The paradigm shift that’s leaving us behind: Upskilling in the age of AI

By Sankarson Banerjee, Ex-CIO and board member

So AI is all the rage nowadays; every company knows that a generational shift is imminent. That they must start adopting AI or fall behind; therefore, the scramble to adopt AI in ways that make a real competitive difference.

Of course adoption is not going to be easy - It is going to take its own sweet time twisting itself into loops about the Gartner Hype Curve. Predicting what difference AI will make, where opportunities will crop up and how people will shape themselves around AI is no simple ask - each company has to evolve this for itself. One thing remains a constant, though; you will need different types of knowledge and different kinds of training to embark on this journey. In this new world the tools, the steps, the technologies, the ways of working are different – very different. What got you here will not get you to success in the new world of AI. You will need a different set of skills.

Adopting AI, the change in approach
The first step is in upskilling the workforce. Enterprises should focus on exposing their employees to a brand new set of tools that will become standard in the AI world. You will need these new basic tools in the new world, and you will need people familiar with them. 

Traditionally, most companies first implement the tools they plan to use, and then train the people. Facing a paradigm shift, it is very hard to adopt tools and build use cases without people who already have some level of skill in the new paradigm. And we have tried this before. We've tried to implement Cloud transformations with people trained on traditional technology. We've tried doing new world businesses with traditional managers. Many of these experiments have led to repeated failures, significant overruns in cost and effort and frequently - very poor business outcomes. Sure you can try to hire this talent, but when companies are handing out gigantic pay checks in a red-ocean world of talent most of you are going to be at the very bottom of the pile.

How to gain adoption within the company
Therefore, I am recommending in this new world that we reverse the approach. We first train the people. And then we use these trained people to help us identify where the use cases are, where the opportunities are, and how to implement the tools. How to gain adoption within the company. This is not entirely a new approach. We have seen this before; companies used to train all their employees on Word and Excel as part of basic training. Or ask every employee hired fresh to learn SAP or Core Banking or whatever enterprise system was being used (because that was not taught in college) even if his or her role did not immediately require it. So it is not an entirely new idea, but it is effective in a paradigm shift where transformations are earthquake-sized.

The approach I want to propose is that enterprises identify a set of core tools suitable for basic AI literacy in their industry, and train all their employees to a minimum competency in those tools regardless of immediate need. Yes, you will get the usual objections that this will lead to attrition, this is a waste of time and all that stuff. And of course you have to handle all that, but you can't really avoid it. If you have an AI illiterate workforce, you are never going to be successful in the AI paradigm.

The culture change
The second kind of upskilling is a larger problem. Being successful in the new world is not just a matter of skill, juniors. There is also a large component. Of change that happens to the culture of a company, to the workflows of a company, to the organizational structures and rewards of a company. Again, this is something that many industries have learned the hard weight during various paradigm shifts, and many super successful industries have failed to do this shift in company culture and fallen flat in the new world. We have seen these again and again in MBA case studies - Kodak, Band Box, HMT were all one-time giants who faded away.

A culture shift also requires upscale. It requires a different kind of skilling. In learning how to think in the new world. A good example in the AI world is Agentic workflows. Of course you need the tools that design and run these workflows, but you also need an organization that can handle. And be quality in our workflow as opposed to the earlier version of workflows with predetermined inputs, logic and outputs. The earlier generation of workflows. Work primarily blind executors. Where is the new generation of work. Workflows are collaborators. But of course, it does take training for managers to know how to manage these new styles of working effectively. It is therefore equally important for senior management to go through what I would call “soft” training – as in Conceptual subjects rather than tools and techniques. There are many available avenues – Conferences on AI, Universities offering these as continuing education programs, or specialised trainers who train senior level employees, Boards, CEOs, CFOs in the new world of AI.

The biggest challenge in any paradigm shift is a natural resistance to change. We have been trained in particular tools, techniques, ways of doing things, management paradigms, and we are very reluctant to challenge the training. And reskill ourselves. In this new world. But for any enterprise, it is now becoming a pressing imperative. Some - very few - will be able to hire their way out of these things but most companies will have to focus on building a large pool of internal talent that is capable of managing this paradigm shift.