The ai.com website crashed during the Super Bowl. Millions watched a commercial for a $70 million domain name. Then they tried to visit. The site went down.
Kris Marszalek spent that money. The same guy who built Crypto.com. He bought ai.com in April 2025, paid in crypto, and kept it secret for months. The plan was simple. Launch it during the biggest TV event of the year. Get everyone talking.
And people did talk. But not the way he probably wanted.
You've seen this before. A company drops millions on a domain. Promises something big. Then delivers a signup page that can't handle traffic. The ai.com launch followed that exact script. The website came back online eventually. But the damage was done. People were already joking about it on Twitter.
Why Anyone Would Pay $70 Million
The previous domain record was $30 million for Voice.com back in 2019. Marszalek more than doubled that. For two letters and a dot.
Here's what actually happens with premium domains. They're not just web addresses. They're statements. When you own crypto.com, you own the category in people's minds. When you own ai.com, you're betting the same thing will work for artificial intelligence.
But there's a catch. The seller, Arsyan Ismail, bought ai.com because it matched his initials. He held onto it for years. Watched AI go from niche computer science to mainstream hype. Then sold it for a historic amount.
That's not strategy. That's luck.
Marszalek is trying to turn luck into legitimacy. The platform promises autonomous AI agents that handle tasks for you. Not chatbots. Actual agents that book things, send messages, manage your apps. And here's the interesting part. These agents supposedly share what they learn across the network.
Sounds familiar if you've been in crypto. Decentralized learning. Shared improvements. The same language Marszalek used to sell blockchain.
The Timing Problem
i used to think domain names didn't matter anymore. Google killed that idea, right? You don't need a perfect domain when search algorithms find everything.
Wrong.
Premium domains still work because they skip the search step entirely.
When someone types ai.com into their browser, they're not comparing options. They're not reading reviews. They're just there. That's worth $70 million if you're building a consumer AI platform.
But timing matters. Marszalek launched during a weird moment for AI. Reddit threads are full of people saying they're tired of the hype. One post from January 2026 starts with "Honestly, I'm getting weary of the 'AI will change everything tomorrow' hype".
Another thread calls the whole AI boom bullshit. These aren't fringe opinions. They're getting hundreds of upvotes.
The problem isn't what you think. AI tools work. They're useful. But the marketing around them has gotten so loud that people tune it out. Every week there's a new model. A new benchmark. A new miracle feature. And most of it doesn't change how regular people work.
So when ai.com launches with grand promises about autonomous agents, people are skeptical. They've heard this before.
What's Actually Launching
Here's a question people always ask. What does ai.com even do?
The website describes it as a personal AI agent platform. You sign up. Create an agent. It handles your digital tasks. Scheduling, messaging, building projects, all of that.
The pitch is that your agent learns. Then shares that knowledge with other agents on the platform. So everyone's agent gets smarter over time.
That's the theory. Reality is messier.
First time users tried the site, it crashed. That's not surprising for a Super Bowl launch. But it's also not great when you're selling automation and reliability.
Second, no one knows what this costs. The press release doesn't mention pricing. The website that came back online just had signups. People immediately started asking about privacy, functionality, and what they're actually paying for.
Third, the term "autonomous agent" means different things to different people. To developers, it means something specific. To consumers watching a Super Bowl ad, it probably just sounds like AI magic.
Most people don't know what they're signing up for yet.
And that's intentional. The launch was about awareness. Not details. Get the domain in people's heads. Figure out the product later.
Premium Domains and Pet Rocks
My friend collects domain names. Not premium ones. Just weird ones he finds funny.
He owns about 30 domains. Most of them were under $15. Things like "sadtoast.com" and "blurrywhales.net." He's never built a site on any of them. He just likes owning them.
When i told him about the ai.com sale, he laughed. Said it reminded him of pet rocks. Remember those? Someone sold rocks as pets in the 1970s. Made millions.
His point was that value is arbitrary. A domain is worth what someone will pay. And right now, someone will pay a lot for AI-related anything.
But unlike pet rocks, domains can actually do something. They can become platforms. Brands. Entire ecosystems. Or they can just sit there, doing nothing, while the owner hopes the hype lasts.
What This Really Means
Let's be honest. The ai.com launch is a bet. Not on technology. On branding.
Marszalek already did this with crypto.com. He spent millions on naming rights for a stadium. Bought Super Bowl ads for that too. Built crypto.com into a recognizable brand during the crypto boom.
Now he's trying the exact same playbook with ai.com. And it might work. Or it might not.
The real question isn't whether AI agents are useful. They probably are. The question is whether anyone needs them from ai.com specifically.
There are already AI tools everywhere. ChatGPT. Claude. Gemini. Microsoft Copilot. Google AI. Every tech company has something. Most of them are free or cheap.
What does ai.com offer that those don't? Right now, no one knows. The marketing says "autonomous agents that act independently". But that's vague. And vague doesn't win in a crowded market.
This is overkill for most people. If you're just trying to organize your calendar or send emails, you don't need a $70 million domain doing it for you. You need a tool that works.
The ai.com launch matters because it shows how far the AI hype has come. A crypto CEO spending $70 million on a domain. Launching it during the Super Bowl. Crashing the website from traffic. That's peak hype cycle behavior.
It also matters because it adds pressure. When you spend that much, you have to deliver. People expect substance. Not just a signup page.
The Part Where It Gets Weird
The Super Bowl ad aired in the fourth quarter. By then, most people were watching for the commercials anyway. The game was low-scoring and kind of boring.
So the ai.com ad had a captive audience. And what did it show? Not much. Just the domain name. A tagline about AI agents. A call to visit the site.
Then the site crashed. Marszalek had to post on social media acknowledging it. Said they were working on it. That's not ideal for a platform promising to handle your digital life seamlessly.
But here's what's weird. The crash might have helped.
When something goes down because too many people tried to use it, that's social proof. It means demand was real. People actually cared enough to visit. In a strange way, the crash validated the $70 million bet.
i still think about that. How failure can look like success if it happens the right way.