My team asked if AI would replace them.
I said no.
They didn’t believe me. Honestly? I wasn’t sure I believed it either.
Here’s what I should have said:
“I don’t know exactly how this plays out. But we’re figuring it out together. And I won’t lie to you about what’s changing.”
That honesty would have built more trust than my reassuring lie.
Why This Is Different
ERP implementation was painful but straightforward. Learn the interface. Follow workflows. Train the team. Done.
AI doesn’t work that way.
AI is rewiring how work gets done. Not a tool to learn. Infrastructure that changes what’s possible.
61% of companies struggle not with AI technology, but with leadership and cultural challenges.
The technology works. Leading people through change is the problem.
AI exposes how work actually happens.
Someone took three days for analysis? AI does it in three minutes. Now you ask: what were they doing for three days? That question threatens people.
AI changes power structures.
Information control was power. Knowing where to find things was value. AI democratizes access. Power shifts. Gatekeepers feel threatened.
AI requires continuous learning.
You can’t “finish” learning AI. Exhausting.
AI forces decisions in uncertainty.
Which tools matter in six months? Which skills to develop? Expensive decisions without clear ROI.
Uncomfortable for managers trained on data-driven decisions.
You can’t avoid decisions. The decisions you make now determine whether AI amplifies capabilities or creates chaos.
Seven Decisions You Can’t Avoid
Not abstract strategy. Concrete choices in the next 90 days.
Decision 1: Lead or Let It Happen?
Are you shaping AI adoption intentionally, or letting people figure it out and dealing with mess later?
Most teams are stuck in experimentation. Haven’t decided if AI is strategic or tactical. Adoption happens chaotically—different tools, no shared practices, no learning.
Intentional approach:
Make AI a team priority. Allocate learning time. Pick 2-3 tools to standardize. Create shared practices.
Chaotic approach:
“Use AI if you want.” Some dive in. Others ignore it. No shared learning. Six months later, half the team uses tools you don’t know about. Half does everything manually.
Neither is wrong. But choose.
How to decide:
Is AI central to how you compete in two years? Or a nice productivity boost?
Central? You need intentional adoption.
Tactical? Chaos might work. Let people experiment, share what works.
Choose consciously. Don’t drift.
Decision 2: What Work Gets Redesigned First?
Wrong question: “What can AI automate?”
Right question: “What should we be doing differently?”
I learned this hard. We automated reporting taking 10 hours weekly. Great, right?
Except we never asked if we should do that reporting at all. AI made us efficient at producing reports nobody used.
Start with outcomes, not tasks.
Not: “Which tasks can AI do?”
But: “What are we trying to accomplish? What’s the best way with AI available?”
Example: Instead of “automate monthly spend report,” ask “what decisions does leadership need? What information do they actually need?”
Maybe the answer isn’t “faster reports.” Maybe it’s “real-time dashboards with AI insights when anomalies appear.”
Pick one workflow.
Don’t redesign everything. Pick something that:
- Takes significant time
- Has clear success criteria
- Affects multiple people
- Won’t break critical processes if it fails
Redesign it with AI built in from the start. Not bolted on.
Document learnings. The second redesign goes faster.
Decision 3: How Honest Will You Be?
Hardest decision. How honest about what AI means for jobs?
Top barrier to adoption: unclear value proposition. People don’t know how AI fits their work or what it means for them.
Options:
Maximally transparent: “This AI tool will eventually eliminate 30% of manual analysis. We’ll use that capacity for work we couldn’t do before. Here’s what that means for each role.”
Reassuring: “Just a productivity tool. Your role isn’t changing.”
Vague: “We’re exploring. We’ll figure out implications later.”
Truth is, you probably don’t know exactly what AI means for every role.
But you can be honest about what you know and what you don’t.
The framework:
“Here’s what I know: [be specific]
Here’s what I don’t know: [be honest]
Here’s how we’ll figure it out together: [process]
Here’s what won’t change: [core principles]”
People handle uncertainty if you’re honest. They can’t handle lies discovered later.
Decision 4: Who Picks the Tools?
Centralized control or distributed experimentation?
Centralized: You pick tools. Everyone uses the same. Standardized. Easier to manage. Slower to adapt.
Distributed: People pick their own. Faster experimentation. More innovation. Chaos when collaborating.
Most organizations land in the middle: standardize some categories (communication, project management), let people choose others (personal productivity, analysis).
Define boundaries, not prescriptions:
“Approved tools for customer data: use any of these. Don’t use tools outside this list without security review.”
“For internal analysis: use whatever AI you want. Share what works in weekly sessions.”
“For client-facing work: standardize on [tool] for consistent outputs.”
Governance prevents chaos without killing innovation.
Decision 5: How Do You Measure Success?
“Time saved” is obvious. Also mostly useless.
Saved 10 hours weekly on reporting. Great. What did you do with those 10 hours? If the answer is “other busywork,” you didn’t create value. You shifted time.
Better metrics:
Outcome improvement: Better decisions? Faster launches? Fewer errors? Higher satisfaction?
Capability expansion: Doing work you couldn’t do before? Serving customers you couldn’t serve? Analyzing problems you couldn’t analyze?
Learning velocity: Getting better at AI how fast? Sharing practices? Iterating workflows? Building knowledge?
Adoption quality: Using AI effectively or checking boxes? High-quality outputs or extensive human correction needed?
Pick 2-3 metrics that matter. Track monthly. Adjust approach based on results.
Be honest when something isn’t working. Tool seemed promising but isn’t delivering after three months? Kill it. Don’t pay for unused software.
Decision 6: What Skills Get Investment?
Can’t train everyone in everything. Choose.
AI literacy for all vs. AI specialists?
Does everyone need baseline fluency (prompts, when to use AI, when not to)? Or a few with deep expertise, everyone else just using tools?
Most teams need both. Baseline for everyone. Deep expertise for a few.
Training roadmap:
Month 1-2: Everyone gets baseline. What AI is. What it’s good at. What it’s terrible at. Prompts. Output validation.
Month 3-4: Hands-on with your actual work. Not generic examples.
Month 5-6: Advanced training for 2-3 becoming your AI experts. Evaluating tools, building workflows, helping others.
Ongoing: Monthly sessions sharing what works and what doesn’t. Continuous building.
Model the behavior. If you’re not using AI, your team won’t take it seriously.
Decision 7: How Do You Handle Failures?
AI will fail. Question is whether failure stops experimentation or becomes learning.
Someone will use AI generating confidently wrong content. Someone will share embarrassing AI output with clients. Someone will waste time on tools that don’t work.
Create psychological safety:
Make it explicit: AI experimentation includes permission to fail. Not careless failure. Thoughtful attempts that don’t work.
“Try something with AI and it doesn’t work? I want to know so we all learn. Try something and it creates a problem? Tell me immediately so we fix it together.”
Learning framework:
When something goes wrong:
- What were we trying to accomplish?
- What did we expect AI to do?
- What actually happened?
- What did we learn?
- What will we do differently?
Not: “Who screwed up?”
But: “What did we learn?”
Your 90-Day Plan
Decisions made. Now execute.
Weeks 1-4: Assessment and Transparency
Week 1: Use AI yourself. One regular task. Use AI for it. See what works and doesn’t. Can’t lead without using it.
Week 2: Talk to your team. One-on-one. What AI are they using? What’s working? What worries them? Listen more than talk.
Week 3: Make the seven decisions. Write them down. Share with team.
Week 4: Communicate transparently. Team meeting. What we’re doing with AI. Why. What it means for each person. What you don’t know yet. Questions welcome.
Weeks 5-8: Pilot and Learn
Week 5: Pick one workflow to redesign. Not most complex. Not simplest. Something meaningful.
Week 6: Implement with 2-3 people. Document everything. What works? What doesn’t? Surprises?
Week 7: Iterate. Fix what broke. Double down on what worked.
Week 8: Share with full team. What we tried. What happened. What’s next.
Weeks 9-12: Scale and Measure
Week 9: Roll out successful practices. Training on what worked. Support for adoption.
Week 10: Start measuring. Pick 2-3 metrics that matter. Baseline. Track weekly.
Week 11: Adjust based on feedback. What’s harder than expected? What needs more support? What’s working better?
Week 12: Review and plan next phase. What accomplished? Learned? Next?
At minimum, everyone needs 30% digital and AI mindset to use tools, ask questions, redesign work. That’s baseline. You’re building toward it in 90 days.
Three Fatal Mistakes
Mistake 1: Delegating AI to IT
AI isn’t an IT project. It’s work transformation using technology.
IT helps with security, infrastructure, vendors. But IT can’t redesign your workflows. Only you and your team can.
Leaders struggle with unclear AI strategy—balancing automation against reskilling and work redesign. Leadership work, not IT work.
Mistake 2: Treating AI as Efficiency Instead of Transformation
If your AI strategy is “do same work faster,” you’re missing the point.
Value isn’t speed. It’s capability.
What can you do now that you couldn’t before? What problems can you solve that weren’t solvable? What decisions weren’t possible?
Efficiency is table stakes. Transformation is competitive advantage.
Mistake 3: Moving Too Fast
Frontline managers are 3x more concerned about AI readiness than executives.
People doing the work feel change more acutely than leadership.
Move faster than they can adapt? You create resistance.
Slow down enough to bring people with you.
The adoption gap is also a leadership gap.
Close it by showing up. Invest time. Stay curious. Model behavior.
By 2026, more than 100 million workers will collaborate with robo-colleagues.
Not speculation. It’s happening.
The question isn’t whether AI transforms your team.
It’s whether you lead that transformation thoughtfully or let it happen chaotically.
These seven decisions determine which path.
Make them intentionally. Communicate honestly. Execute with your team, not to them.
AI doesn’t replace leadership.
It exposes it.
Leaders who understand that—who focus on judgment, trust, communication, continuous learning—their teams will thrive.
The rest won’t.