Today, AI is a reality reshaping the way we live, work, and even think. Recently, I encountered an article from Stanford's Erik Brynjolfsson aptly calling out the "Turing Trap": designing AI to mimic human intelligence rather than augmenting it. The question isn’t whether machines can do what humans do, but how they can amplify what we are capable of achieving.
I think the idea is simple: identifying tasks humans can do that AI can’t today, human tasks that machines can automate, and finding new tasks humans can do with AI. As a thought experiment, I am looking at cases in roles like marketing, finance, legal, HR, and consulting. In the end, I want to develop a simple methodology that we can follow personally or in our organizational transformation projects.
The Three Categories of Tasks:
Tasks Humans Can Do These are the bread-and-butter activities humans have performed for millennia. Creative problem-solving, strategic decision-making, and emotionally intelligent interactions fall squarely into this category. Think of a lawyer navigating the subtleties of a complex negotiation or a doctor comforting a patient.
Human Tasks That Machines Can Automate This is where automation comes in. Machines excel at repetitive, high-volume tasks that require precision. For example, AI tools can analyze vast datasets, crunch financial numbers, or screen resumes at lightning speed. While these tasks are vital, they don’t necessarily require human ingenuity.
New Tasks Humans Can Do with AI (or simply Augmentation Cases): Here lies the magic. AI opens doors to entirely new possibilities by acting as a co-pilot or your wise friend. With AI, a marketer can use predictive analytics to anticipate customer needs, a financial analyst can simulate market scenarios in real-time, and a consultant can tailor solutions using deep behavioral insights. Today, for simple projects, AI tools can write code for data science, front-end applications, and back-end applications. For complex architectures, it’s still a bit problematic, but for those projects, as a co-pilot, it works perfectly. I have never written a Flutter application before, but I recently completed an end-to-end application by just using Claude and ChatGPT.
Role Cases and Method
To explore how AI can augment human tasks, I selected two distinct roles: Marketing Manager and Registered Nurse. These roles represent diverse domains and responsibilities, offering a comprehensive view of AI's potential across professions. For clarity and consistency, I used O-Net's detailed role descriptions to outline the key tasks, skills, and expectations.
Marketing Managers oversee strategies to promote products and services while ensuring alignment with organizational objectives. Key responsibilities include:
Developing and executing marketing plans.
Analyzing market trends to inform pricing and distribution strategies.
Managing teams and coordinating with cross-functional stakeholders.
Using tools like CRM software and data analytics platforms to optimize campaigns.
Registered Nurse Registered Nurses provide essential healthcare services, focusing on patient care and medical interventions. Their core responsibilities include:
Monitoring patient conditions and administering medications.
Coordinating with healthcare teams to develop care plans.
Performing diagnostic tests and educating patients on health management.
Managing medical records and ensuring compliance with healthcare standards.
These roles were selected to illustrate AI's impact on tasks requiring both strategic decision-making and hands-on execution. Using these descriptions, we categorize tasks into three areas: tasks AI can fully automate, tasks AI cannot yet automate, and tasks where AI can serve as a tool for augmentation.
Analyses of Tasks for Our Distinct Cases: Augmenting Humans with AI
Marketing Manager
Task | Category | Potential AI Tools and Strategies |
Identify, develop, or evaluate marketing strategy | Augment | Predictive analytics (Tableau), strategic planning tools (Miro) |
Formulate, direct, or coordinate marketing activities | Augment | Collaboration tools (Asana), CRM systems (HubSpot, Salesforce) |
Evaluate financial aspects of product development | Automate | Financial modeling tools (AI-driven budgeting tools) |
Develop pricing strategies | Augment | Dynamic pricing models (Shopify AI, Python-based pricing algorithms) |
Compile lists describing product or service offerings | Automate | Data organization tools (Airtable, Google Sheets with AI plugins) |
Direct hiring, training, or performance evaluations | Cannot Automate | Requires nuanced interpersonal skills |
Consult with product development personnel on product specifications | Augment | Design collaboration platforms (Figma, Adobe XD) |
Use sales forecasting or strategic planning | Augment | Machine learning-driven forecasting (Google Analytics, Oracle Cloud forecasting tools) |
Negotiate contracts with vendors or distributors | Cannot Automate | Requires high emotional intelligence |
Coordinate or participate in promotional activities | Augment | Event management tools (Eventbrite, Monday.com) |
Initiate market research studies or analyze findings | Augment | Data analytics tools (SPSS, Python pandas library for data analysis) |
Registered Nurse
Task | Category | Potential AI Tools and Strategies |
Record patients' medical information and vital signs | Automate | Wearable monitoring devices (Apple Watch, Fitbit Health) |
Administer medications to patients | Cannot Automate | Requires human judgment and interaction |
Monitor, record, and report symptoms | Augment | Patient monitoring systems (Philips IntelliVue, AI-driven alert systems) |
Provide health care, first aid, immunizations | Cannot Automate | Hands-on care needed |
Consult and coordinate with healthcare team members | Augment | Telehealth platforms (Teladoc, Zoom for Healthcare) |
Direct or supervise less-skilled nursing or healthcare personnel | Cannot Automate | Supervisory and decision-making role |
Order, interpret, and evaluate diagnostic tests | Augment | AI-powered diagnostic tools (IBM Watson Health) |
Develop health improvement programs | Augment | Program design tools (Canva for communication, Miro for strategic planning) |
Conduct research activities | Augment | Research databases with AI-assisted searching (PubMed, EndNote with AI capabilities) |
Based on my analysis for Marketing Managers, approximately 40% of their tasks can be automated, freeing them to focus on high-impact areas. About 30% require uniquely human capabilities like creativity and leadership, while the remaining 30% can be significantly enhanced with AI tools, such as campaign optimization and predictive analytics.
For Registered Nurses, the automation potential is lower, at 30%, reflecting the hands-on and empathetic nature of their work. However, AI can augment 20% of their tasks, particularly in diagnostics and patient monitoring, leaving the remaining 50% squarely within the realm of human expertise.
For both roles, we can definitely add AI-centered education and skill upgrades. I only performed analyses for several tasks and activities for both roles to demonstrate there are great augmenting opportunities. Of course, some tools may not seem AI-driven at first glance, but both roles can use those tools with help of AI, increase tool usage efficiency, or directly use a tool's AI capabilities for automation or semi-automation.
Framework for Organizational AI Augmentation Projects
To implement AI-driven augmentation effectively, organizations can follow this structured framework:
Task Inventory: Create a comprehensive list of tasks and expectations for each role or process. Use detailed role descriptions and engage employees to ensure no task or expectations are overlooked. Combine your internal research with external research. Many case studies and market researches available and don't miss them.
Categorization: Divide tasks into three categories:
Fully Automatable: Repetitive tasks that AI can handle end-to-end.
Non-Automatable: Tasks requiring human judgment, creativity, or empathy.
Augmentable: Tasks that can be improved with AI tools but still require human input.
Technology Mapping: Identify AI tools and technologies that align with each task category. Ensure tools are user-friendly and fit the organizational context. Also tool assesment is highly critical for such projects. Prioritizing tools that come to you with built-in AI capabilities can create both cost advantages and fast learning curve.
Pilot Implementation: Begin with a pilot project for one role or department. Use measurable KPIs to track the impact of AI augmentation. Focus on value driven activities, measure their outcomes
Skill Development: Train employees to use AI tools effectively. Offer workshops or online courses tailored to the tools being deployed. If you can don't rush, go one by one and create realistic roadmap.
Feedback Loop: Continuously gather feedback from employees and stakeholders. Refine processes and tools based on real-world usage and outcomes. Combine feedbacks with KPIs
Scale and Optimize: Expand successful pilots across the organization, optimizing for efficiency and scalability while ensuring alignment with strategic goals.
By following this framework, organizations can ensure that AI-driven augmentation adds value, enhances productivity, and fosters employee satisfaction without undermining the human element of work. These kinds of internal projects are crucial today. Many companies focus solely on automation projects, overlooking augmentation opportunities and failing to enable people to realize their true value and potential contributions. The goal should always be to empower people to bring their best to the table while letting AI handle the mundane and amplify the extraordinary.
Comments