AI Over 40 Series - Week 14: How AI Moved from Experimentation to Expectation

For the first thirteen weeks of this series, I focused on my personal AI journey—my frustrations, breakthroughs, and hard-won literacy. But while I was learning AI one prompt at a time, something larger was happening across our organization.

We were conducting an experiment without realizing it: What happens when you encourage AI adoption without expecting it?

The answer was clear. After more than a year of encouragement, only a third of our developers consistently used AI. Nearly half had tried it but stopped using it. A fifth had never tried it at all. That’s when it became apparent: encouragement isn’t a strategy. It’s delegation disguised as empowerment.

When the models changed, the stakes changed

In early 2025, OpenAI, Anthropic, and Google released model updates that significantly expanded the capabilities of AI. The tools developers tried in 2023—limited, brittle, easily dismissed—were nothing like the systems available in 2025. These new models understood entire codebases, delivered accurate suggestions, and accelerated work instead of interrupting it.

Suddenly, the gap between AI-enhanced developers and those without modern tools was no longer abstract. It was measurable—and costly.

The question changed from “Will they use AI if we encourage them?” to “Can we afford developers who don’t?”

Moving from encouragement to expectation

1. Visibility

We switched from reimbursed individual Copilot subscriptions to team subscriptions. It doubled the cost but gave us what we lacked: visibility. We could finally see usage patterns, model access, and engagement.

That’s when we discovered some developers were still using free plans to “save money,” unintentionally shutting themselves out of the most capable models.

2. Addressing Real Objections

Visibility allowed for real conversations:

• For those who tried early and stopped: Their experience was with outdated models. They needed a fresh start.
• For those who had never tried: We set expectations, accounts, and deadlines.
• For the resistant few: We were honest—AI-enhanced development is now table stakes.

3. The “No Time” Illusion

One developer told me he’d try AI once he “had time.” After a single afternoon with the latest models, he called back to say AI had immediately given him the time he thought he needed to begin.

That’s the paradox: AI solves the very problem delaying its adoption.

What happens when you hit critical mass

When AI use became expected instead of optional, our learning curve accelerated. Developers began sharing prompts, workflows, and patterns. They built an organizational MCP server with our coding standards. They used Microsoft Learn MCP to compare custom work to base Business Central functionality. They indexed years of solutions, so no one has to start from scratch.

The transformation became collaborative—and exponentially faster.

The ripple effect beyond development

Other teams followed quickly:

• Consultants use AI to navigate Microsoft’s dense documentation conversationally.
• Business Analysts generate complete, industry-specific discovery questions in seconds.
• Project Managers use AI with Monday.com to see patterns across all active projects simultaneously.
• Sales uses AI-powered call analysis to improve listening, engagement, and client dialogue.

Adoption is no longer top-down. It’s contagious.

The ROI that matters

The ROI isn’t about calculating minutes saved. It’s about raising the performance baseline.

AI-enhanced employees produce higher-quality work, learn faster, and tackle problems we once declined. Juniors ramp faster. Seniors experiment more. And the cost—$20 per seat per month—is trivial compared to the productivity gap between those using AI and those working without it.

This isn’t about efficiency. It’s about competitiveness.

The leadership lesson

I spent thirteen weeks documenting my own journey because you cannot lead a transformation you haven’t experienced. You can’t mandate what you don’t understand. And you can’t meaningfully coach someone through challenges you’ve never faced yourself.

Your Week 14 Challenge

1. Count your team: Who uses AI regularly, irregularly, or not at all?
2. Pick one person: Talk for 15 minutes with someone whose early AI experience was disappointing.
3. Calculate one gap: What would 30 reclaimed minutes per day mean for your team?
4. Set a date: When will AI move from experimentation to expectation?

The goal isn’t full transformation—yet. It’s to stop pretending encouragement is enough.

The Bottom Line

Shifting from optional to expected wasn’t comfortable. It required investment, transparency, and hard conversations. But once AI-enhanced work became the norm, everything changed.

The real experiment wasn’t whether AI could transform our work. It was whether we were willing to lead the transformation.

And that starts with personal transformation first.

This post is part of my “AI Over 40” series. It first appeared on LinkedIn: AI for the Over 40 [Week 14]: When Personal Transformation Becomes Organizational Expectation.

Next Up: Why No One Can Show You How Agentic AI is Going to Transform Your Business.

Read more AI and Copilot blogs.

Trending Posts