Optimize vs Reimagine: Why the Companies Winning with AI Are Rethinking the Work Itself

Most companies initially approach AI the same way: find a process, add AI, make it faster. That’s a reasonable starting point. But it’s not where the value is. The companies seeing bottom-line impact from AI aren’t just making existing work faster. They’re rethinking how the work gets done.

The difference

Optimizing asks: how do we speed this up?

Reimagining asks: If we were designing this today — knowing what AI can do — what would it look like?

That second question opens up a completely different set of answers. It’s not about adding a tool to an existing workflows. It’s about questioning whether the workflow itself is sitll the right one.

The Uber lesson

In a recent conversation with Kara Swisher, Uber CEO Dara Khosrowshahi walked through how this played out in their customer service operation — hundreds of millions of interactions a year.

Their first move was the one most companies make. They layered AI on top of the existing agent workflow. The AI would research the issue, check the facts, and recommend a resolution. The agent would review and act.

It backfired. The AI was wrong about 5% of the time — enough that agents stopped trusting it. Instead of saving time, they did double work: reading the AI’s output, then doing their own research anyway. Same process, same steps, same roles — just with AI bolted on.

What worked was reimagining the work itself. Uber gave AI full ownership of low-stakes interactions. Then came the real breakthrough: they stopped giving AI rigid rules and instead gave it general guidance — treat the customer well, check the facts, use good judgement. That approach produced the best results.

The version that worked looked nothing like the original process.

It’s a people exercise

Reimagining work isn’t a technology project. It starts with the people doing the work every day — they see things leadership doesn’t. Where time gets wasted, where judgement actually matters, and where it doesn’t.

But what separates reimagining from optimization is what you do with those insights. Optimization streamlines what exists. Reimagining asks: should this work exist in its current form at all? Does a human need to be in the loop? What changes if we design around outcomes instead of steps?

Uber’s optimization move would have been fixing the AI’s accuracy so agents trusted it. Their reimagination move was asking why a human was in the loop at all for a $12 refund.

Three signs your team is ready to think bigger

You’ve already seen the easy wins. Your team is using LLMs as copilots — drafting emails faster, summarizing documents, streamlining day-to-day tasks. That’s a good start. But now you’re wondering: is there a bigger opportunity here? That instinct is right.

Your team is asking better questions. When the people doing the work start saying “why do we do it this way?” instead of “can we do it faster?” — that’s the shift. They’re seeing possibilities that the current workflow can’t capture.

You’re thinking about outcomes, not just efficiency. Speed is a fine goal. But the real question is whether the work itself is designed for the results you actually want. AI gives you the chance to redesign around outcomes — not just optimize for time.

At MindFrame Partners, we help companies hmake this shift — from automating what’s there to reimagining what should be. If you’re wondering where to start, get in touch: info@mindframe-partners.com