Ready to Lead in the Age of AI? How the S.T.E.P. Framework Can Guide Your Transformation

?Omar Andrade Organisational Effectiveness

?

This is a summary of a feature article in the November-December 2023 issue of the Harvard Business Review. The article discusses how to capitalise on generative AI models like Chat GPT. I was motivated to summarise this article because it proposes a simple framework. The framework is called S.T.E.P.; it stands for Segmentation, Transition, Education and Performance. If one universal law about adopting new technologies existed, it would be this: People will use digital tools in ways you can’t fully anticipate or control. The S.T.E.P framework tries to address this law.

?

1?? SEGMENTATION

?No single AI will do everything one person does in a work role. Informed leaders should ask, “How will AI affect the various tasks my employees engage in?’ To determine the answer, have your employees create three categories: (1) tasks that AI can’t or shouldn’t do, (2) tasks for which AI can augment workers’ actions and (3) tasks that AI can automate.

Example. (1) Tasks that AI can’t or shouldn’t do. AI would not help determine how to comply with federal policies or how to safeguard company IP from external consultants.

(2) Tasks for which AI can augment workers’ actions. It accurately cross-referenced contract details against a request for proposal (RFP). Here, ChatGPT can be handy. After reading through an RFP and a standard contract template, it could generate a draft contract that reflects the terms of the contract. Paralegals could then review the drafts for specific areas of concern that need to be amended.

(3) Tasks that can be completely automated. Laborious emailing of outside parties requesting changes to contracts. AI can automatically generate those emails by reading the revised contract language.

Once employees had segmented their tasks into the above three buckets, they began figuring out how AI could augment or automate some tasks, freeing up five work hours weekly for additional tasks.

Successful companies adopting AI encouraged employees to take the lead on segmentation and asked employees to experiment with the tool. Their leaders convened meetings at which employees discussed the results of their experimentation. They allowed the employees to reach a consensus on best practices. Being part of the experimentation and planning gave the employees insight into how their companies would use AI, reassuring them that automating part of their jobs wouldn’t put them out of work.

?

2?? TRANSITION

Because AI either helps complete work tasks faster and more accurately (augmentation) or takes some of them over completely (automation), some employees will have less to do after AI is deployed. In some cases, companies may reduce their headcount. Yet, among the ten companies studied for this article, only one eliminated jobs in response to the efficiencies gained by augmenting and automating work. The common strategies were to transition work roles by deepening or upgrading them. Deepening roles allows employees to devote more time to tasks they previously could not find time for. Upgrading roles means using the extra time to give employees more critical work. Example.

Deepening. In a marketing department, ChatGPT was used to help marketers create marketing collateral, such as PowerPoint presentations. That gave them time to spend on more value-adding tasks like competitor analysis and evaluating campaigns. Managers noticed which employees had an aptitude and interest in deepening their knowledge and worked with them to offer further training. Upgrading. It involves having employees perform tasks typically done by someone more senior employees. Junior civil engineers who previously built land-based models could now use their time for scenario planning, a higher-level task generally reserved for senior engineers. This meant that new tasks had to be found for senior engineers, so the lead engineer turned his responsibility of managing relationships with city planners over to the senior engineers. “After careful evaluation, I decided to give them a chunk of my job”. “That allowed them to feel comfortable giving up scenario planning. Now I can focus my efforts in new directions, too, since I’m freed from maintaining all those relationships”.

Deepening work roles is the most common strategy in this study. Most leaders find it easier to help employees identify new value-adding activities within their current roles than to take over tasks from more senior employees.

?

3?? EDUCATION

The nature of large language models like chat GPT is that their self-learning algorithms constantly evolve and take in new data points. This means that new ways to use AI crop up continually. Employees, therefore, need to know how to use it best. How to create effective prompts (prompt engineering) and evaluate the validity of AI’s predictions. Employees can't learn new skills once, and then they will be done with them. They must revisit the segmentation process periodically and continually refresh their learning about AI’s capabilities. The suggested education cadence (from a three-year study in this area) is annual.

Example.

A data company embraced the need for continual reskilling by hosting a “boot camp” for its employees to create custom teaching programs for its staff and devised a test that employees had to pass to be “AI-ready”. (This is an expensive example.) Elsewhere, companies with a smaller budget purchased subscriptions for short courses on AI? from companies that offer corporate learning, i.e., LinkedIn, Udemy and Udacity. Employees were encouraged to complete one course each month.?

Companies that framed AI as a learning opportunity had two things in common. (1) Learning culture. Employees were expected to learn to use AI ideally or segment their tasks. Instead, they were expected to explore its capabilities and take time to determine how best to incorporate it into their work. (2) Company-sponsored learning time. Employees were given two days to attend the boot camp or three hours a week to take online courses to which they subscribed.?

?

4?? PERFORMANCE

Most discussions regarding AI's impact on individual performance relate to productivity. As a result, performance evaluation has shifted in two key ways. (1) Segmentation has quickly changed the expectations regarding what tasks employees should perform and how they should perform them. (2) Evaluation cycles have shortened. The speed at which AI improves increases its ability to augment and automate tasks, thereby changing the nature of the employee’s role. Evaluation cycles have had to be shortened in reply. (From yearly to quarterly to keep up with the speed of change.)

KEY TAKEAWAYS

Integrating the S.T.E.P framework promises to optimise our workflows, cultivate forward-thinking, and increase workforce adaptability.

Segmentation Classify tasks into three categories: those unsuitable for AI, tasks where AI can support human efforts, and tasks suitable for full automation. This allows us to identify where AI can best complement our skills, freeing valuable time for more strategic work.

Transition Jobs will evolve as AI takes on routine tasks. Instead of reducing headcount, we can deepen or upgrade the existing roles, allowing team members to focus on higher-value activities. This strategy aligns with the Agile principle of continuous improvement and adaptation.

Education AI advancements will require an emphasis on continuous learning. Remaining proficient in leveraging AI tools will enhance our problem-solving and innovation capabilities.

Performance We must adopt more frequent performance reviews to reflect the rapid pace of change and its impact on roles and responsibilities.

?

Omar Andrade // Open to roles in organisational effectiveness.