You're Not Behind on AI. You're Just Playing the Wrong Game.
Quick take
- A viral tweet exposed a $1.4M Microsoft Copilot rollout with 1.2% adoption (47 out of 4,000 employees actually used it).
- AI isn't failing because of technology. It's failing because leaders buy tools without redesigning how work happens.
- Most organizations layer expensive AI on top of broken workflows instead of recoding their operating model.
- The future belongs to "agent bosses" who orchestrate AI workflows, not executives who just purchase software.
- 82% of leaders said 2025 was pivotal for transformation, but most refuse to touch the processes that actually need changing.
- Real transformation requires asking harder questions: What work doesn't need humans? What workflows need to die, not be optimized?
The $1.4 Million Email Summarizer No One Asked For
Let me start with a truth most leaders won't say out loud:
If AI feels overwhelming, underwhelming, or wildly overhyped in your organization, you're not failing. You're just watching expensive theater masquerade as transformation.
And nothing illustrates this better than a tweet that detonated on X with 21+ million impressions because it hit a nerve so precisely it hurt.
I'm quoting it in full because it deserves to be read slowly:
"Last quarter, I rolled out Microsoft Copilot to 4,000 employees, $30 per seat per month, $1.4 million annually. I called it digital transformation. The board loves that phrase.
They approved it in 11 minutes. No one asked what it would actually do, including me. I told everyone it would 10x productivity.
That's not a real number, but it sounds like one.
HR asked how we'd measure the 10x. I said we'd leverage analytics dashboards.
They stopped asking.
Three months later, I checked the usage reports:
47 people had opened it.
12 had used it more than once.
One of them was me. I used it to summarize an email I could have read in 30 seconds.
It took 45 seconds.
Plus the time it took to fix the hallucinations, but I called it a pilot success."
Attributed to Peter Gurnas, as discussed on The Artificial Intelligence Show, Episode #186 (Dec 16, 2025).
It's funny. It's brutal. And it's painfully familiar.
Because this isn't a Copilot problem. It's not even an AI problem.
It's a leadership problem. (No joke.)
Why AI Adoption Is Failing in Most Organizations
According to Microsoft's 2025 Work Trend Index, we're sitting on a productivity powder keg:
- 80% of the global workforce says they don't have enough time or energy to do their work
- 53% of leaders say productivity still needs to increase
- 82% of leaders said 2025 was a pivotal year to rethink strategy and operations
That tension (the gap between demand and human capacity) is exactly where AI is supposed to help.
But here's what actually happens instead:
- Company buys expensive AI tool
- Leadership announces "digital transformation"
- Tool gets dumped into existing chaotic workflows
- Adoption flatlines
- Executives blame the technology, the timing, or "change resistance"
Sound familiar?
The real problem isn't that your people don't want to use AI. It's that you gave them a Ferrari and told them to drive it through the same congested streets, with the same traffic patterns, using the same outdated GPS.
Tools don't transform companies. Operating models do. (Re-read as necessary.)
And most organizations are spending millions on tools while refusing to touch the operating model.
What Separates AI Theater From Real Transformation
Here's the mistake I see over and over again:
Leaders buy AI, announce "digital transformation," and then change absolutely nothing about how work gets done.
They add AI to the chaos instead of using it to eliminate it.
We're entering what Microsoft calls the era of the Frontier Firm (organizations built around human-agent teams, where AI doesn't just assist work on the margins). It runs entire workflows. And humans provide direction, judgment, and guardrails.
This shift changes everything.
In the old model, your value was:
- Typing faster
- Answering more emails
- Attending more meetings
- Processing more information
In the new model, your value is:
- Deciding what work shouldn't require a human at all
- Redesigning workflows instead of layering tools on top of broken processes
- Becoming what Microsoft now openly calls an "agent boss"
That last phrase is critical. An agent boss isn't someone who manages people. It's someone who orchestrates the work of AI agents, human teammates, and hybrid workflows to achieve outcomes that were previously impossible.
But becoming an agent boss requires something most executives haven't done yet: recoding how work actually flows through the organization.
Why the Viral Tweet Struck a Nerve
The tweet went viral because it exposed the quiet truth most leaders are living but not saying out loud:
Most AI "failures" aren't technical. They're cultural. They're structural. They're strategic.
If you're wondering why:
- Adoption is low
- ROI is fuzzy
- Enthusiasm fizzled after the launch email
- Usage reports are embarrassing
- The board is asking hard questions you can't answer
This is why.
AI doesn't fail because it's too advanced, too complicated, or "not ready yet."
It fails because organizations refuse to recode how work actually happens.
They want the results of transformation without the discomfort of transformation.
And that simply doesn't work.
What Actually Needs to Change (The Uncomfortable Truth)
Real AI transformation isn't about buying the right tool.
It's about asking harder questions like:
"What work are we doing that doesn't need a human anymore?"
This question makes people nervous. But avoiding it means you're wasting human intelligence on tasks that could be automated, freeing your people to do work that actually requires judgment, creativity, and strategy.
"What workflows were designed for the pre-AI era and need to be completely rebuilt?"
Most processes in most companies were built 10, 15, or 20 years ago. They're layered with workarounds, approvals, and manual steps that made sense when email was the most advanced tool in the building. Those workflows are now anchors.
"Are we training people to use AI, or are we training AI to fit into outdated ways of working?"
If your AI implementation strategy involves "change management" but not "workflow redesign," you're doing it backwards.
"Who in our organization has the authority to kill workflows, not just optimize them?"
Because transformation isn't about making bad processes 10% faster. It's about eliminating them entirely.
These questions are hard. They expose inefficiencies, challenge sacred cows, and force difficult conversations.
But they're also the only way forward.
The Path Forward: From Tool Buyer to Agent Boss
If this made you laugh and cringe a little, good. That's usually the moment before real transformation.
Because here's what I know after spending 3,000+ hours training on AI and decades scaling businesses:
The companies winning with AI right now aren't the ones with the biggest budgets.
They're the ones willing to recode their DNA.
They're asking:
- What do we stop doing?
- What do we redesign completely?
- Who becomes an agent boss, and what does that role actually look like?
- How do we measure value in a world where speed and volume are table stakes?
This is exactly what I unpack in my keynotes and leadership sessions. Not future-AI hype. Not tech-bro nonsense. But Monday-morning AI: what changes, who decides, and how to do this without burning trust, people, or money?
Because the alternative is spending $1.4 million to summarize emails slower.
And we both know you're smarter than that.
If you want to go deeper and learn how to actually lead AI transformation (not just buy tools and hope), let's talk. Learn more about my keynotes and leadership briefings at susanfrewspeaks.com.
Stay sharp.
— Susan












