I've just come back from a three-day marathon of rebuilding our website. Three days. Twelve hours a day.
The goal wasn't simply to launch a new website. I wanted to answer a question that seems to be everywhere right now:
How easy is it to build a professional website with AI?
At the same time, I wanted to experiment with new SEO strategies built around AI-generated content and AI-assisted workflows. The internet is full of people showcasing what Claude, Cursor, and other AI models can supposedly do, but there is still very little discussion about what the process actually looks like when you're building something real.
Three days later, the website is live: https://metanow.com
Whether the design is good or not is for you to decide. As for the SEO and AI optimization (AIO), that will take time. Ask me again in six months.
Was it worth it?
After spending roughly $250 on Cursor AI tokens and pushing Claude MAX to its limits, my answer is yes.
But probably not for the reason most people think.
Despite what many social media posts suggest, AI is not a magic button. You don't write a single prompt, sit back, and watch a complete website appear.
The reality is much more hands-on.
You still need to think about structure, user experience, messaging, branding, technical implementation, and hundreds of small decisions. AI accelerates the work, but it doesn't replace the creative process.
In many ways, it becomes a collaboration between you and the AI.
The biggest advantage
What impressed me most wasn't that AI replaced developers or designers.
It was the speed.
In three days, I was able to go from concept to a finished website.
If I had followed a traditional process, I would probably have needed to:
- Find the right agency or freelancers.
- Prepare a detailed brief.
- Go through multiple design revisions.
- Wait for development.
- Review milestones.
- Request changes.
- Hope the final result matched the original vision.
Not to mention the cost.
Instead, I stayed in complete control of every decision while moving at a pace that simply wasn't possible a year or two ago.
AI still isn't automatic
One thing became very clear throughout this experiment.
AI isn't replacing expertise. It's amplifying it.
The better you understand marketing, design, SEO, development, and your business, the better results you'll get from AI.
Without that foundation, it's easy to generate lots of content that looks impressive but ultimately doesn't solve the real problem.
Now comes the real test
The website is finished.
Now the interesting part begins.
Will the SEO strategy work?
Will AI-assisted content perform better?
Will search engines reward this new approach?
Those questions can't be answered today.
I'll have a much better picture in about six months.
The tools I used
The stack for this project included:
- Cursor
- Claude Code (Claude MAX)
- Pantalytics
- Odoo
- GitHub
Each played a different role, but together they made it possible to build, iterate, and launch far faster than I expected.
Final thoughts
I'm convinced AI is changing how websites—and many other digital products—will be built.
Not because it removes humans from the process, but because it allows one person to achieve what previously required an entire team.
If you're experimenting with similar workflows, whether in web development, marketing, software, or another industry, I'd genuinely like to hear about it.
We're all figuring out where this technology is heading, and I think we're only seeing the beginning.
Comments