When it comes to AI adoption, one theme keeps coming up in conversations with executives: enthusiasm is high, but focus is scattered.
As one Chief Data Officer recently told Merav Yuravlivker, Chief Learning Officer of Data Society Group, “Everyone is AI-ing all over the place.”
It’s a striking phrase because it captures where many organizations find themselves today. Leaders are racing to implement generative AI training and tools in as many places as possible, but often without a clear sense of purpose. The result is scattered pilots, overlapping systems, and uncertain returns. Merav is quick to point out that while enthusiasm is essential, it’s not enough. “My recommendation is always to start with the solution, start with the objective, and then determine what the parameters are and where AI can fit into that strategically.”
Why Every AI Training Program Needs a Baseline
That clarity starts with something deceptively simple: knowing where you are right now. Merav has seen too many companies skip this step. “One of the biggest mistakes that I see with the organizations that I talk to is that there isn’t a baseline for what they want to improve,” she explains.
It’s easy to assume that launching a generative AI training program will automatically create efficiency gains or boost retention. But Merav emphasizes that you can’t know if that’s true unless you measure what things look like before you begin. “Sometimes a 5% improvement can mean tens of millions of dollars, and sometimes not. It depends on the size of your organization and the projects you’re working on.”
The takeaway: don’t just expect AI training to move the needle, define what the needle is and how much movement matters.
Expanding AI Training Beyond Technical Teams
When Data Society launched in 2014, most AI training programs centered on technical staff: data scientists, engineers, and IT teams. Fast forward to today, and the conversation looks completely different. Executives now come to Merav with a new priority. “Executives [are] coming to me and saying, ‘I wanna make sure that everybody in our company is AI literate, is data and AI literate.’ And I love that.”
This reflects a broader understanding of how generative AI training creates value. It’s not just about coding, it’s about giving everyone the confidence to think differently. As Merav puts it, “Not everybody needs to be able to program these models, but to be able to understand it empowers every individual to think about, ‘Well, how can this help me? Here’s a problem that I’m seeing that nobody else really sees: how can I bring this to the table and work with my colleagues to solve this challenge?”
When literacy spreads across the organization, it creates a culture of collaboration. “Being able to upskill non-technical people and help them understand what AI does and doesn’t do helps create this shared vocabulary and this shared knowledge. It improves communication and improves collaboration.”
MUST READ: Shadow AI Is a Signal of Systemic Gaps
Why Community is the Secret Ingredient in AI Training
But Merav also cautions that training alone is rarely enough. To sustain the benefits of an AI training program, companies need to build communities of practice. “One of my favorite aspects that I’ve seen work really well to help ensure that people are not only learning the skills but also implementing the skills is the building of a community within an organization.”
This matters even more in remote and hybrid environments, where informal learning rarely happens organically. Communities of practice help employees go beyond the classroom to apply AI training, share real-world projects, and solve challenges together. As Merav explains, “By creating communities where people can share their projects, they can share their best practices, they’re able to show off the work that they’re doing or they have a challenge that they need to solve, you’re building connections within a company, you’re building that knowledge repository so people are not duplicating work, and you’re increasing the engagement.”
The ripple effects are powerful: higher engagement, stronger collaboration, and better ROI from your AI training investment.
Measuring Success in AI Training Programs
Finally, Merav emphasizes the importance of tracking what matters. Adoption and efficiency are two metrics every AI training program should measure. “If you’re implementing this tool across your organization, how many people are actually using it on a regular basis? One way to improve the usage is always through upskilling, through training programs, through initiatives to use these tools.”
Tracking adoption shows whether employees are embracing the tools they’ve been trained on. Measuring efficiency, through before-and-after comparisons of processes, reveals whether AI is saving time and resources. Without those measurements, companies risk assuming success when the reality is just noise.
Generative AI training holds enormous potential, but the organizations that succeed do three things differently:
– They set baselines before rolling out AI training programs.
– They expand training beyond technical teams to build a company-wide culture of AI literacy.
– They create communities that keep skills alive long after the training ends.
– They measure adoption and efficiency to prove impact.
As Merav Yuravlivker reminds us, excitement alone isn’t enough. Clear objectives, strong foundations, and intentional AI training programs are what separate organizations that “AI all over the place” from those that achieve real transformation.Want to chat with Merav, book a meeting here.
FAQ: Stop “AI-ing All Over the Place”
A baseline tells you where you’re starting from. Without it, you can’t measure progress. Merav points out that even a 5% improvement could mean tens of millions of dollars—or very little—depending on your company’s size and goals. Baselines make it possible to track real ROI from your AI training program.
