What Changes When Ideas Are Cheap?

I keep noticing how often “good enough” is offered with a shrug.

We recognize that more time might yield higher quality, but the phrase and shrug imply “If we come up with something better later, we’ll change it.” It’s a soft agreement that we’re moving on now, even if no one could clearly explain what we’re moving on from.

For a long time, that shrug was about time. We don’t have enough of it. We need to ship. We’ll clean it up later. But that’s not really the tension anymore. We have tools that can move as fast as we want them to. Speed stopped being the bottleneck.

Quality didn’t.

And quality, it turns out, is a decision. One we’re making constantly, whether we admit it or not.

I think about this a lot when people talk about “vibe working.” I understand the appeal. Less self-editing. Fewer internal gatekeepers. Just getting ideas out of your head before you kill them yourself. Most ideas fail. but not because they’re bad. They fail because starting from scratch is exhausting, and we’re our own worst enemies in those first few moments. We get that spark of a new idea, but instead of cultivating it, iterating and growing it, we self-edit. We nip that creativity in the bud right then and there.

AI changes that. It gives you a place to begin.

I love the concept shared over on Microsoft’s Worklab blog:

“With next-generation AI, the possibilities for invention and expression expand. Instead of starting with a blank page, we can chat with AI to generate options—ideas for a speech; a rough draft of a white paper; or multiple versions of a customer presentation. This doesn’t make AI a substitute for human capabilities. Instead, it augments them, enabling an entirely new way of working.” (AI: A Whole New Way of Working)

That framing matters. Because once something exists, even in a rough form, creativity gets easier. Ideas beget ideas. Iterations beget iterations. You’re no longer inventing. You’re responding. Shaping. Choosing.

In that sense, AI really is the Energizer Bunny of idea generation. It doesn’t get tired. It doesn’t overthink. It just keeps offering you something to react to.

But here’s the part that keeps nagging at me. Once everything can exist quickly, what does “done” actually mean?

Vibe working forces that question into the open. Not because it answers it, but because it removes the old excuses. When drafts appear instantly, quality can’t hide behind effort anymore. All that’s left is the standard itself.

And that’s where things get messy.

Not all work deserves the same bar. Some things should be loose by design. Early thinking. Internal notes. Explorations that are meant to spark conversation, not end it. Over-polishing those moments can actually make the work worse. It freezes ideas too early and creates false confidence.

But there are other moments where “good enough” quietly becomes dangerous. Work that leaves the room. Decisions that compound. Outputs that other people have to build on. When quality thresholds aren’t shared, everyone fills in the gaps differently. One person assumes rough is fine. Another assumes refinement is implied. No one says it out loud.

That’s how silent rework shows up.

The extra pass before something gets sent upward. The quiet cleanup before a client sees it. The internal rewrite that never gets tracked because it’s easier than having a conversation about expectations. On paper, the team moved fast. In reality, they just moved twice.

The WorkLab article had another line that really jumped out at me:

“We spent less time generating the raw material and more time applying the best of our knowledge and insight.” (WorkLab)

That’s the real trade-off. AI doesn’t eliminate work. It relocates it. From production to judgment. From creation to decision-making.

Which means quality can’t be accidental anymore.

The opportunity hiding inside all of this is that we should have more time to think about how to make the most of the ideas we have. If AI makes thinking visible sooner, then teams have to decide sooner what thinking actually matters. What level of clarity is required. What level of polish is expected. What “done” really means for this kind of work, right now.

I don’t think the answer is raising the bar everywhere. That just recreates old bottlenecks with shinier tools. I think the harder, more honest work is naming the bar. Low, medium, high. Sketch versus artifact. Work meant to provoke versus work meant to endure.

“Good enough” isn’t a compromise. It’s a choice. One that only works when everyone understands the choice being made.

AI gives us momentum. That’s the gift. But momentum without shared standards doesn’t create better work. It just creates more of it.

And maybe that’s the unfinished thought worth sitting with on this bitterly cold, ice-stormy Sunday afternoon: the future of quality isn’t about speed or tools at all. It’s about agreeing, together, on what we’re actually trying to make before we decide it’s done.