphosis. However, the emergence of generative AI and, more recently, agentic AI—systems capable of autonomous goal-setting and execution—is forcing a realization that the transition is not linear. It is a paradigm shift from a "human organization" where people use technology as a tool, to an "autonomous organization" where AI and humans function as colleagues, with work assigned to the intelligence best suited for the task.
The Automotive Parallel: Understanding the Three Phases of Autonomy
To understand the trajectory of this transformation, business leaders are increasingly looking toward the automotive industry, which has spent over a decade grappling with the transition from human control to machine intelligence. In September 2025, during an address to CarNewsChina, Candice Yuan, the Director of Autonomous Driving at XPeng, provided a crucial insight into this evolution. Yuan noted that developing advanced driver-assist systems (ADAS) for passenger vehicles is often significantly more challenging than building fully driverless systems. This observation serves as a cornerstone for understanding why business transformation is proving so difficult for the modern enterprise.
The journey toward autonomy, whether in a vehicle or a corporation, can be categorized into three distinct phases: Hands On, Hands Ready, and Hands Off.
Phase 1: Hands On (The Human-Led Era)
In the "Hands On" phase, the human is the primary driver. In an automotive context, this involves traditional driving where the human manages steering, braking, and acceleration. In a business context, this represents the traditional "as-is" state. Humans perform the vast majority of cognitive and operational tasks, using software like Excel, CRM systems, or ERP platforms as passive tools to record or organize their work. AI in this phase is used primarily for basic data analysis or "subtractive" cost-saving measures, such as automating invoice processing.
Phase 2: Hands Ready (The Hybrid Bottleneck)
The "Hands Ready" phase is the most complex and dangerous stage of transformation. In a vehicle, this is equivalent to Level 3 autonomy, where the car can drive itself under certain conditions but requires the human to be ready to take control at a moment’s notice. Yuan’s insight highlights that this is the hardest phase to engineer because the system must account for two "masters" simultaneously. It must manage the machine’s algorithmic logic while accounting for human unpredictability, emotional overrides, and varying reaction times.
In the corporate world, most "AI-forward" companies are currently stuck in this phase. They are attempting to run hybrid workflows where AI generates content or suggests decisions, but humans must review, verify, and approve every step. This creates a friction-filled environment where the "human-in-the-loop" becomes a bottleneck. The organizational structure must manage the seamless transition of control between human managers and AI agents, often leading to confusion over accountability and decision rights.
Phase 3: Hands Off (The Autonomous Organization)
The "Hands Off" phase represents the "to-be" state of the autonomous organization. In a vehicle, this is Level 4 or 5 autonomy, where the system is optimized for a single mode of operation: machine control. In business, this is the realization of non-linear value. Here, AI agents manage entire workflows—from supply chain adjustments based on predictive weather patterns to real-time dynamic pricing—without requiring constant human intervention. The human role shifts from operational "driving" to strategic "mission control," focusing on governance, ethics, and long-term vision.
Why Business Autonomy Surpasses Technical Autonomy in Complexity
While the automotive industry provides a useful roadmap, the transition for a business is exponentially more complex. A vehicle operates with three primary actuators: the accelerator, the brake, and the steering wheel. In contrast, a modern enterprise operates through dozens of "actuators" that are often interconnected in non-linear ways. These include:
- Financial Actuators: Pricing strategies, capital allocation, and resource budgeting.
- Market Actuators: Product mix, sales campaigns, and marketing spend.
- Operational Actuators: Decision rights, approval authorities, and supply chain logistics.
- Cultural Actuators: Organizational identity, performance measurement, and employee morale.
Each of these levers transforms at a different rate. For instance, an organization might successfully implement AI-driven dynamic pricing (a financial actuator) while still relying on manual, human-led performance reviews (a cultural actuator). This discrepancy creates "organizational drag," where the speed of AI-driven processes is hampered by the slower pace of human-led legacy systems.
Data and Industry Response: The Push Toward Agentic AI
The transition toward the "Hands Off" or autonomous state is being accelerated by the rise of agentic AI. Unlike standard Large Language Models (LLMs) that respond to prompts, agentic systems are designed to pursue goals. According to Gartner, by 2028, at least 15% of daily work decisions will be made autonomously by AI agents.
Industry reactions to this shift have been polarized. In recent earnings calls, tech leaders such as Microsoft’s Satya Nadella and Salesforce’s Marc Benioff have emphasized the shift from "copilots" (Hands Ready) to "agents" (moving toward Hands Off). Benioff recently remarked that the future of the enterprise is not about humans using AI, but about "AI-powered workforces" where autonomous agents handle the mundane to allow humans to focus on high-value innovation.
However, labor advocates and organizational psychologists raise concerns about the "Hands Ready" phase. A 2024 study by the MIT Task Force on the Work of the Future found that workers in hybrid AI environments often report higher levels of stress and "automation complacency," where the requirement to remain "ready" to intervene in an automated process leads to cognitive fatigue.
Chronology of the Transformation Journey
The path to the autonomous organization typically follows a recognizable timeline, though the pace varies by sector:
- The Pilot Era (2022–2023): Organizations experimented with generative AI for individual productivity (subtractive value).
- The Integration Crisis (2024–2025): Companies attempt to scale AI across departments, hitting the "Hands Ready" bottleneck where human-machine coordination creates operational friction.
- The Agentic Shift (2025–2026): The introduction of autonomous agents that can execute multi-step tasks begins to move organizations toward the "Hands Off" model for specific functions like customer service and IT operations.
- The Autonomous Paradigm (2027 and beyond): Leading organizations restructure their entire hierarchy to support a "machine-first, human-governed" model, capturing non-linear value.
Strategic Implications and the Human Role
The most profound implication of this shift is the changing role of leadership. In a traditional organization, leadership is often synonymous with operational oversight. In an autonomous organization, operational control is handed over to AI systems that can process data and execute tasks at speeds unattainable by humans.
This does not render humans obsolete; rather, it deepens the necessity for "mission control." Humans must take on a more explicit role in defining the "why" of the organization—its strategy, ethics, and purpose. The governance of AI systems becomes the primary responsibility of the C-suite. As the automotive industry learned, the transition from "Hands On" to "Hands Off" is not just about better sensors; it is about a fundamental redesign of the vehicle. Similarly, AI transformation is not about better software; it is about a fundamental redesign of the corporation.
The "Hands Ready" phase remains the greatest challenge. To navigate it, businesses must move beyond seeing AI as a series of disconnected tools and start viewing it as a structural evolution. Those who fail to recognize the complexity of this hybrid stage risk becoming stuck in a state of permanent friction, unable to reach the transformative value that defines the autonomous era. The journey is not linear, and the end state is not just a faster version of the past—it is an entirely different kind of entity.
