3 Thoughts on Cars

Autonomous rides. Car dealerships. Uber losses.

Ranjan here, and this week I'm going to talk about cars.

Let's start with the fact that I know almost nothing about cars. The only car I ever bought was during college in 2001. It was a Maroon 1992 Corolla and I think it cost $1,200. One of my favorite parts living in NYC is not having to own or think about cars, save for those obscenely expensive occasional rentals.

However, a few recent developments have me thinking about cars a lot. This, being The Margins, made me want to overanalyze them in the contexts of technology and business. I also realized that each development represents a very different outlook on the impact of technology on society.


I rode in an autonomous vehicle! I've been working at a client's office at the Brooklyn Navy Yard, though it's a bit misleading to call it an office, as it's a really cool warehouse-y, industrial, tech-y space called the RLab. The Navy Yard just happens to be the first place in NYC where a startup called Optimus Ride just launched six autonomous shuttles.

Riding in the autonomous shuttle was incredible.

photo via OptimusRide

I know I might come off as a bit cynical when it comes to technology, but let me explain. For all of my life, I unquestioningly LOVED all things tech and considered every new product or feature as unadulterated progress. I’ve now, perhaps, swung too far on the cynical pendulum because I sometimes feel like a jilted lover, who loved and lost, rather than never loved at all. But riding in this goofy little shuttle felt like young love. Or maybe the first iPhone. Clunky and awkward, fuzzy and warm, innocent and full of possibility.

I have never experienced the self-driving capabilities of a Tesla. Friends in San Francisco, who all work at big tech firms, all, naturally, have Teslas. They love to talk about how far along the autonomous driving features are, but I've never seen it, so watching a steering wheel turn with no human hands was insane to watch.

More importantly, Optimus Ride did this right. They understood the current moment and adapted accordingly. There are two people up front the entire time. One has a big laptop open that shows you a constant stream of everything the car's LIDAR is "seeing". The other is behind the wheel as a safety precaution. They talk to you about safety and walk you through what's happening, like tour guides. Some people used the multiple “drivers” as an occasion to snark that we’re still far away from autonomous, but that is the point. We’re still on the way. It’s worth doing this slowly, and not promising fully driverless cars as soon as this year.

The company also chose the perfect place to try this; the Navy Yard is huge space, but it's all enclosed with little thru or pedestrian traffic. It's a very controlled environment, with just the right amount of bikes crossing your path or cars passing by where the shuttle is lightly challenged.

And it was just plain, nerdy fun. The two people up front were genuinely giddy to talk about the cars, and the "cars" looked like gutted, futuristic PT Cruisers. For context, the walk from the ferry stop to the main entrance of the Brooklyn Navy Yard, the route these vehicles travel, is 0.8 miles. yet, most people were just walking by the cars ignoring this was a transport option. One person yelled "just get on a bus!"

This all made it feel like good old-fashioned early adopter fun! Cool technology, trying to solve a defined problem. People ridiculing and/or ignoring it. The founder is an MIT PhD and longtime MIT Media Lab researcher, and the entire experience seemed to reflect this.


I'll admit it. I've been looking at buying a car. I feel ashamed to write this, as I strongly believe in the importance of both public transit (shared spaces!) and making cities better for bicycles. But now, with two kids, I've been looking at something reasonable to get out of the city more, and have been shopping for a Honda Fit. For those unfamiliar, they kind of feel like the modern, urban cousin of a 1990s minivan.

(Side note: The ultimate endpoint to a formerly free-wheeling bachelor life is when you google “honda fit trunk space uppababy vista double stroller”.)

I honestly thought tech had "solved" car-buying. As I avoided the process for nearly two decades, I assumed there were heavily disruptive forces that would allow me to simply define my requirements, be shown a few prices, tap a button on my phone and a shiny, new car would show up with all insurance and ownership requirements fulfilled.

It's not that, at all. There is certainly more information with which to arm oneself. But, almost every sleek aggregator site ends up leading you to a local franchise where your interactions are, well, what I always imagined car-buying to be. Given I work a lot with email, I have very strong thoughts on emails like this:

The most surprising part of this was how aggressively every dealership pushed for getting your email and phone number. To simply view prices, most of them force you to enter both, but it makes sense. The cost of the information capture is far lower than the high-value sale. You do whatever you can to get that phone number, and then spam the buyer incessantly.

The process made me think about my co-host Can's writing on Data-as-a-Liability. Imagine how this equation will change if (hopefully when) the dealer assumes some cost in accumulating and storing tens of thousands of email addresses and phone numbers. How will that change site UX and overall buying process?

This entire experience has been a reminder that in many industries, for all the talk of technological disruption, many things have not changed.

(Note: Carvana was the closest thing to delivering the digital-first experience I wanted, but only deals in used cars. Given I'm looking for a fairly inexpensive car, I was looking for something new.)


Last winter, I was debating the impact of Uber and Lyft on public transit infrastructure with a self-declared neoliberal friend. The New York City subway has certainly been seeing its increasing share of issues (this 2017 NYT piece is the best explanation I’ve seen on why), and my friend’s argument was that you can only see "profitable" innovation, or operational and cost structures that aren't wasteful, from companies in the private sector.

After the conversation, I sent her Michael Lewis's The Fifth Risk.

Yesterday, we learned in Uber’s earnings that the company again lost over $1 billion in a quarter (the stock-based compensation charge, something Can has written about, is what pushed it to a ludicrous sounding $5 billion). They’ve lost over $16 billion in their lifetime. And to clarify, that ~$16 billion is not what they’ve spent. It’s what they’ve lost!

Is Uber really a Koch Brothers-ian wet dream, or is it just a public infrastructure cost structure with a private market compensation structure? The common Heritage Foundation / right-wing perspective is that public infrastructure is wasteful and a vibrant private sector is where you find innovative and profitable organizations. But is Uber destined to eternally lose money on a per-unit basis?

Their go-to market strategy has certainly worked. Thanks to cheap capital, they were able to build consumer habit off of artificially deflated prices. If my Uberpool costs less than a subway ride, there is no question I’ll take it. But when they raise prices to break-even levels, do we all go back to taking the subway (this is admittedly a very NYC-centric view of the issue)?

The only paths to profitability for Uber have always felt like:

  • Uber eliminates the cost of drivers with self-driving cars

  • Uber builds a strong enough monopoly where they can simply raise prices and users have no choice but to stay with them.

Neither of these sound like a case study in private market efficiencies, but instead, forms of financial arbitrage (#EverythingIsArbitrage). Even more worrisome, on this road to a moonshot break-even, could they have already obliterated public transit systems?

This felt especially relevant after reading a NY Times piece from this week which informed me Uber will begin selling public transit tickets within its app in some cities. There are already a number of cities where the local government heavily subsidizes Uber rides in low-transit option areas. This seems laudable - but isn’t a per-unit, money-losing service that is heavily subsidized by government, in this case both local and foreign (via Saudi PIF/Vision Fund) just…..public transit.

The question of whether Uber can ever be profitable is important if they are to become an increasingly integrated part of public infrastructure. We need to better frame debates around public vs. private market innovation to account for the possibility that some private companies were never meant to be private.


These are admittedly three disparate ramblings, and this has probably been the most I’ve thought about cars in my lifetime. My love/fear relationship with technology remains something I continue to feel the need to work through via writing newsletters. In the meantime, I’m going to go test drive a positively gangster Lunar Silver Honda Fit.

A Video

This is one of the darkest and funniest short clips I’ve ever seen. If Black Mirror and Saturday Night Live had a baby, this would be it.