We built an online arcade using only AI: Here's why (and how)

Tl;dr: AI-generated content has exploded by 8,362% since ChatGPT's launch, making standing out through writing nearly impossible. Instead of adding to the noise, I built an entire arcade using only AI tools to create something genuinely worth talking about. This experiment proved marketers can use AI as an engineering partner to build interactive experiences without dev resources. The differentiator today isn't who writes best—it's who creates the most interesting experiences worth sharing.

Creating written content has never been easier. Anyone can pump out a decent blog post with a few prompts—heck, anyone can create THOUSANDS of posts with a simple prompt and some cash. The result? A flood of content drowning the internet.

AI-generated content has exploded by 8,362% after ChatGPT-3.5 launched in November 2022. Some experts estimate that as much as 90 percent of online content may be synthetically generated by 2026. I edited and wrote parts of this post with AI. Some sentences aren't my own ideas. That's just our world now.

Standing out with writing is harder than ever. Being a good writer used to set you or your company apart. Now everyone's at least "ok" at writing if they bother to edit the AI output. You need to be not just good, but interesting—and even that's not enough anymore.

To make matters worse, this content flood is killing a reliable growth channel: SEO. It's not dead, but it's dying. Keyword competition is nearly impossible because people aren't limited by... people anymore. You don't need a content team—just one person with a ChatGPT Pro subscription can run an entire blog.

Elena Verna recently published "Company blogs are no longer worth the investment," citing alarming traffic drops from companies who built their names through blogging (HubSpot, Gong, etc.).

While her headline is deliberately provocative, her core point hits home: "The brands that win will adapt, experiment, and figure out how to meet people where they actually are. The question isn't whether blogging still works. It's whether your content is built for the internet of today – or the one that existed 10 years ago."

The problem: Content isn't enough anymore

As a marketing lead at a small startup like Clarify, I'm constantly banging my head against the wall trying to figure out how we can stand out without existing brand recognition and without strong written content as a differentiator. It keeps me up at night. Literally.

I don't have the perfect answer (sorry if you came here looking for a silver bullet), but I do have a hypothesis that's been rattling around in my brain: marketers should spend time doing things worth talking about rather than just creating content. Ryan Law frames this perfectly, comparing it to Hunter S. Thompson's "gonzo journalism."

Law writes, "as Thompson proved, short, intense periods of firsthand experience can yield incredible insights and stories. So what would happen if you radically reduced your content output and dedicated half of your content team's time to research and experimentation? If their job was doing things worth writing about, instead of just writing?"

Credit: Ryan Law at Ahrefs

Simply put: to create good content, you have to be interesting. To be interesting, you need good subject matter. To get good subject matter, you actually have to do shit worth talking about.

The opportunity: Using AI as an engineering partner rather than a replacement

This is where AI offers opportunities instead of taking them away. Marketers have always been held back by code. Want to build a tool? Fight for engineering resources. Want to create landing pages for an ABM test? Hope your web team approves it, or worse, pray your CMS isn't hard-coded.

But now, we're not constrained.

So we asked: what could we create with AI that's interesting, different, and bold? The answer came through our early culture. With many gamers on our team, what if I could create a simple online Clarify arcade using only AI?

Spoiler alert: we did. And I'm going to walk you through exactly how.

The process: Building interactive games without coding experience

Getting Past the Doubt Gap

The first thing I needed to do was get over the doubt gap: inherently dismissing that it would be possible to build games with AI. That, in my opinion, is the biggest reason people haven't seen what's possible with AI. They either don't want to or don't even try.

It's the same mindset trap we've always fallen into with new technology:

• "AI can write a blog post, but it can't be creative" → Then it wrote poetry
• "AI can write poetry, but it can't code" → Then it started coding
• "AI can code simple things but not complex applications" → Yet here we are

The doubt gap is a moving target of our own limitations, not the technology's.

Game #1: Starcraft CRM

When I set out to create the first game, I wanted players to experience the evolution of CRMs through the ages - a story that paralleled our CRM manifesto:

• The humble beginnings with simple rolodexes
• Today's challenges that demand reimagined relationship management
• Our vision for a truly modern CRM built around core guiding principles

From Haunted Houses to Space Battles

My first idea was a haunted house where every room was a different legacy CRM. The development journey was anything but linear. I asked Claude how to prompt it to build this game, and it gave me detailed instructions. What it produced wasn't perfect (heck, it wasn't even functional), but it got me close enough to know that it was possible. In other words, this was my "aha" moment.

But something about it didn't quite click with our brand voice.

That's when the real iteration began:

I started with that haunted house concept, then experimented with a Pokémon-style movement system. Still not quite right. Then inspiration struck - what about StarCraft? Being a huge part of Clarify's early culture (we've got some serious gamers here), it felt like the perfect thematic fit.

The StarCraft reimagining brought everything together. I added twinkling stars and planets for immersion and designed unique door graphics for each CRM era. The universe was taking shape.

Gameplay With Purpose

An interactive experience needs more than just aesthetics - it needs mechanics that reinforce the message. I created enemy ships that represented the problems with legacy CRMs, gave players defensive capabilities (just hit spacebar to shoot!), and implemented satisfying explosion animations when defeating these outdated systems.

The final touch was a cohesive start screen that tied everything back to our brand message.

What emerged was more than just a game - it was a StarCraft-themed journey through CRM history that combined educational content with genuinely fun gameplay elements, all while telling our story in a way that resonated.

Game #2: CRM vs CDP pong battle

For my second game, I wanted something more straightforward yet equally impactful - a reimagined version of Pong where CRMs and CDPs face off across the digital court, battling for control of data.

The Beginning Serve

The concept was elegantly simple. I started with a clear prompt: "Create a simple Atari Pong game where one side is CRM, other side is CDP and the 'ball' is data. First to 7 wins, and you can pick easy, medium, hard."

This classic arcade format provided the perfect canvas - immediately recognizable, yet ripe for storytelling through gameplay.

From Game to Metaphor

Once the basic mechanics were working, I transformed it into something that carried our message:

I added contextual commentary about how CDPs and CRMs shouldn't be locked in competition, but rather working together. The game itself became a perfect illustration of the inefficiency in today's fragmented data landscape.

The end screens weren't just "Game Over" moments - they connected directly to our product's value proposition, showing how Clarify resolves this unnecessary battle.

I carefully calibrated three difficulty levels to ensure the game remained challenging yet winnable, and designed intuitive onboarding screens that explained both the controls and the underlying concept.

What began as a simple Pong clone evolved into a powerful metaphor for the very problem our product solves - turning a fun diversion into a memorable message about data integration.

Leveling up: From individual games to a complete arcade

But individual games weren't enough. I wanted to create a true arcade experience—something cohesive and immersive. That's when I discovered V0.

Discovering more advanced tools: Enter V0

If you haven't played with V0 yet, you're missing out on one of the most mind-blowing AI tools available right now. While Claude is exceptional for certain tasks, V0 is specifically designed for building interfaces and functional applications.

I started with a stupidly simple request: "I want to build an interactive landing page that lets you select different arcade games that I've coded elsewhere."

That's it. No elaborate prompt engineering. No technical specifications. Just a simple request in plain English.

Holy moly, what happened next was impressive. V0 immediately grasped what I was trying to do and started building:

What I loved about working with V0 was how conversational the process felt:

• Me: "The INSERT COIN button isn't clear enough"
• V0: Makes it larger and adds animation
• Me: "The background texture is making the text hard to read"
• V0: Adjusts contrast and adds drop shadow

I found myself slipping into real creative direction mode, suggesting themed games like "RevTech Stack Tetris" where the blocks would represent different marketing technologies. V0 not only understood these requests but executed on them creatively.

By the end, we had a working arcade landing page with four built out games.

Taking it live: From prototype to production

This is when I had to call in the cavalry. I worked with our engineers to copy the codebase and merge it into production. I felt a little guilty asking them to help with my weird AI experiment, but when they saw what I'd built, they got genuinely excited.

One of them literally said, "Wait, YOU built this? With AI?"

We did a little bit of UX work to make it closer to brand, but honestly, not that much. AI had already done 90% of the heavy lifting.

The final result? A fully functional arcade with themed games that tell the Clarify story in an interactive, engaging way. No marketing fluff, no SEO-optimized blog posts that sound like everyone else's—just a genuinely interesting experience that gives people something to talk about.

The bigger takeaway: Creating experiences worth talking about

The point here isn't that every marketer should go build an arcade (though honestly, why not?). The point is that in a world drowning in AI-generated content, the way to stand out isn't to create more content—it's to create experiences that give you something worth writing about.

This experiment taught me several important lessons:

1. The doubt gap is the biggest barrier

Most people don't even try because they assume AI can't do something. Start with the assumption that it can. I'm not special—I just asked, "I wonder if this is possible?" and then actually tried it.

2. Iteration is everything

None of my games worked perfectly on the first try. They were buggy, ugly, and sometimes straight-up non-functional. It was the back-and-forth, the specific feedback, and the willingness to experiment that got us there. I must have gone through 20+ iterations on each game.

3. Different tools have different strengths

Claude was great for detailed games with specific mechanics, V0 excelled at creating the arcade interface and visual elements, and our engineers helped bring it all together. Knowing which tool to use for which job is half the battle.

4. The end result isn't just content—it's a story

Now when I talk about how AI is changing marketing, I don't have to speak in theoretical terms. I have a real experience to share: "Let me tell you about the time I built an entire arcade using just AI..." is a much better conversation starter than "According to a recent study on AI in marketing..."

What will you build?

As marketers in 2025, we need to recognize that creating content is table stakes. Everyone can do it, and AI makes it easier than ever. The differentiator isn't who can write the best blog post—it's who can create the most interesting experiences.

So I challenge you: What weird, impossible thing could you build with AI that would be worth talking about? What experiences could you create that would make your content stand out not because it's optimized for search engines, but because it's genuinely interesting?

The brands that win won't just adapt to the content flood—they'll rise above it by doing things worth writing about in the first place.

Don't just write about what everyone else is writing about. Build something. Create an experience. Do something worth telling a story about.

Then tell that story.