Six levels of autonomous work: How AI augments, then replaces

The AI revolution is unfolding before our eyes, yet it remains a daunting challenge for business leaders to chart a clear course and set a definitive vision. Amidst this uncertainty, we can still outline some observations about current and future AI capabilities that provide a foundation for understanding this transformative period. These insights include:

Also: Welcome to the AI revolution: From horsepower to manpower to machine-power

- AI's generative and predictive capabilities are already impressive and poised to become even more so.
- The field is buzzing with substantial investment and enthusiasm, a trend that shows no signs of slowing down.
- CEOs are perpetually seeking ways to achieve growth and improve margins with fewer resources.
- Many jobs or job components are routine, procedural, or algorithmic, making them prime candidates for AI automation. According to H. James Wilson and Paul Daugherty in the Harvard Business Review (Sept-Oct 2024), AI can augment most business functions and over 40% of all US work activities.
- New companies will soon be AI natives, meaning they'll primarily use AI resources and hire humans only when necessary. These companies will likely lead the way in demonstrating where human contributions are still valuable, and others will follow at varying paces.

Inspired by the "6 Levels of Driving Automation" from the Society of Automotive Engineers, we've developed a framework to illustrate how AI capabilities will evolve and impact companies over the next decade.

As AI resources continue to improve, they'll have a dual impact on businesses and the workforce. Initially, AI will augment human efforts by taking over low-value tasks, allowing humans to focus on strategic and creative work. However, within about five years, AI will start to assume entire job roles, beginning with the most procedural or rules-based positions. Eventually, AI will gain enough decision-making and orchestration capabilities to manage entire teams and even lines of business.

We've identified these as the Augmentative and Replacement phases, respectively. While the transition will be gradual at first, it will accelerate over time, with adoption varying by industry, function, team, and individual.

The Six Levels of Autonomous Work

Let's delve into the chart's details row by row:

Level: Each level of autonomous work is numbered from 0 to 6 and labeled with a title reflecting the amount and complexity of work AI can handle. Starting with the simplest unit, a Task (level 1), AI can perform basic functions. Moving up, a Sub-Process (level 2) involves a series of tasks in sequence to complete a part of a business process, such as gathering information for a customer case. At level 3, AI can manage an entire business process like taking a customer order or managing a case from start to finish. At level 4, AI can handle multiple processes, effectively covering roles like sales or service agents. At level 5, AI can perform most roles within a commercial team, including managerial positions. Finally, at level 6, AI can orchestrate the work of multiple teams, functions, and processes, essentially managing a business or line of business. This applies to all types of operations, from commercial to manufacturing.

Also: How your business can best exploit AI: Tell your board these 4 things

Phase: The progression of AI isn't a linear career path. Instead, it will transition through two distinct phases. Levels 1-3 represent the Augmentation phase, where AI assists and enhances human capabilities, creating new opportunities. Levels 4-6 mark the Replacement phase, where AI takes on increasingly complex responsibilities and begins to replace human roles.

AI Role: This section outlines AI's capabilities and its relationship with human colleagues at each level, from a non-technical perspective. We plan to provide a more detailed technical perspective if there's interest.

Human Role: This complements the AI role, focusing on the evolving relationship and shared responsibilities between humans and AI.

Adoption: We estimate the dates when mainstream adopters (encompassing early and late majority categories) will start using AI at each level. Innovators and early adopters will precede this, while laggards may lag unless spurred by a crisis. Adoption rates will vary widely by industry, department, and individual. Some employees will eagerly embrace AI, particularly when it frees them from monotonous tasks, while others might resist, especially if they fear job replacement. Yet, we're already seeing AI applications across industries, with more advanced bots and agents on the horizon.

Also: AI will change all businesses and most leaders are not ready

Autonomous Work Implications for Business

We've identified three key implications of this AI evolution for businesses, urging leaders to recognize and prepare for them:

- Planning for Augmentation vs. Replacement: The six levels don't represent a linear path for AI. The Augmentation phase (levels 1-3) is widely discussed because it's reassuring, with AI enhancing productivity and freeing humans from repetitive tasks. However, the Replacement phase (levels 4-6) will see AI taking over entire roles, and this transition will happen rapidly. HR and change leaders must start planning for this now.

- Accelerating Responsiveness: AI will speed up business operating cycles. In our 2023 book "Boundless," we introduced the SUDA model (Sense, Understand, Decide, Act) as the operating model for the AI era. AI will enhance a company's ability to navigate these stages, leading to quicker, more informed decisions and what the military calls "decision dominance" and "overmatch." Success will hinge on minimizing the time between each stage, ideally shrinking the Sense to Act delta to near zero. Each level of the Autonomous Work model increases AI's capacity in one or more SUDA stages, accelerating decision-making and action-taking across various scales. Companies unable to reduce their Sense to Act delta will be outpaced by those that can.

- Beyond Human Capabilities: At levels 5 and 6, AI will surpass human abilities, handling situations beyond what any number of humans could manage. It will be measured in terms of machine power, reflecting complexity, accuracy, and speed rather than just computational units or transactions per second.

Leadership Call to Action

AI is here to stay, and leaders must acknowledge its permanence, even if the current hype might be unsustainable. Even if not ready to fully embrace AI, leaders can take steps to prepare:

They can develop and implement a comprehensive data strategy, crucial for business regardless of AI. Streamlining primary business processes by eliminating, simplifying, and standardizing them is another wise move. On the HR and change management fronts, leaders must plan for both AI phases before they're upon us.

Also: Leadership alert: The dust will never settle and generative AI can help

While AI might seem like a challenge, it's also part of the solution for navigating increasingly uncertain times. It can play a vital role in helping leaders and teams make strategic, data-driven decisions and take effective action.

These are exhilarating times, and we hope our model provides enough structure amidst the uncertainty to help leaders take action.

*This article was co-authored by Henry King, a business innovation and transformation strategy leader and co-author of "Boundless: A New Mindset for Unlimited Business Success."*