"AI-First Mindset" - Is this just another rodeo?

Before even I welcome you into the Agentic AI era, it feels like we are already in the middle of it. Everyone that I meet with are progressing much faster in AI than I have ever seen them in the previous trends. Cost, quality and skillset availability are the grounding forces for their ambitious AI journey.

I remember my younger days when I was shifting from 'mainframe' to 'client-server'. The best advice I got then was to think of "modularity" in the design. It helped me in the next 30 years as the world evolved through client-server to cloud to microservices to agentic to physical AI.

Along came the push for the Automation-First mindset which helped us rethink work and face the challenges of job losses. Introducing automation was a massive undertaking and we were successful to some extent.?

Remember the days when we were all preaching and practicing? "Agile Mindset". We faced many challenges such as resistance to change, unfit org culture to agile, micro-managing teams, Infra challenges, communication and collaboration issues etc. Unless a firm is agile-native, most have not yet achieved 100% agile adoption. For many, Agile-Mindset is only on the organizational mission statements and on few walls inside the company.

More than anything, it is our mindset that helps us steer towards the most relevant context and constraints for a given task during the technological evolutions.

Right now, should we adopt an AI-first mindset?

Let us look at the top 3 classic errors we made during our agile adoption and compare that with the current AI-First scenarios. (If your organization has not made the below errors, pat on your back!)

1. We boiled the ocean: Agile adoption was centered on changing the organizational culture.? Agile challenges the way 'how we do things around here'. So, we trained everyone in IT, business and operations, that gave us the notion of moving the entire organization towards a major change. Change management was a nightmare.
2. We cut our legs for the shoes: Some of us created iterative waterfall process and called it agile. Many of us brought standard textbook theories and force-fitted the organization to show obedience. Micromanagement under the name of daily standups irritated some of the innovative and liberated souls among the development communities.
3. Our measures of success were wrong: Instead of measuring the outcome of agile adoption by means of customer satisfaction, competitive advantage, market expansion, customer acquisition, etc., our measures stopped with individual project metrics. Even if somebody was tracking the outcome, it was not recognized by the development community.??

Unlike Agile, AI-First mindset is not cultural, but technical (Is it?).

If we approach the AI adoption similar to Agile adoption, we will fail.

AI is everywhere. It integrates into all aspects of our lives - personal and professional. It also changes some of our core behaviors. Unless we convince ourselves on the benefits of AI, businesses having the AI-first mindset will be waste of resources. So thinking about AI-First mindset in our personal lives is equally important.

There was a time when I was driving the Automation-First approach in my organization. It was just around the Agile adoption days. We did the following steps and found it successful. (Till date this is a proven model for accelerating automation!)?

1. Baseline the Automation Maturity: Start with an automation assessment to evaluate the firm's maturity level, potential areas that could/should be automated, coverage, leverage etc. That way, we not only identify gaps, but also establish a baseline.
2. Run pilots: Choosing select business and/or technical processes and running pilots to test the automation framework, technologies, and preset KPIs, so that the foundation is laid off.
3. Building an expansion plan: Identifying potential expansion areas for automation with a well-defined plan for resources is key.
4. Talent development: An important step is to address the talent demands through hiring new talent, upskilling the existing workforce, that is aligned to the automation plan.
5. Establish the Automation Authority:? Effective management is key to establish standards, frameworks for repeatability, continuously monitor to improve the automation maturity of the organization.

Building the AI-First Mindset can follow the same process as that of Automation-First approach and it will work.??

However, be careful with the following:

1. Most of the times, the lack of the leadership's understanding and support is the main reason these initiatives fail.
2. Lack of clear priorities and directions make these initiatives no-one's responsibility OR everyone's responsibility.
3. This is not once-and-done initiative. There is a long road ahead of us.

After all said and done, I am expecting the same old problems/challenges that we faced in the earlier rounds of automation-first, agile-first, cloud-first initiatives in this time as well. Good luck.