A Talent & Transformation Executive on How AI Is Quietly Redesigning Organizations—Leaders Need to Catch Up

I also sit on a number of steering committees at the World Economic Forum. I am a member of the faculty at the Kellogg School of Management at Northwestern, where I am also an executive fellow. And I am the author of six books and a couple of hundred articles on the future of work and the impact of AI on work and organizations.

Ravin's Tip

There’s this belief that you can lead with technology and somehow magically get a return, as opposed to doing the hard yards of redesigning the work and the architecture of work that surrounds the organization. This misconception is where the fundamental obstacles to technology adoption lie.

How AI is shifting models of work from "fixed" to "flow"

I see leadership changing quite dramatically in an AI-first world, and not just in the leadership role that I have, but also in the work that I do with my clients.

Specifically, I'm seeing that leaders need to orchestrate a much wider and more distributed ecosystem of options for getting work done and for executing their strategies — whether that's with employees that are part of their organization, AI that they've developed internally, AI that they have procured, the capabilities of an alliance partner, or the skills of a gig worker.

All of those capabilities come into play in an AI-augmented world as the velocity of work increases and new skills are demanded by AI. And this is replacing the notion that work is bound up in a job with more agile ways of working.

In my last book, I describe three models for engaging talent:

1. The fixed model: This is the traditional job-based model.
2. The flexible model: This is where people are in jobs, but have the flexibility to express their skills in different domains or acquire new skills.
3. The flow model: This is where talent only connects to work through projects, assignments, and gigs.

AI in organizational design now affects every facet of work and leadership, so we are seeing businesses move from fixed to flexible to flow models of work.

AI now affects every facet of work and organizations and leadership, so we are seeing organizations move from fixed to flexible to flow models of work.

Share This Quote on:

• Share on Twitter
• Share on LinkedIn
• Share on Facebook

Ravin Jesuthasan

Global Leader for Transformation Services at Mercer

Why leaders must redesign work before deploying AI

There's this belief that you can lead with technology and somehow magically get a return, as opposed to doing the hard yards of redesigning the work and the architecture of work that surrounds the organization.

This misconception is where the fundamental obstacles to technology adoption lie.

Thanks to generative AI, we've overcome that last-mile problem that AI used to have when it was not fit for purpose. Generative AI is a technology that is incredibly easy to use. It's conversational and it's deployed to us in technology that we use every day. But what continues to be the most vexing challenge is the architecture of work.

That's where leaders in progressive organizations are actively working. And for good reason — it's where they can capture exponential gains.

Leaders must be willing to look beyond the machine substituting humans to actually see all of the potential for a distributed ecosystem.

Focus on these types of questions: "Where can work be potentially moved to a lower skill level?" And, "As work is substituted, where do we need someone with a higher skill level to take on the work?"

Here's more on how to redesign work and leverage AI-enabled productivity.

A real-world example of a "work-forward" approach

Here's an example from work done with one of our clients. At a division of a very large global financial services organization, we redesigned work around the capabilities of a new proprietary technology platform that included computer vision for data ingestion, process automation, machine learning, and Generative AI.

Redesigning the work first allowed us to take tasks, move them onto the platform, substitute them, and augment them with the capabilities of the platform. And that wasn't the end of it. Other pieces of work were then able to move from talent at more senior levels all the way down to novices. This was possible because the tasks placed a premium on experience, but the technology reduced those skill premiums.

In fact, in some instances, we were able to eliminate the need for certain types of skill entirely. The technology substituted this entry-level work.

We used a "work-backward" approach to redesigning the work, rather than the "tech-forward" approach that so many organizations try to implement — often unsuccessfully. The benefit of our approach is that it frees up and exposes not just the primary effects of the technology, but the secondary and tertiary effects as well.

Ravin's Tip

We used a “work-backward” approach to redesigning the work, rather than the “tech-forward” approach that so many organizations try to implement — often unsuccessfully. The benefit of our approach is that it frees up and exposes not just the primary effects of the technology, but the secondary and tertiary effects as well.

Why a "work-backward" approach improves AI adoption

A "work-backward" approach helps with AI adoption as well.

When implementing AI in strategic planning, it is really easy for organizations to spray and pray — spray the technology and pray that people will adopt it uniformly. But that rarely happens.

Instead, they need to do the hard work of:

• Redesigning jobs
• Redesigning processes
• Changing the requirements of new talent
• Ensuring the discipline with which the technology is used and governed and managed to ensure consistent application and deployment.

I know, you know, we all know that change management is the thing that makes or breaks transformations. Yet somehow, AI in change management — and the active redesign of work — is still the biggest impediment for organizations.

Ravin's Tip

Every leader needs to be AI fluent, and you don’t get there unless you are using this technology every day. You don’t get there without experimenting, without trying. Most importantly, you don’t get there without role modeling for the rest of your organization and leading from the front

How building AI literacy accelerates adoption

We have been using our own LLM for the better part of two years now and, like many organizations, we democratize access.

To ensure adoption and literacy, our leaders role model its usage. We also train people on the use of the technology. We've trained them to understand what a hallucination looks like. We've trained them to validate and check the data. We've trained them on what appropriate usage of AI in business operations looks like, what to upload and what not to upload.

Training happens through a variety of means:

• AI-based simulations
• Peer-to-peer learning
• Workshops

We have invested in getting our people fully up to speed so that AI in learning and development is truly a partner they turn to every single day for every piece of work.

But I'll say it again: You can't get to the ROI from AI if all you do is democratize access without intentionally redesigning the jobs and processes.

How an in-house AI stack powers faster work redesign

Like virtually every business leader, I'm obsessed with Gen AI.

It has allowed us to build on a legacy of process automation by progressively enhancing the level of intelligence and workflow integration. And it's helped us power an agentic capability that is truly transforming every aspect of our work.

We use our own LLM for this, as well as an in-house work-design tool that helps our clients redesign their jobs and processes.

How agentic workflows are transforming talent, operations, and consulting

Speaking of agents, agentic workflows permeate every aspect of our business model and culture — from the operational parts of consulting to talent processes to our financial management processes.

Agentic AI is creating opportunities for new services and adding value to existing services. It is really difficult to overstate how dramatically it is transforming our strategy and our business model.

Our internal work-design tool has allowed us to transform a six-week consulting process into a six-hour, three-workshop process that allows us to help our clients redesign their jobs and processes to better enable AI adoption

The tool helps us deconstruct jobs and processes, analyze the work and identify redeployment options where work can be substituted or augmented by different types of technology (RPA, machine learning, Gen AI, agents, robotics, etc.) and reconstruct new ways of working.

Why leaders must become AI fluent and lead from the front

Soon, teams and clients will expect expertise at much greater velocity with a level of insight and testing that is unprecedented — and increasingly so with the development of digital twins. Tested, actionable solutions will need to be deployed in days and weeks versus months.

With AI in leadership development, you don't get there unless you are using this technology every day. You don't get there without experimenting, without trying.

Most importantly, you don't get there without role modeling for the rest of your organization and leading from the front.

Follow along

You can follow Ravin Jesuthasan's work on LinkedIn as he continues implementing AI via his "work-backward" approach.

More expert interviews to come on People Managing People!