Many speculated. Many of you reached out. Here's the condensed version.

It was 4AM and I was getting out of bed to prepare for a call with someone on the other side of the world.

I could barely walk. My body felt tight, my head hurt, I felt nauseous. This feeling wasn't new — it had been there so long I'd forgotten what normal felt like.

After much encouragement, I went to see a specialist. After a series of tests, the doctor looked up and asked: "Do you experience a lot of stress in your daily life?"

"No more than usual," I said.

"Tell me about it."

After a long chat, she shared her prognosis.

I wasn't surprised. I was relieved. Relieved because it made sense. I'd accumulated so much bottled-up stress that my stomach had started eating itself from the inside out. My body was not able to process food anymore and my body was in constant fight or flight mode.

That was the moment I knew I'd been running at full speed for too long.

What broke

For fifteen years I'd prioritized work above everything else. The problem-solving, the impossible-looking challenges, the mission — I thrived on all of it. Hundred-hour weeks. Obsessing over every detail. That was just how it worked.

The same was true for Music Health. I'd poured everything into it — including all of my own money. We were trying to use music to help people with dementia. That mattered to me, deeply.

Then the belief started to crumble. Not in the mission itself. In everything around it. My body had been keeping score, and then it gave up.

What followed

In the twelve months that followed my partner and I welcomed our first child, moved countries, and rebuilt the life we actually want to live. Most of that year was startup recovery. I couldn't even think about what came next.

Slowly, my body started to heal. I started working out, setting small goals — with weights, then with running. 5km. 10km. 15km. Last month I ran a half marathon in Ibiza.

While my health was improving and time with my little family was giving me peace, I started to feel something pulling at my brain again.

The excitement to build was coming back.

The 15-year view

I've been working with AI for fifteen years. I was late to the game then, which makes it strange to watch how buzzy it has become in the last three. It's stranger still to hear the same conversations I was having with engineers fourteen years ago.

I once asked a PhD on my team to break down how our neural network was learning from our data. His answer: "We don't know. It's a bit of a black box. If we give it more data it should improve."

A lot of that is still true. The people building LLMs don't have a clean picture of how their models are training themselves or what they're capable of. And the leading problem today is the same one we had then: shit in, shit out.

LLMs have consumed 99% of the data on the internet and they still get plenty wrong. So how do we improve them? We make more data for them to learn from. Who's making the data? The AI is.

You start to see the issue.

Zoom into a smaller dataset and it gets clearer. Music. The dataset is tiny compared to the rest of the internet. To improve the AI music models, more music gets generated by AI for the AI to learn from. The output? More content. More content that all sounds the same.

Here's what I keep coming back to. Two things are true at once:

AI is replacing millions of people in the workforce. Right now. I've sat at the table with founders of billion-dollar businesses and small businesses alike. Sometimes the AI system costs more than the employee did. Their words: "less hassle." No culture mismatch. No sick days. Harsh but true.

The cost of bringing a product to market is collapsing from human time to token cost. Simple products cost less than $10 in API calls. But agentic systems replacing humans? I've seen basic agents costing companies $25K a month and more. If that replaces a senior engineer earning more than that, the math works. If it replaces a $10K-a-month salaried role, the math is upside down.

Add those two together and you get the thing nobody's pricing in. AI could erode the middle class — the part of the economy that does most of the spending. When the workforce is made redundant, who's left to buy what the AI is building?

Where the opportunity is

I've never been a doom-and-gloom person. I see opportunity everywhere and this moment is no exception.

Every single business can benefit from some form of AI. Most are using it badly. In order for AI to work for your business you need proper systems. The people who actually understand these systems have a near-infinite funnel of customers to help. A few of them are quietly building 7- and 8-figure businesses doing exactly that.

While I was figuring out what to build next I was consulting. Helping founders separate hype from what AI is genuinely useful for, then implementing it so it actually gets used. But something curious kept happening. The conversations would start with AI and end with strategy. Where the business was really stuck wasn't the tooling — it was the foundation underneath it.

I've failed more times than I've succeeded. That's why I've been successful. 16 years behind the DJ booth. 15 years building tech companies. Raising money in markets that didn't want to give it. Building things that weren't supposed to work. Securing partnerships nobody thought were possible. Every win came after a hundred quiet losses.

As Seth Rogen put it on Diary of a CEO: "The only way to mitigate not being successful is to not quit."

Turns out the pattern recognition that came from all of that was useful to the people I worked with. What started as a temporary thing turned into serving thirty businesses, then a system I now call the FOG® Diagnostic to support the next 100.

The FOG® Diagnostic. It ended up being the biggest value I could bring to my clients, so I gave it a proper structure and system. We work through what phase the business is really in, and why they're stuck at Foundation, Optimize, or Growth. They leave with a FOG Map they can execute against. Read more →

Optimize Labs. When the work needs a team — AI systems, implementation, occasional custom builds — that's Optimize Labs. 9 times out of 10 there's already a software solution for what you need. In the rare case there isn't, our team builds it. We've helped charter companies, dental clinics, real estate agents, and coaching businesses install systems that let them grow.

I've spent fifteen years building tech. This is the first year I'm actually turning every idea I have into a launched product. Three so far, all with free tiers. I built each one because I needed it.

  • mywebsitesucks.io — a free audit of your site for the age of AI. I run it on every site I own.
  • bigmoneyideas.ai — an idea validator. Every idea I have goes through it first. 147 validated already.
  • first100 — finds where your audience actually hangs out. Because ads are saturated and AI agents are eating the rest.

Why I'm doing this in public

When I was DJing I always felt like I was working behind the scenes. As an operator I've defaulted to the same posture. This new direction doesn't give me that option.

So I'm leaning into the discomfort. Social media. A weekly WhatsApp newsletter. This blog. I'll make a million mistakes along the way. The products will break sometimes, but I show up every day. That's the process.

I'm building this from Ibiza — back on the island I grew up on, 17 years after I left it. My baby is asleep in the other room. The work fits around the life now, not the other way around.

If any of it gives you the nudge to start your thing, that's worth it.

One last thing

There's an essay by James Baldwin called The Artist's Struggle for Integrity that has shaped how I think for years. The takeaway, paraphrased: the story of who we are as a society right now isn't best told by politicians, religious leaders, or business executives. It's told by the creatives — the writers, the speakers, the painters, the singers.

In an age of AI-driven misinformation, that's worth holding onto.

Be genuine. Be kind. Never quit.