AI, your workforce multiplier

Will you get 10% or 10X better? Put differently, are you using AI to make existing work slightly faster, or to fundamentally redesign how work gets done? That was the driving question at the YegaTech AI Summit I recently attended in Palo Alto. While many of us are taking a scattershot approach to AI, with small AI pilot projects or allowing staff access to tools like ChatGPT, true AI adoption requires structure. Unlike previous technologies, AI’s ubiquitous nature gives it the potential for a far greater and more unifying impact when deployed effectively. What follows are my takeaways from the summit which I believe can be a starting point for other utilities.

To start, AI adoption is not primarily a technology challenge. Like all change management, it is a leadership challenge. From front line workers through the executive office, we need to prime our people for the deployment of AI solutions. In spending time at the summit with multiple CEO’s, their experience was the same; leadership determines AI success. As leaders, we must create an environment that allows for, encourages and rewards experimentation. Organizations should recognize and reward staff who are trying practical AI ideas, especially those that improve service, reduce friction, or create value.

One cultural primer for AI is the use of small group engagement, such as workshops or task forces. Group members are given the seeds of change, which they bring back to their divisions for germination. Leaders must also recognize that we are not always the most knowledgeable when it comes to AI and may benefit from upward mentoring. Pairing junior staff, particularly recent college graduates, with senior leaders can be highly valuable because they are often closer to emerging tools, more comfortable experimenting, and less attached to legacy processes. This kind of cross-generational collaboration can create a meaningful innovation pipeline, helping organizations identify practical AI use cases, challenge outdated workflows, and build a culture of continuous learning.

Beyond an openness to change, we must cultivate a high appreciation for data quality and ownership. AI depends on the quality and accessibility of organizational data. Several speakers emphasized that bad data plus good AI still produces bad results. Valuable information often sits across disconnected systems such as local drives, cloud storage, email threads, field reports, and enterprise platforms. The goal is not necessarily to build one massive data lake, but to make fragmented data more connected and usable. As leaders we have a responsibility to promote an organizational culture that values data.  

Another critical point is that business units, not IT, must own AI use cases. Each case needs a clear owner, a defined pain point, measurable outcomes, and an early plan for scaling if it works. Orphan ideas rarely get implemented. One trick is to assign the use case generator as its champion. They don’t need to have the technical know-how (though it helps), they just need passion to campaign and promote it. On the flipside, organizations must have the discipline to sunset use cases that do not deliver value. While difficult, ending an underperforming pilot frees resources and capacity for higher-value opportunities.

Bottom line, AI will not automatically transform your organization. The benefit will come from leadership choosing where to focus, creating space for experimentation, connecting data and knowledge, and helping employees adapt to a future where success depends less on doing tasks manually and more on designing, guiding, and improving AI-enabled work.