Why “Faster with AI” Is Slowing Teams Down
AI is supposed to save time, but leaders are not seeing the productivity boost they expect. Here’s what they’re missing.
The Productivity Pressure
I’ve always felt the pressure to make my teams more productive.
Sometimes it came from my leaders demanding better results, faster delivery, and more impact.
Other times, it was internal. A quiet voice whispering: “You can do better.”
When I led engineering teams at Google, Roblox, and later at Benching, I poured energy into refining collaboration processes. I believed better collaboration would unlock better decisions, leading to faster execution.
And it did… until the bar moved higher.
The AI Shift in Expectations
With the rise of AI, the definition of “efficient” changed.
Suddenly, it wasn’t about a 20% improvement. It was about a 10x, even 100x, acceleration.
And that pressure isn’t coming from middle management. It’s coming from the top.
Your CEO isn’t asking if you’re using AI. They’re expecting to see results.
Let’s be honest:
“We’re testing ChatGPT internally” no longer counts. Not if you can’t demonstrate real, measurable outcomes.
The Illusion of Speed
Most companies start with the obvious:
They roll out ChatGPT or Gemini and say: 👉 “Go be more productive.”
It feels like a good start until you realize what’s happening.
Chat-based AI tools give you the illusion of speed... at the cost of quality.
When you ask ChatGPT to write a PRD, it doesn’t just format your thoughts. It makes hundreds of silent decisions on your behalf.
You get a PRD of a product. But it’s not your product.
- Assumptions are hidden.
- Trade-offs aren’t called out.
- And the most dangerous: it looks pretty, polished, and professional, even if it’s empty of essence.
And when engineers build from that PRD, they build something… wrong.
→ Then rebuild → Then rebuild again...
Suddenly, your “faster” process is moving slower than ever.
This Is What Speeds Up Teams, Always
If you’re serious about using AI to improve productivity, don’t start with tools.
Start with bottlenecks.
Ask yourself:
- Where does your team actually lose time?
- Which process creates confusion or rework?
- Which activity, if done better, would make everything downstream faster?
For me, that bottleneck was always the definition of the product requirements.
Not because it took too long to write the document, but because getting it wrong wasted weeks, sometimes months, not just for me, but for the entire team.
I had the right tools: We used Confluence and AI, but we weren’t using them correctly.
What Changed
Eventually, I shifted how I thought about AI.
At first, it was just a writing assistant.
Then it became a reviewer, challenging my thinking and surfacing blind spots.
Eventually, it grew into a coach, helping me clarify the why, the what, and the edge cases that would otherwise trip us up two sprints later.
But it wasn’t just about helping me.
For AI to help larger teams, it had to meet people inside the tools they already use, like Confluence.
That’s when I saw the shift.
When teams walk through the same guided thinking process, they move faster not by skipping steps, but by thinking more clearly.
AI That Helps You Think
AI isn’t magic. It won’t save your team from chaos if your inputs are wrong.
But when applied well, AI becomes a multiplier, not a shortcut.
That’s the difference.
That’s the productivity boost we should aim for inside our Atlassian ecosystem. Using AI:
Not “generate more.”
But “think better, together.”
How We Built That into Wisary
This shift in mindset led to the building Wisary, a Confluence App that empowers product managers with AI to create better PRDs, not just create them faster.
Instead of generating entire documents for you, Wisary prompts you with critical thinking steps:
- “What’s the real goal of this feature?”
- “What constraints might your engineers push back on?”
- “How will success be measured?”
Wisary doesn’t just help PMs move faster. It helps your entire team move faster with more clarity and alignment.
If you’re curious how it works, you can learn more here.
Final Thought
Simply asking your team to “use ChatGPT” won’t unlock productivity.
Encouraging critical thinking at the PRD creation stage within Confluence?
That’s where everything downstream begins to accelerate.
I’ve seen this transformation firsthand working with teams.
Better requirements lead to better outcomes: fewer revisions, faster launches, and more aligned delivery.
Are your AI tools making your team faster… or just busier?
Was this helpful?
Thanks!
Ala _Wisary_
2 comments