Moving at the Speed of AI: The Reckoning

On January 4, 2025, I built my first AI-powered, AI-native application in a weekend. At the time, I had a major constraint in my creative building process: Character asset creation and transformation.

You see, a big part of MuseKat, my family-friendly learning app is this cute, curious meerkat mascot that I invented with AI named Miko. Over the past two months, the rapid technological advancement with AI have helped me completely blow past this constraint. This experience felt like a striking (if not scary) analogy for what it takes just to keep up in today's technological tsunami.

But before we get into all of that, let's bring it back to Miko.

Here's what Miko looked like in January 2025, when I first invented her in partnership with Flux.

Back in January, when I asked AI-powered image generation tools online to help me transform Miko into other poses, or carrying different objects, I got results like this:

Cute meerkats. But not exactly brand consistent assets that I could reliably use or repurpose.

When I tried to take my exact character and put her in different backgrounds, like parks or zoos, I'd get an entirely different meerkat altogether:

Still cute. Still furry. Still not reliable.

Well, last week, ChatGPT's release of its 4o image generation service changed the game when it comes to effective and reliable image generation. The first thing I did was take my current meerkat and ask for a version that as a little simplified and on a transparent background. Here's what I got:

Not bad. Having spent some time studying other animated characters since January, I could tell the version on the right conveyed a bit more simplicity, dynamicism, and play.

The real test however, would be running it by Miko's biggest superfan: My 2-year-old.

"Who's this?" I asked her, showing the new and improved meerkat.

"Miko!!" she shouted gleefully.

Excellent. Now onto step two. Character transformations. I immediately asked for Miko in a variety of poses and set against different backgrounds.

Not bad.

But the real fun for me came when I started co-prompting some of this Miko play with my 4-year-old. I asked her, "What would like to see Miko do?"

She said: "Dance!"

From there, we went on a bender of Miko-oriented AI-generated games. We made songs about Miko on Suno. We put those songs to animation with Hedra. We made up stories together.

Then, we did something even crazier. I created AI-generated images of my kids and put them into the same cartoon world as their new meerkat friend. Here's what happened:

That in just two months, something that was impossible is suddenly immediately available overnight is both terrifying and inspiring. Exciting, isn't it, that I can now try my hat at brand design, bolstered by these new tools. Terrifying, isn't it, that the human designers I would have relied upon even as recently as last week suddenly have a different role to play.

There have been two weekends when I had nightmares about technology's takeover with society. The first was the weekend when ChatGPT first came out (in November 2022). The second was this past weekend.

After all, it's only a matter of time until whatever I'm building gets disrupted too. As a founder today, my only options seem to be: Keep up, or keep out.

Of course, this feeling isn't just limited to those of us building (or attempting to build) on the bleeding edge of innovation. It's a feeling for anyone who touches a computer in their work or lives. In other words: All of us.

When I woke up this morning, I read my friend Saadiq's blog post titled, The AI Reckoning is Here: Our Careers and Companies Need an Immediate Reset. In it, he talks about the so-called "new work reality" that's slowly (and quickly!) settling upon society right now. It's an "adapt or die" kind of mentality right now.

That a single person with AI can work at the pace of two humans without it is astounding. But there's a double-edged sword to this superpower: Completely skewed expectations. For the AI-fluent among us, it's no longer good enough to just work at the speed of a single human. If you can't double your typical output, you can't keep up in organizations that have AI-charged every aspect of their business. Never mind the learning curve that's required to achieve such productivity gains; if you can't keep up, you're getting kicked out.

Maybe that's why part of why this technology is already tearing companies apart at the seams. Nobody knows how to manage the onslaught, how to navigate the pressure, how to hire, how to plan, how to prepare. A question I'm starting to ask is, "How am I going to reliably make money, for the rest of my career? How many backup plans do I need to not get erased from relevance?"

Saadiq's theory is that entrepreneurship is perhaps the only defensible path right now. He's not alone. And I can't say that I disagree.

Those among us who have earned through experience the hard-worn habit of continual reinvention know what it takes to constantly pivot and move with agility through the world. But it's hard enough to do this on your own; even harder to move with a group. In that way, right now, it's weirdly easier to run a company as a solopreneur. After all, when your human team can't keep up, your AI agents sure can.

While entrepreneurial thinking seems to be a must-have to survive in this Darwinistic take on technological revolutions, it's certainly not enough to thrive. Every day invites another possibility for disruption. A business model that can be completely sucked up by a megalith with billions more dollars than you, or access to more training data, more compute power, or more human agents to put to the task of figuring things out. Sure, the chase may be fun for awhile. But eventually we may grow tired. What may result is people opting out of certain cycles entirely, leaving our present-day society to feel more like a time capsule of experiences, imprinted with where you got left behind.

Suffice it to say, I have a feeling it’s going to take more than cute meerkat renderings—no matter how charming—to make my app stand a real chance in the long term. But hey, at this rate, I'll just be happy it it survives longer than a week.