An assortment of opinions

September 1st, 2024

After reading How China Works, I’m more convinced than ever that centralization and decentralization of policy is way more important than liberal and conservative.

Seeing Like a State, and it’s accompanying critic really dive into this tension. Centralization can bring tremendous gains in production and distribution, yet the limited legibility of reality makes it impossible to design a perfect system. A good system carefully chooses what to centralize and leaves everything else decentralized.

It’s not just governments either. It’s any organization. Facebook’s famous metric-driven culture produced immense results during its growth period. You recognize that inevitably, the humans that “work for you” have their own priorities.

Cars

I hate cars, but America is built sparse. We’re a mishmash of races and cultures and giving everyone space is good. But cars are effectively human’s number one predator, after disease. They’re inhuman hunks of metal traveling at murderous speeds. An orderly city should only have professionally driven cars (taxis, delivery, buses). I have given up on effective public transit in America. Cities should just raise the cost of parking and subsidize ride-share (especially autonomous) to the point that a car is a luxury, not a necessity. The positive externalities seem entirely worth it.

This is a case where China’s party-led model works. The existing populace who own cars will never support this, but when it’s clear that it’s the right thing for society, sometimes you need a force external to society to un-entrench society.

Drones and robots

Drones are for surveying (not surveilling) because they don’t perch. They’re more like Google indexing the web than Datadog monitoring your stack. Drones are ideally deployed in cases where surveillance is uneconomical and unnecessary: “crawling” major highways and roadways every few hour for accidents, obstructions. They can also be used to deliver payloads to remote locations, but that’s only useful to a limited number of people (infrastructure and military). Most people live in cities and suburbs. Why not delivery? Any pilot will tell you the hardest part of flight is landing. Drones face the same problem. Uncertain conditions (obstructions, wind, animals) make it risky.

Why robots? Robots are “backwards compatible” with humans. So much of industry is already automated (factories, shipyards, even fast food) that what’s left to automate is residential/commercial. However, this world prioritize humans above efficiency. Robots are therefore useful for navigating these human spaces, not only from a practical point of view (climbing stairs), but from an emotional one too (nightmare spider robots are not welcome). And again, the efficiency of robots in this context is hardly important. Wood slat walls aren’t efficient or useful. They’re just nice looking.

Of course, robots don’t need to look like humans to do that. The dog format seems better than the human one and more emotionally aligned with their servile role. Help load the luggage. Bring in the groceries. Water the garden or indoor plants. Tidy up the house.

Robots will largely be indoor pets. The outdoors are much more dangerous. Robots are bound to be expensive, at least the cost of a car. And we don’t want 300lbs robots in our homes, but a 100lbs robot can too easily be stolen or vandalized.

Looking Glass

Why can’t you use two high-res camera and a short-throw projector to have a life-sized video conference with someone? Maybe through in a laser scanner? I want a portal wall, where I can call someone and it feels like they’re right there, through the looking glass.

I’d definitely want this for a remote personal trainer. But it’d be nice for therapy as well. Anytime you need a personal connection.