Everyone wants exciting. I want boring and repeatable.
Compounding isn't about one big win. It's about how many times you can repeat a small one and scale.
Here's the model:
→ Build a cohort of 20 stocks, make it your micro-fund
→ Hold for 3-4 months
→ Capture the gains
→ Rotate into the next cohort
→ Repeat, 4x a year
No moonshots. No hero bets. Just a data model that finds the same small edge, over and over.
Three different teams. Three different problems solved. Same ending: the model labs didn’t need the product, just the people.
AI is starting to feel like a rat race for pre-seed startups
and a risky bet for investors
Is AI becoming a dangerous asset for investors?
I saw a dozen of "great AI apps" going down for the last 12 months
some priced out by Claude Code,
others featured out by OpenAI’s next release.
Convogo started as a weekend hackathon solving a real problem for executive coaches. OpenAI took the founders. The product didn’t survive the deal.
Your bet on those great founders: GONE.
3/ Pre-trade is the actual game. Execution is what’s left over after the real decisions are already made.
4/ Retail obsessing over execution is like optimizing your free-throw form while the other team already knows the final score.
1/ Retail investors love to talk about “strategy.” Entry points. Exit points. Dollar-cost averaging. Technical setups.
None of it matters. You’re optimizing the wrong stage of the game.
I don't invest like everyone else.
I'm an engineer.
I like to see things work. I test things constantly.
I build things to make sure they are useful and that they don't fail.