4 min readMay 23, 2017


Among my drafts, I have an article concept about what I might call the “Manhattan Fallacy” and your response here is the exact line-of-thinking I was wanting to provide a counter-balance to.

I realize that you may not be able to read this article due to changes that Medium has made (kind of a shame IMO), but this article had a fantastic takedown of the budgetary implosion of efforts for light rail in the United States. My city the same as most US cities of similar size, in that a light rail proposal has come and gone, and has a price of ~$2 billion for a token single line amongst a massive sprawl in all directions. We’re already paying sales tax for it, although it will almost certainly never be built, but I’m digressing, there’s so much to cover…

For a moment I’ll go into personal stories because you asked. A few months ago I visited London for the first time in my life, of course I took the tubes, and it was a wonderful trip. I’ve been among the people pushed into cars by train workers in Tokyo. I also have personal friends who can attest to the fact that NYC is crippling expensive, and many people living there are doing so unsustainably, and so they eventually move on and move out.

Myself, I have lived and worked in around a dozen locations in the US, and I can go through them one-by-one if you’d like. In almost all cases, I started with a work location and found housing afterwards, most commonly using the open markets for rentals. Several times, I was fortunate to live a walkable distance (although one of those times I still drove because crossing one particular intersection would have seriously made me fear for my life). I do check these things, and public transportation was a viable option for 1 of those locations I lived (a bus, of course). Currently, I could do a mix of walking and public transit to work, but it would take >3.5 hours one-way. This is not uncommon. In many places I lived, there simply were not any public transportation options. I kid you not, I would not have been allowed to walk to work (again, no public transit) in 2 of the places I’ve worked.

Carpooling is very workable in the US, and we should absolutely be doing more to facilitate more of this.

The best of us have all fallen into the fallacy of advocating something for the “greater good” that was not realistic due to the cold hard reality of economics. We should first start out accepting that cost is a real thing. If a light rail line costs a 10% of the city’s yearly economic product to build, we’re not going to build it. So the issue of public transportation comes down to 1) reasonable predictions of cost and 2) consideration of the merits and drawbacks of alternatives.

On the other hand, I believe in the value of discussing the fundamental physical reality at work. There is a first-principle at work. The novel technology that will soon hit mass-market is robot chauffeurs. Cars and trains, themselves, are old tech. The only primary effect is reduce the cost of driving services. An overlooked secondary effect is allowing those chauffeurs to coordinates amongst each other. I think we are both in agreement (considering your articles like this) that this secondary effect has massively positive implications for mass-transit.

Allowing accurate coordination between robotic taxis will allow drop-offs and pick-ups to be timed down to the second, which allows transferring people from small vehicles operating locally to large vehicles operating over a larger distance. This is such a deployable, scalable, and powerful system that municipal mass-transit systems will either adopt this model or be destroyed.

But rail has no place in this. Change happens too fast for bulky outdated technology like this. The efficiency argument for light rail over buses is insufficient to make up for their inefficiency due to logistical inflexibility. Public transit systems operate with extra capacity the majority of the time because they need to cope with demand spikes when they come. American public-transit systems are very often less efficient than personal vehicles because of this form of waste (keep in mind that per-person-mile metrics are distorted because the bus system travels more distance to get someone to the same destination). Robotic dispatch algorithms can allow the number of buses in service to scale to meet demand. They will be able to string together busses in a train-like thing with each unit earmarked for a different service model. Again, light rail has no chance of adapting fast enough to use this model.

I’m totally sympathetic to the human-side considerations that you bring up. We could find ourselves building a psychologically torturous future, and I’m not going to argue with you on that point. It just doesn’t matter compared to the bigger picture. The future just isn’t up to you or me to decide, it is determined by powerful forces outside of our control. If we want to make any difference, action should start from understanding those trends and figuring out actions have a sufficient economic multiplier to affect the direction technology moves in.




Obligatory analytical writing, online participation account for Medium. Engineering, software, books, space, constant daydreaming.