AI, technology, education, and ideas about what comes next.
AI is becoming extraordinarily good at doing things. Yet it can still misunderstand what a person actually means. The deeper problem may not be intelligence, but judgment: knowing what matters in the first place.
IQ matters, but it is not destiny. Research suggests that success emerges from the interaction of cognitive ability, execution, education, environment, opportunity, persistence, timing, luck, and compounding.
If AI can teach you almost anything, why go to college?
AI’s breadth makes it powerful at exploration, but human expertise can be better at narrow execution. The future may be AI that knows when to explore—and when to stay in the box.
AI is making intelligence, coordination and information cheaper. Could that eventually change how money, local economies and even government work?
I tried to use Microsoft OneNote as a long-term archive for my ChatGPT work. The experiment exposed a deeper problem: storing AI-assisted work is easy; designing a useful system for organizing, retrieving and working with it is much harder.
A humorous but serious look at the gap between AI model capability and the amount of useful work a user can actually extract from it: fuel, interface bandwidth, context mobility, execution access and reliability.
A real-user review of ChatGPT’s model intelligence, memory, retrieval, UI/UX and tool limitations — and why a Pro upgrade is not the obvious answer when information flow is the real bottleneck.
Many important ideas begin as tiny seeds rather than fully formed projects. Multi-Seed Theory proposes planting many small ideas, testing them cheaply, pruning weak branches and investing more deeply in the few that show real potential.
I work in the second half of the day, when I take classes. On other days, I trade on the share market in day time or make videos…