“Hey Boomer, Watch Us Prompt”: Millennials Are Outpacing Leadership on AI Adoption (And How to Catch Up)
My advice to enterprises deploying AI tools is simple: talk to the younger people in your organization, particularly those in their 20s, 30s or early 40s, sprinting up the learning curve.
Friends and colleagues know the passion I feel for the promise of AI. We are in a moment as consequential as the rise of the steam engine. What is unfolding is as disruptive as the dawn of the internet. The future is as full of promise as any innovation in human history. Generative AI, I firmly believe, will unlock superhuman productivity, transform industries, and create profound value.
Yet, it’s sobering that so many enterprises remain stuck in the early innings of AI transformation.
The volume of work on this subject of adoption is daunting. Some of the best comes from The Wharton School’s Ethan Mollick, whose insights in his blog and videos are without peer. But I really recommend that anyone interested in the future of AI, in commerce and society, read a profound new report published by McKinsey, “Superagency in the workplace: Empowering people to unlock AI’s full potential”. More than incidentally, the report’s authors acknowledge, it was inspired by LinkedIn co-founder Reid Hoffman’s new book, “Superagency: What Could Possibly Go Right with Our AI Future”, another great read, which I reviewed just in February.
The Data Doesn’t Lie: Employees Are Ready, Leaders Aren’t
The McKinsey report paints a startling picture: employees between their 20s and early 40s — millennials specifically — are adopting generative AI tools 3x faster than C-suite leaders estimate. Their expectations for near-term growth are more than double than those of their bosses. They want to unleash that employee joy that I talked about last year.
That’s not an employee problem. That’s a leadership gap.
At one level, I don’t find this surprising. A year and a half ago, I published a series on AI and education in which I argued that, Youth Will Lead Us to a Much Needed, New ‘Learning Ecology’. I see this phenomena of understanding rising up from the younger, rather than being handed down from the older, in watching my own AI pioneer son Levi, who is 15. In a podcast he recorded last year with author and AI entrepreneur and prolific author Bryon Reese, Levi coined a phrase I’ve used many times since: “The better name for the age of AI is the age of a billion dreams.”
Nothing New Under the Sun: Youth Has Always Led Innovation
The history of the young as early adopters of new technologies is a long one. Martin Luther was only 34 when he embraced the printing press and triggered the Reformation. We wouldn’t have had the rock and roll revolution of the 1950s and 1960s without broadcast radio technology, then just a few decades old. The pace at which the world rushed online during the pandemic — a global exercise in distributed decision-making — was due in no small part to the fact a huge cohort of the workforce had picked up that skillset through gaming.
Boomers and Gen-Xers like myself can still contribute to the conversation. At 53, I’ve spent my career building data-driven companies and I now lead data.world, my sixth entrepreneurial venture and now the world’s leading knowledge graph based data catalog and governance platform. I’ve seen firsthand how AI efforts can flounder and that the core issue is not technical. It’s not even data science. It’s organizational will, data readiness, and leadership alignment.
That’s why your data teams are struggling.
Most organizations are trying to build AI on top of shaky data infrastructure. According to Cisco’s AI Readiness Index, 81% of companies believe AI will significantly impact their business, but the same number lack the centralized, low-latency, governed data infrastructure to support it.
The implications are enormous. Poor data quality leads to unreliable outcomes. Worse, without data context, LLMs “hallucinate” — a fancy way of saying they make things up with their predictive-engine-based approach. In enterprise settings, this isn’t just annoying; it’s a systemic risk.
This is why we built the AI Context Engine™ at data.world — to connect LLMs to business-context-rich knowledge graphs, delivering explainability and accuracy at scale. We’ve shown through deep studies that LLMs are 4.2 times more accurate when grounded in knowledge graphs than when relying on SQL databases alone. That’s not incremental. That’s transformational.
The Five-Point Playbook for AI Success
The playbook as we’ve learned at data.world, with plenty of help from Mollick and others, boils down to five points:
- Prioritize the data foundation: Before implementing AI, fix your data. Adopt knowledge graph-powered catalogs to contextualize, connect, and clean your data. Without this, your AI efforts will be unreliable at best — dangerous at worst.
- Create cross-functional AI teams: Integrate data scientists, business minds, and domain experts. Build with the business, not in isolation from it. As I wrote last week, making your enterprise data-driven is not an end in itself, it’s the means to empower your company’s larger mission.
- Invest in agile data governance and explainability: Establish agile data governance, the approach my co-founder Jon Loyens, our Chief Data and Strategy Officer, envisioned and coined as a new movement: Agile Data Governance.
- Make upskilling universal: Use platforms like Workleap or build your own AI Academy. Take a look at our effort, the data.world University. As Mollick’s classroom experiments show, anyone can become 25–50% more productive with just basic AI training. That’s not a luxury. It’s a necessity.
- Adopt an experimentation mindset: Not every AI initiative will succeed. But the ones that do will reshape your business. At data.world, we appointed a VP of AI Ops, Brandon Gadoci, who helped deliver a 25% productivity gain, making our company more resilient and unleashing employee joy (and increasing creativity as a result). The key? Constant iteration.
Mind the Gap: What’s Really Holding Us Back
The disconnect between the generals and the troops on the front lines is a major barrier to scale. In my experience, employees are eager to use AI but often lack access, training, or support. They need leadership willing to invest in upskilling, embed AI into workflows, and, most importantly, trust them to innovate.
We need to empower millennial managers — who are already the most confident AI users — to be the champions of this transformation. They’re ready. The C-suite needs to catch up. The real blockers are cultural, organizational, and strategic. A few more insights from McKinsey that hammer this home:
- Leadership inertia: 47% of executives report their companies are developing AI tools too slowly.
- Data silos and governance gaps: Only 39% of C-suite leaders use benchmarks to evaluate their AI systems.
- Skill gaps: 46% of leaders identify talent skill gaps as a significant barrier to adoption.
Meanwhile, pilot programs fail to scale because they aren’t connected to real business objectives. As Hoffman writes in his earlier book on AI, Impromptu, and expands in Superagency, AI success requires superagency: amplifying human potential through systems that align technology with human goals.
This alignment doesn’t happen by accident. It requires design.
We are now entering what I call the “post-demo era” of AI. The time for playing with ChatGPT is over. This is about transforming business systems, cultures, and outcomes.
AI is not just another tool. It’s the next chapter of human agency.
As I wrote a year ago in Toward Data Science, “Generative AI is still a gamble for the enterprise. But it’s also a necessary one”. Companies that avoid this leap won’t just be late — they’ll be irrelevant.
So let’s get our data right. Let’s trust our employees. Let’s lead boldly into this new era.
