Private notes, made public. · 3rd Jul, 2026

Hey, it’s Gabriel.

These are my private notes this week on building wealth, raising my family, and breaking generational patterns.

This week
⚡ I'm sharing the AI that took over my ad account
❤️ the $700 line item that protects our marriage
💰why I'm buying Tesla LEAPS before the news hits
🧬 and the only five supplements that survived years of testing

Two issues ago I mentioned I was using Fable for my health data analysis before it got pulled, and that I hoped it would come back. It's back. And I have barely slept since, and for once it isn't the newborn's fault.

I've spent the week prompting and building with it, and it lives up to the hype. Projects that had been stuck in my business for months finally moved. But one experiment stopped me cold.

I connected Fable 5 to my Meta ad account through an MCP and gave it one mega prompt: read the last 90 days of the account like a CMO would. Where the spend is concentrated. Cost per result. CTR. Which ad sets are stuck in learning phase. Then tell me what to fix.

It came back with five specific things I could do to improve the account.

I've been a marketer for 10 years. I read the output twice. It was the kind of briefing I'd expect from a senior media buyer after a full day inside the account, except it took minutes. Imagine handing 90 days of receipts to a forensic accountant and getting the verdict before your coffee cools.

Then I pushed it further. I had it write a creative brief from the winning ads, connected it to Higgsfield, and asked it to produce the next batch. It generated an image ad and a video ad. The video came back edited, with music, with direction.

I couldn't believe what I was watching.

I posted the actual video it produced on my Instagram. Watch it here.

Most people would say I'm overreacting to a demo. Maybe. But I run a marketing company. I know what this work costs when humans do it. That one conversation replaced my entire creative team and a CMO that would have easily cost over $30,000 a month.

What an amazing time to be living in.

If an AI can read your business better than you can, what exactly is your job now?

My answer: taste and judgment. The AI found the five fixes. I still decided which ones were worth doing. That's the job that's left, and it's the part I teach every week inside Two Hour CEO

We have a nanny right now, plus two helpers. My sleep hasn't suffered. Our routines held. Life has continued pretty much as usual.

I know how that sounds. The guy who writes about breaking generational patterns is admitting his newborn season is smooth because he hired help. But that's exactly the conversation I want to have.

The nanny leaves in two months. So Zoey and I sat down and asked the real question: can we survive with three kids and two helpers?

Yes. We can.

But survive was the wrong word, and we both caught it. The actual question was: do we want to survive, or do we want to be present? Because there's a version of parenting where you white-knuckle through every day, and there's a version where you have enough margin to actually see your kids.

We're blessed to live in Singapore, where a helper costs around $700 to $800 a month. And here's the reframe that settled it for us: that money doesn't buy labor. It buys back our mental health, our marriage, and our ability to be with our children instead of just around them.

Help is not an expense. It's what presence costs.

What are you white-knuckling through in survival mode right now that a few hundred dollars of help would turn back into presence?

Nothing major changed in the portfolio. But I'm entering a new position, and I want to show you the breadcrumbs behind it.

The Setup: Cybercab production started at Giga Texas in April. Musk confirmed it on the Q1 call, and Tesla says there's no cap on the ramp. Meanwhile the robotaxi network is live: Austin went fully unsupervised, and Dallas and Houston launched in April with a fleet of 573 vehicles, mostly Model Ys.

The Signal: This week Tesla published the Cybercab Emergency Response Guide for first responders. Rescue sheets, how to disable the vehicle, all of it. You don't write the manual for firefighters unless the car is about to meet the fire truck.

The Speculation: Cybercabs are being produced faster than they're showing up in the fleet. My read, and this is speculation, not confirmation: Tesla is stockpiling until it has enough vehicles to flood the network all at once instead of trickling them out.

The Wilder Bet: A SpaceX and Tesla merger within the next year or two. SpaceX IPO'd in June, and the structure tells you everything: Musk keeps supermajority voting control through super-voting shares, reported north of 50 percent. At Tesla he holds roughly a fifth of the votes and has to answer to everyone. Elon thinks in decades, and public markets punish decades. The New York Times is already writing about a mega-merger, and Polymarket prices one before May 2027 at 46 percent. I've held Tesla for many years, and consolidating everything under a structure he controls is the most Elon move on the board.

The Position: The shares stay untouched. The LEAPS are a separate one-year news trade. Tesla trades on headlines, and right now the potential good news (mass robotaxi rollout, a merger announcement) outweighs the bad. Any one of those headlines rerates the stock, and the calls capture it with leverage.

The Track Record: Over the last two years I've taken 10 LEAPS positions like this one. 8 made money. That's an 80% hit rate, and past performance says nothing about the future. The edge isn't a crystal ball. I read a lot of news and study the patterns before they become headlines, my own version of what Chris Camillo calls social arbitrage.

The Honest Part: This is purely speculation, and options can expire worthless. This position is sized so that if I'm wrong, it stings and nothing more. And so you know I'm not selling you a highlight reel: about 80% of what I've put into crypto has lost money. Haha. Clearly not my game. If you want to invest, do your own due diligence. (None of this is financial advice. It's just what I'm doing with my own money.)

Back in #004 I told you why I stopped taking 30 pills. This is where that journey ended up. If you forced me to empty the cabinet and keep only what I can actually feel, these five stay:

1. Fish oil, EPA-dominant, about 2g of EPA a day. For mental clarity and inflammation. Check your label: most fish oil is underdosed on EPA, and the ratio is the whole point.

2. Creatine, 5g a day. Everyone knows it as a gym supplement. Almost nobody uses it as what it also is: a nootropic. Your brain runs on the same phosphocreatine system as your muscles, and the research shows the cognitive effect is strongest when you're sleep-deprived. I have a three-week-old. This is not theoretical.

3. Vitamin D3 + K2. These two travel together for a reason. D3 increases how much calcium you absorb. K2 activates the proteins that direct that calcium into your bones instead of your arteries. Taking D without K is raising the shipment without telling the warehouse where to put it.

4. Magnesium glycinate, 300mg at night. Deeper sleep, and it's also a cofactor your body needs to actually use the vitamin D.

5. Glycine, 3g at night. It lowers your core body temperature, which is one of the triggers for deep sleep. And here's the quiet synergy: magnesium glycinate is literally magnesium bound to glycine, so the two stack the same pathway from different doors.

The Verdict: Five things. Not thirty. Two for the brain, one for the skeleton, two for sleep.

As promised, my private notes. No fluff.

See you next week,

Gabriel Judah

P.S. Inside Two Hour CEO, this week's live call was me teaching the community exactly how to prompt Fable 5, and I've created a loop where my agents cleaned, transcribed, and published the replay all on their own the very same night without me lifting a finger. Over 200 members are building "Loops" like this that automates the work you hate. You don't need to be a developer. Grab a seat here.