🧭 Agents as App Aggregators (Stratechery)
Not unlike the iPhone, AI agents are transcendent products that aggregate multiple services in one screen. If you win the agent war, everyone is your supplier.
🐏 Everyone Hired Mr. Claude (NOEMA)
Companies are getting Airbrain — the common affliction where the AI ends up doing the work and critical thinking. The corporate brain atrophies, the tools create the same outputs as the competitors and differentiation ceases.
🥇 Manufacturing Eileen Gu (The Science of Parenting)
Almost every parent wants their kid to be as polished and accomplished as Eileen. Few want to trade childhood for eight-hour round trips to ski training or summers spent in intense schooling. She might feel the polish was worth it; many others might not.
🏢 Voluntary In-Person Collaboration (Work Forward)
Atlassian is a remote-first company, but its workers are spending more time together these days. Having a great collaboration space incents workers to gather and do their best work. In a world where execution is cheap, coordination is worth the in-person investment.
🐝 The Swarm Manages Itself (One Useful Thing)
One might assume that your agent swarm needs a manager. It turns out part of their superpower is to organize themselves.
🖐️ Winning Ugly, Not Dirty (Box of Amazing)
Winning dirty is not new — Maradona did it in 1986, Volkswagen tried to do it with emissions testing. Winning ugly is tougher. Principled innovation requires setting your principles when it is nil-nil. How will your agents align?
💪 Great Planner, Terrible Hands (Understanding Robots)
OpenAI’s Astra briefly topped the robot-control leaderboard. Then it met reality and could not quite figure out how to use its robot hands. Small models might have some advantages there.
🏟️ Live Beats Digital (Apricitas Economics)
Economic data illustrates how the creative class is hollowing out. 200k jobs in four years. Live events are booming as the value of experiences increases in a sea of entertainment.
🚕 Uber Made More Milkshakes (Bet On It)
In 2012, Americans rode some 4.5 billion miles in taxis. Today, Americans ride some 28 billion miles in ride-share and taxis. Imagine what self-driving cars might do.
🚲 2026 in LLMs So Far (Simon Willison)
Language models have jumped ahead. Agents are doing the easy things. Now to make them do the hard things.
🍽️ The Decline of the Dinner Party (Derek Thompson)
In mid-20th-century America everyone threw dinner parties. Now very few people do. Did the machines take our guests too?
🧺 Nobody Misses the Washboard (Risk & Progress)
Laundry, meals and cleaning once took 58 hours a week. Progress has reduced that to an average of 15 in modern times. Machines took our chores long before they took our code.
The Look
The More or Less Pod asks the interesting questions about AI — what if there is no moat around your foundation model?
Did you enjoy reading this post? Hit the ♥ button above or below because it helps more people discover great Substacks like this one and it helps train your algorithm to get you more posts you like. Please share here or in your networks to help us grow!