I came back from Davos with a familiar feeling, and it was reinforced by a World Economic Forum article published recently: “Davos 2026: Leaders on why scaling AI still feels hard - and what to do about it.”

The point is straightforward: even as AI adoption spreads, scaling AI beyond pilots still feels hard, and the “hard part” is not only technical. Moving past early wins requires new strategies, capabilities, and organizational designs.  

That is why this thought followed me home: it’s a year later, and we are still having the same conversations about AI.

The good news is that this is a problem we can now solve. The path forward is clearer than it was a year ago. In many ways, the organizations that move now can still make up a lot of ground, arguably more than they would have last year. But it requires a change in mindset.

Déjà vu and what it says about the AI maturity curve

The most surprising takeaway from Davos was not a breakthrough idea or a new model. It was déjà vu.

Many conversations sounded like the same conversations we were having a year ago. Leaders are still asking foundational questions about where to begin, what matters, how to build confidence, and how to create business impact without creating new risk.

That tells me something important about the AI maturity curve inside most organizations. The technology has moved fast. Organizational adoption has moved slower. The headlines make it sound like everyone is either far along or completely failing. The reality is more uneven, and for many companies, it’s still early.

I see this especially in the mid-market. These organizations are arguably best positioned to create impact because they can move quickly, stay close to the work, and make decisions without the size that slows change in the largest enterprises. The opportunity is meaningful, but it still requires a deliberate approach.

So why are we still stuck in the same place?

Why we are still having last year’s conversation

The “stuck” moments I see are rarely about ambition. They are about fragmentation.

Many organizations still treat AI as separate.

Separate from how work gets done. Separate from the operating model. Separate from how decisions are made. Separate from team design and accountability.

At the leadership level, that separation often shows up as parallel tracks:

• A technology conversation about tools, platforms, vendors, and pilots

• A people conversation about roles, skills, and org structure

Both tracks matter. But when they remain disconnected, progress stalls. You get pockets of experimentation and impressive demos that never become a durable way of working.

That is also why the WEF article resonates. Scaling AI is not simply about adopting tools. It is about changing how the organization works so the tools actually create outcomes.  

In client conversations, the questions usually come down to three practical ones:

1.) Where do we start so we can see ROI quickly?

2.) How much do we do internally versus externally?

3.) How do we make this stick for teams, beyond everyday use of ChatGPT-style prompts?

Those are practical questions. They deserve practical answers.

The mindset shift that changes everything: human + machine

This is not a new belief for me, but Davos reinforced it. It made me more emboldened about saying it plainly, because it matches what we see every day with business leaders.

The formula for success is human + machine.

Not human versus machine. Not AI as a separate initiative. Human + machine as a design principle for how work gets done.

At Quantum Rise, this design principle has informed how we think about transformation: define where judgment stays human, where machines can accelerate and recommend, and how the handoffs, accountability, and measurement work in practice.

A design principle is useful because it forces clarity. It pushes leaders to ask: where should humans lead, where should machines support, and how do we combine them so performance improves without losing accountability, trust, or good judgment?

It also helps move the conversation away from hype and anxiety and toward design choices and outcomes.

For example, during the V-LAB panel I joined at Davos, “A Global Dialogue on Humanity and AI,” the discussion naturally touched on serious questions about jobs, safety, and governance. Those concerns deserve attention. At the same time, every industrial revolution has changed work. Some tasks fade, new roles emerge, and organizations adapt.

Human + machine gives leaders a way to engage those concerns without getting trapped in extremes. It creates a shared frame for building capability responsibly while still making progress.

When leaders embrace a human + machine mindset, the conversation changes in productive ways:

• We stop treating AI as an experiment and start treating it as an operating capability

• We stop debating the tool and start redesigning the work

• We stop waiting for certainty and start building confidence through measurable iteration

There has never been a better time to feel behind on AI

If you feel behind, you are not alone. Many organizations are earlier than they think, even sophisticated ones.

There is an old proverb, often paraphrased this way: the best time to plant a tree was years ago, the second best time is today.

The best time to start building a human + machine operating model may have been earlier. The second best time is now.

Why? Because the path is clearer, the tools are better, and the early efforts can produce meaningful ROI. You do not need to solve everything to start. You need to make progress in the places that matter.

This is a great time to solve for practical adoption and measurable impact. That starts by moving AI from an isolated capability to an embedded part of how work happens.  

How leaders get unstuck in 2026

Here are three moves that create momentum:

1) Build organizational fluency

Fluency is shared understanding at the leadership level so the organization can make faster, higher quality decisions about where to apply AI, how to manage it, and what “good” looks like.

2) Start with outcomes, not tools

Pick one measurable business outcome, then work backward to the workflow and decisions that drive it. This keeps the effort grounded in results, not experimentation.

3) Redesign one workflow with a human + machine mindset

Choose one workflow that matters, clarify where humans lead and where machines support, and measure improvement. Then repeat.

Closing from Davos, looking ahead to 2027

I am grateful to the teams who convened thoughtful conversations in Davos, including Handshake Strategies and V-LAB.

My hope for Davos 2027 is not that the debates disappear. Concerns about safety, governance, and the future of work are serious, and they should remain a critical part of the conversation.

My hope is that alongside those debates, we hear more leaders sharing what they actually implemented, what they changed inside their operating model, and what they learned as they scaled. More examples of how work was redesigned as human + machine, and how that translated into outcomes people can trust.

If the thought that followed me home is that we are still having many of last year’s conversations, the optimistic counterpoint is this: we do not have to keep having them next year.

The path is clearer. The building blocks are better. And with a human + machine mindset, the maturity curve can move faster than it has in the past.