Thoughtful experimentation—not hype—will determine long-term success.

“AI is doing 30% of development work,” said one tech CEO.

This bold claim, recently made by the CEO of a major tech company, was crafted to catch headlines—and it did. It sounds like progress. It sounds like a leap forward.

But when I spoke with VPs and directors at the same organization, a different story emerged. Behind closed doors, they expressed confusion and growing pressure. The top-level messaging didn’t align with operational realities. Teams were left asking: 

The message is unclear from the top leaders.

How do we make sure the AI output is validated? What does quality look like? Who is responsible if something goes wrong? 

This disconnect isn’t unique to one company. It’s a common pattern I’ve seen across my executive coaching work with leaders from leading organizations, including those in the Fortune 50. AI is here, it’s powerful, and it’s confusing the people who are expected to deliver on its promise. 

The AI Hype Is Real—and So Is the Chaos

We’re in the midst of an AI gold rush. The pressure to “use AI for everything” is intense. Tech leaders are being told to integrate AI yesterday. Teams are being tasked with innovation, often without clear guidance, adequate resourcing, or defined success metrics.

Let’s be clear: AI is transformative. It will reshape how we work. But right now, we’re seeing a top-down push driven more by market optics than thoughtful planning.

The result? Teams are overwhelmed. Leaders are unsure whether they’re supposed to move fast and break things, or uphold the hard-earned standards of product quality and user safety.

A Crucial Leadership Disconnect

From the CEO’s perspective, promoting AI aligns with shareholder expectations. But when that same message trickles down the org chart unfiltered, it can backfire. Engineers and product teams take it literally, feeling pressured to release AI-driven features without proper validation.

For decades, engineering leaders have been trained to ship reliable, secure, and high-quality products. That discipline is second nature—and for good reason. With AI, suddenly, that ethos is being disrupted. But it shouldn’t be discarded.

As one senior leader told me, 

“It feels like we’re being asked to throw things at the wall to see what sticks—without knowing if it’s safe for the customer.”

Experimentation Is Not Randomization

If we want to embrace AI fully, we need a new mindset—one that blends experimentation with responsibility.

Here’s what experimentation means:

• It requires space to fail. Not every AI project will succeed. Teams need permission to learn through iteration, not fear being punished for imperfect outcomes.
  
• It demands curiosity and openness. Leaders must foster a culture where exploring possibilities is encouraged.
  
• It also demands accountability. Experimentation is not a hall pass for chaos. It must be grounded in ethical use, quality control, and validation.
  

Experimentation is not the opposite of quality. It’s the path to quality through learning, iteration, and ownership.

Let’s not mistake speed for strategy. There’s a fine line between moving fast and scrambling.

Why AI May Not Be Boosting Productivity Yet

A recent study by METR found that developers expected AI tools to boost productivity by 24%—but in practice, productivity dropped by 19%.

The reasons?

• Low acceptance rate: Only 44% of AI-generated code was accepted. Most required major edits.
  
• Debugging tax: Fixing flawed suggestions often took longer than writing from scratch.
  
• Context limitations: Many enterprise codebases exceed current LLM context windows.
  
• Complexity ceiling: AI still struggles with complex, multi-layered problems.
  

This is not a failure of AI—it’s a failure of expectation. AI is not magic. It’s a tool that needs proper input, structure, and governance to deliver value.

Slow Down to Speed Up

Troy Scoffield, a Senior Learning Consultant at Microsoft, shared this after training thousands of employees on using AI:

“The TL;DR? Slow down to speed up.”

His advice is clear:

• Learn and educate: Understand how AI works. Don’t just use it—train it with intentional, contextual input.
  
• Be selective: Just because AI can do something doesn’t mean it should. Choose use cases thoughtfully.
  
• Set boundaries: Define what AI will handle—and what must remain human-driven.
  

This mindset builds long-term momentum. It prevents burnout, rework, and user harm.

A Call to Tech Leaders: Clarity, Agility, and Accountability

If you’re in a leadership role—especially at the director or VP level—here’s what your teams need from you:

1. Clarity over chaos: Don’t pass down top-level hype without operational translation. Help teams understand what’s expected and what’s possible.
   
2. Room to Try, Fail, Learn: Create psychological safety for experimentation. Give teams permission to question, refine, and learn.
   
3. Accountability, not urgency: Empower your teams to deliver quality AI work at a sustainable pace.
   

We’re not in a race to launch the most AI features. We’re in a race to figure out how to use AI for good.

TL;DR: AI Is Not Just a Technology Shift—It’s a Leadership Shift

AI is forcing a redefinition of leadership itself. As Francessca Vasquez, VP at Amazon Web Services, emphasized in a recent GeekWire interview:

“Culture and talent strategies are as important as the technology. Successful AI adoption requires rethinking how teams work, not just what tools they use.”

It calls on us to move from control to curiosity, from urgency to responsibility. It invites us to slow down, build understanding, and experiment with intention, not just to move fast, but to move wisely.

We must stop assuming that a haphazard AI implementation will magically bring exponential productivity overnight. This moment is like the rise of computers in the ’90s. Back then, we knew computers would revolutionize manual work—but few grasped the full depth and breadth of their impact. We’re now facing the AI equivalent of that awakening.

To lead effectively through this transformation, we must invest in:

• Training that enables both technical fluency and ethical awareness
• Team alignment on goals, boundaries, and quality benchmarks
• Clear guardrails to ensure responsible experimentation
• And most importantly, a mindset of curiosity, reflection, and ownership

This isn’t about resisting change. It’s about meeting change with the emotional and strategic maturity it demands. Especially in high-stakes sectors like tech, aerospace, healthcare, and finance, quality and trust still matter. As I mentioned in another piece, navigating complexity and ambiguity is not foreign to the tech community. It requires an experimental, iterative mindset—one rooted in both openness and responsibility.

Call to Action: Be the Leader Your Team Needs Now

We are at a pivotal moment in history. AI will not just change how we work—it will shape who we become as leaders.

So ask yourself:

• How are you setting the right expectations around AI for your team?
  
• How are you encouraging experimentation while reinforcing accountability?
  
• How do you balance between being fast and being responsible?
  

The future will be shaped by the people who can balance innovation with integrity.

If you’re ready to help your team make this mindset shift, let’s talk.
Schedule a strategy session to explore how you can adopt and develop the mindset required for a responsible AI strategy for your organization.

Because how we lead now will define what we build next.

---

Feature image courtesy pixabay.com

Other articles: No, AI is not Making Engineers 10x as Productive, MIT report: 95% of generative AI pilots at companies are failing