Posts

As a father of teenagers, I’ve been thinking hard about what skills they’ll actually need in a world where AI is everywhere. Not the hype, not the fear-mongering—just practical skills that will serve them well whether they’re applying to college, starting their first job, or building something of their own. This guide is for teens, but also for parents who want to understand what’s worth focusing on. It’s based on research, but written in plain language with real exercises they can actually do. ...

TLDR: Regulation is inevitable—if democratic institutions don’t write the rules, corporations will through their “permission systems” Current complex regulations favor large companies while crushing small competitors, creating digital feudalism where platforms extract rent from participation The solution isn’t eliminating rules but creating thin, open, transparent protocols like the early internet that enable innovation without permission How We Traded Democratic Rules for Corporate Permissions Regulation is the landscape, not the weather. Whether you’re building an app, writing a newsletter, selling crafts online, or just trying to share your work with the world, you will encounter rules that shape what’s possible. If democratic institutions don’t write them, the market’s largest players will. The question isn’t whether we’ll have rules—it’s who writes them, and whether they serve the many or the few. ...

TLDR: LLMs are like Deep Thought from Hitchhiker’s Guide—they provide answers without context, making them less useful for specific tasks AI agents solve this by providing specialized intelligence with domain knowledge, tools, and appropriate context boundaries The future isn’t one omniscient AI but a constellation of specialized agents working together, each excellent at their specific domain In Douglas Adams’ “The Hitchhiker’s Guide to the Galaxy,” the supercomputer Deep Thought spent 7.5 million years calculating the answer to the Ultimate Question of Life, the Universe, and Everything. The answer? 42. The problem? Nobody knew what the actual question was. This comedic masterpiece offers a profound lesson for our current AI revolution: answers without context are meaningless, and context without specialization is overwhelming. ...

TLDR: We’ve industrialized knowledge work into teachable frameworks, making it vulnerable to AI automation—when you can standardize expertise, you’ve done most of the work needed to automate it Unlike previous revolutions that automated manual labor, AI is automating knowledge work itself, collapsing entire career progression pathways The remaining human roles will require managing ambiguity, relationships, and creative problem-solving—skills our education system doesn’t prioritize Success means recognizing this shift early and developing uniquely human capabilities rather than competing with AI on systematized tasks Every major technological shift creates two types of people: those who cling to the old ways and get swept aside, and those who recognize the change early and position themselves to ride the wave. We’re at one of those inflection points now. ...

TLDR: Labels like “left/right” or “elite/worker” don’t just describe reality—they create and reinforce the divisions we claim to oppose Inverting our language (“cancer got this person” vs “person has cancer”) externalizes conditions rather than making them defining features AI could amplify collective wisdom without the distorting lens of political categorization, helping us see each other as humans first Small linguistic shifts can create cascading changes in how we perceive reality and relate to each other In a world increasingly defined by division, we find ourselves trapped in a paradoxical cycle: the categories we create to understand our society reinforce the divisions we claim to oppose. Left or right, Democrat or Republican, worker or elite, pro-choice or pro-life—these labels don’t just describe reality; they actively shape it. ...