I just read two books on celebrities whose lives were ruined by addiction: Friends, Lovers, and the Big Terrible Thing, Matthew Perry. The Book of Sheen, Charlie Sheen. I’m so glad this disease does not afflict my wife or myself, and I really hope my sons are spared. It also reminds me of a Charlie Munger quote along the lines of “the hard thing about being rich/famous/powerful is not losing your mind” ...

I fatten quickly when eating carbohydrates. Not just sweets and ice cream, but bread and pasta. You might even say I’m carbohydrate intolerant. I think it’s because I was plump as a young kid and I have too many fat cells, especially around the mid section. They’re just sitting there all day waiting to grow. Begging. This is a reminder to myself that my eating operating system is to eat meat and vegetables. ...

Yesterday, I was listening to the Modern Wisdom podcast interview with Eliezer Yudkowsky about his new book (which I’ve read). The podcast: #1011 – Eliezer Yudkowsky – Why Superhuman AI Would Kill Us All The book: If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All, Eliezer Yudkowsky & Nate Soares, 2025. Yudkowsky always gives great analogies, this interview was no different. The climbing the ladder analogy hit for me: ...

I was travelling for some time recently. Many weeks. During that time, I pulled back on reading books and even on consuming long-form podcasts. Instead, I consumed vast amounts of social, specifically Instagram and TikTok. They are amazing at narrowing in on my interest of the moment. But I found that I felt off. Wrong. The best way I can describe it is my brain was fuzzy. I could not think clearly. I didn’t care so much about “stuff”. I was lethargic. I didn’t want to eat right or go to the gym. I just stopped caring. ...

I read “I’m Too Young To Die” by Bitmap Books last week. Here’s the official homepage with lots of cool photos of the book: I’m Too Young To Die: The Ultimate Guide to First-Person Shooters 1992–2002 Here’s the blurb: I’m Too Young To Die: The Ultimate Guide to First-Person Shooters 1992–2002. Covering the early, experimental years of the first-person shooter, we celebrate more than 180 games as we track the genre’s explosive entrance onto the ’90s gaming scene. Whether obscure oddities or genre-defining behemoths, first-person shooters transported players to alien worlds, alternate universes and the shores of Hell itself, and in doing so helped to trigger the 3D-graphics arms race. ...

I read “Wisdom Takes Work” this week. Good book. too much on Elon Musk. There was a good reminder to work on things that make your heart flutter. Great advice. I need to remember this.

Back from long holidays and I’ve put on 5kg/10lbs like it was nothing. Calorie deficit time. To get me in the mood, I sat down with Claude we wrote ~400 tips for maintaining a calorie deficit. I then hooked this up to my auto-tweeter and created a new twitter account: Calorie Deficit Tips It tweets new tips every 2 hours around the clock. I make sure to read the last few tips every time I’m on twitter to stay on the path.

I just just looking over my “Machine Learning Mischief” project from December last year: Machine Learning Mischief Good stuff. Very fun to think about, write about, and develop examples about. I was idly thinking of turning this into a book. No audience though… oh well. Maybe expand a subtopic and do that? e.g. seed hacking/p-hacking. Or maybe get out there and market/promote the ideas.

I run 5km 2-3 times per week as part of my cardio day at the gym. I have a routine of setting the pace at 6 minute kilometres initially and increasing every km. For example, the pace in minutes looks as follows over the whole 5km: 6, 5.5, 5, 4.5, 4. This takes 25 minutes. If I’m tired, I’ll run the first 2km at a 6min pace, then increase after that, for example: 6, 6, 5.5, 5, 4.5. This takes 27 minutes. ...

Every movement has a counter or opposite movement. The opposite of vibe coding? Let’s call it: “Discipline Coding”. Tenets Self-Reliance over Delegation Belief: If you didn’t write it, you don’t truly understand it. Practice: Hand-roll core algorithms, avoid generated code, and limit dependencies. Deliberate over Emergent Belief: Good code isn’t discovered in a “vibe,” it’s designed with intention. Practice: Careful planning, diagrams, upfront modeling before writing a line of code. Transparency over Black Box ...