Hacker News Reader

Comment by niftich | Hacker News Reader

> Movable houses (...)

Sounds intriguing, if skeptical about cost, material strength, coordination, labor costs, longevity, etc.

> All-robotic car fleets are an obvious first step.

This is a truly big ask.

The rest of the essay makes a point similar to homesteading, where you pick an arbitrary lot you like and move your movable house to it. It has a section about land value taxes that offers a total non-solution to the supply-and-demand problem of land.

There are some good, insightful observations, and some wishful thinking; the essay even ends with the admission that it's more of a "wistful sigh" and the author is "nearly certain that none of that will actually happen". But intellectually rigorous analysis or proposal this is not.

View on HN

> This is a truly big ask.

From the article: The technology for an all-robotic car fleet is probably already available... if we take all the non-robotic cars off the road.

Many of the challenges in autonomous vehicles (maybe even all the remaining ones) are related to modelling/predicting an uncertain world. Will a pedestrian jump out at me? Will a truck do a left-turn across my path? What in the world is that bicyclist doing?

Take all that out, and things get vastly easier.

Well yes, it's the same sort of argument that teachers jokingly make, saying school would be so nice without all the students.

It's been suspected for some time that a mixed environment of autonomous vehicles and human-operated everything else (including pedestrians) will be the worst of both worlds, as like you said, those algorithms will be face with un-anticipated inputs.

The problem is, while cars we can automate way (given enough time), pedestrians we can't, unless we're willing to expend even more money in re-engineering our infrastructure to segregate road traffic entirely, keeping the contact surface to as minimal as possible -- only every home, every workplace, every commercial establishment, every rest stop, and every place vehicular maintenance is done. While this goal tends to appear in 'next city' brainstorming, it's easy to see why it's a challenge.

Oh, I thought your "big ask" comment was about the technology.

Sure, you'd have to segregate away pedestrians, but there's already precedent for that with many roads (no pedestrians on the highways!), and aren't we dreaming about making new cities anyway, rather than retrofitting old ones?

"No pedestrians allowed on the lanes which are painted red, not that you'd want to cross that small spiked fence anyway. Just use the overhead walkway like everyone else."

>It has a section about land value taxes that offers a total non-solution to the supply-and-demand problem of land.

Can you expand on this? I'm not exactly sure what you consider the problem to be.

Reply on news.ycombinator.com