Dream Big Or Get Mediocrity

Two things happened this past week that sharpened my view.

The first was a project where the output was animated content. We had a solid machine. We had an efficient workflow. We had prompts that generated great ideas and built the additional prompts we needed for characters, scenes, everything.

It was terrible.

We struggled to develop continuity. Getting the right expression from video assets or audio seemed impossible. Given our timeline, we pivoted, and in the end, we made something awesome. But we learned that yes, this is an incredibly productive tool—but it is NOT set it and forget it.

For digital products, I’ve noticed it’s difficult to do frontend well. The part you interact with needs to be clean, organized, beautiful. The irony of the human part being difficult is not lost on me. I spent days investigating different “machines” that would help me plug into Figma, pull in files, set up components, and output great work. Most of them are terrible.

So what does this mean? Are we in a bubble, and AI is just a massive waste of time?

No. But we do need to think differently about what we’re doing and how.

You should read Frank Chimero’s latest talk—Beyond the Machine: Creative Agency in the AI Landscape. You should be reading anything by Artificiality. And this excellent interview with Dylan Field from Figma on Lenny’s Podcast.

My conclusion from my own work and from these thinkers I respect: our creative vision and expectation for what is good must lead the way. Regardless of the metaphor, approach to the problem, or tools you select, you must start with the end in mind.

How is this a shift for me?

The Motherf****** MVP

I love the idea of having an idea, testing it, being pragmatic, and building only what you need when every element costs money. This is still the gold standard for how to build a product or system wisely.

But…

We need to dream more now. We need to do the opposite of what we’ve been taught for so long. If we’re going to do things that express who we are and solve problems, we can’t be satisfied with the MVP solution when we present our idea. We need to explore, use our imagination, go deeper. Push farther. Set the bar higher.

A great deal of what we can do is easily solved by AI now. This is 100% true. I had a conversation with a developer friend yesterday where we gushed over how much time we’ve saved on backend and business logic—the “hard” stuff that does the magic in a digital product. He described a problem that was solved in days instead of a month, as it would have previously taken. But then we lamented how miserable the frontend was. The part we interact with. This was easier to do by hand. Old school.

What I’m getting at: going forward, things will never be the same. Content creation, application development, problem solving—this will be powered by AI from now on. But for these things we create to actually do the work we’re asking them to do, we need to design them well. We need to establish a deep and well-designed context of what the idea is and what the right solution is.

The process from here

Step One: Imagine

This is the part I’m truly excited about. If I don’t have to spend weeks doing data schemas or hand illustrating characters, what does that free up for me? If I can predictably build the URL routing for an application and save ten hours… what can I do then?

From here, our job as makers is to imagine. In grandiose and extravagant ways. To dream about what can be done, not what should be done. To find the north star and work with your audience to design it. This is the utopian experience of developing something that can change the world. Or at least make your users’ lives easier.

Step Two: Communicate

The trick is how you communicate this vision to folks in a way that gets them excited and on board about what you’re trying to do—without the inevitable reaction of “this will be too expensive and there’s no way we can do it.” I think communicating with proofs of concept and demos of the creative effort will be incredibly useful here to show not only is it possible, but you’ve already begun. The steps between the idea and the implementation need to be guided by the vision you’re communicating.

It’s one thing to say we can build interesting and incredible things with AI. It’s another thing entirely to show your stakeholders or clients a proof of concept of something that was done in a few hours instead of weeks or months. Sharing the imagined vision and then showing folks that vision—I believe this is incredibly useful for communicating what’s possible and getting beyond the creative brief or product requirements document.

The shift

So what changes?

Everything and nothing.

The tools change. The timeline changes. What’s “hard” and what’s “easy” flip. But the need for vision doesn’t change. The need for craft doesn’t change. The need to know what good looks like—that intensifies.

AI collapsed the distance between idea and prototype. Between concept and proof. What used to take months now takes days. What used to require a team now requires imagination and clear communication.

This means we can’t hide behind “we don’t have time” or “we don’t have budget” anymore. The constraints that used to justify mediocre work—they’re dissolving. What’s left is the honest question: Do you know what good looks like? Can you see it clearly enough to build toward it?

That’s the work now. Not building the machine. Not optimizing the workflow. Not tweaking the prompts.

The work is imagining something worth building. Then communicating it so clearly that when the machine produces it in days instead of months, everyone recognizes it as the thing you described.

That’s the future I’m excited about.