Tag: analysis

Reality & Straight lines

Does reality drive straight lines on graphs, or do straight lines on graphs drive reality?

Here’s a graph of US air pollution over time:

Some people pointed out that this showed the Clean Air Act didn’t matter. The trend is the same before the Act as after it. This kind of argument is common. If we wanted to be really harsh, we could make a graph like this:

But the same argument that disproves the importance of photolithography disproves the importance of anything else. We’d have to retreat to a 1000-coin-flips model where each factor is so small that it happening or not happening at any given time doesn’t change the graph in a visible way.

Facebook Privacy Pivot

Ultimately, Zuckerberg doesn’t address the biggest trade-off: Are these changes compatible with Facebook’s fundamental business model, which relies on a steady supply of user data? If these changes are truly implemented, there will be a substantial business cost to bear. Until he fully answers that, Zuckerberg’s vision of privacy will be incomplete.

The move towards privacy seems designed to respond to a number of problems. To some extent, it responds to Facebook’s content-moderation problem: if a conversation that violates the company’s “community guidelines” is encrypted, such that even Facebook’s software can’t read it, then the company can’t be expected to expunge it. It responds to the Ralph Northam-yearbook problem (“People want to know that what they share won’t come back to hurt them later”), since, by default, content will auto-delete. And it responds to the China problem, as Zuckerberg vows “not to build data centers in countries that have a track record of violating human rights like privacy or freedom of expression.” It also attempts to address Facebook’s public-relations problem: after a scandalous year, fewer people trust the network than ever.

Data Incumbency

Open-data requirements could make it clearer who is providing a valuable service and who is primarily exploiting information asymmetries between creators and services. They could help identify the genuine rip-offs that thrive on opacity, such as the chronic underpayment of artists or the role of concert organizers in ticket black markets. They could help answer the perennial question of whether streaming payments and other licensing schemes are fair to artists, based not on a notional “value gap” but on who else is getting a cut.

Airships

Giant Airships Could Be a Trillion $ Industry

Mode Cost per ton-km Typical speed (kph)
Airplane >$1 >800
Truck 15-20¢ 100
Giant Cargo Airship (projected) 5-10¢ 160
Rail 3-5¢ 100 (but with transshipment delays)
Ship/Barge ~1¢ 30

Now for the disclaimers. “Per ton-km” estimates of shipping costs always misleadingly oversimplify. Costs tend to be concentrated at transshipment points. It matters a lot whether there’s a payload for the return trip. Modes differ in reliability, and if occasional transport delays make expensive bottlenecks, it might not matter if a mode can offer cheap service and adequate speed most of the time. Handling is crucial for some cargoes, e.g., fragile or perishable ones. Sometimes space is a more binding constraint than weight, and lightweight, bulky cargoes often pay “chargeable weight” proportional to the space they occupy rather than their actual weight. Above all, not all modes go to all destinations: e.g., trucks can go wherever there’s a road, but not over bodies of water, while airplanes can cross water but can only land where there’s a port. And so forth. Logistics is complicated! As for the estimates of cost and speed for a giant cargo airship, suffice it to say, for the time being, that I do have some basis for suggesting that 160 km per hour and 5-10¢ per ton-km are attainable performance goals for a giant cargo airship, though it will take a lot of investment to get there. I’m not asking readers to trust me on that, except for the sake of argument.

2021-10-22: Airships to overcome lack of roads

Airships right now are interesting as a cost-effective substitute for less efficient cargo transportation that’s already happening. But they can be much more: one of the barriers to manufacturing exports in the developing world is that exports require infrastructure, and infrastructure generally requires spending, which needs to be funded, ideally from exports. If the fixed cost of participating in global supply chains drops, the beneficiaries will be people in the poorest parts of the world, who will have a shot at joining the global market. Transportation costs are high in the developing world, which amounts to a tax on imports and exports, benefiting no one. If countries can skip some of the expensive and time-consuming process of building road and railroad networks, and focus on ports, they can jump over a significant barrier to higher incomes.

2023-02-03: Eli Dourado does another market sizing.

Let’s say airships captured half of the 13 trillion ton-km currently served by container ships at a price of 10¢ per ton-km. That would equal $650b in annual revenue for cargo airships, notably much bigger than the $106b Boeing reports for the entire global air freight market. If one company owned the cargo airship market, taking only 50% of only the container market, it would be the biggest company in the world by revenue.

How many airships would we need to fill that demand? A lot. If each airship can carry 500 tons, cruises at 90 km/h, and is utilized 66% of the time, that adds up to 260 million ton-km per year per airship. To produce 6.5 trillion ton-km per year would require 25k such airships. This is about the number of airliners in the world today.

None of this analysis yet assumes any expansion of the market from normal growth, from the availability of a new service class, or from the ability to go point-to-point rather than shipping between existing ports. But it’s easy to imagine new trading patterns and even new companies forming because cargo airships exist. Just as Uber and Lyft massively expanded the vehicle-for-hire market, the added supply chain flexibility afforded by airships would stimulate new demand.

Status as a Service

The gradient of your network’s social capital ROI can often govern your market share among different demographics. Young girls flocked to Musical.ly in its early days because they were uniquely good at the lip synch dance routine videos that were its bread and butter. In this age of never ending notifications, heavy social media users are hyper aware of differing status ROI among the apps they use.

On the dumb things people do to maintain “status”

Changing restaurant culture

The community of people I surrounded myself with ate and drank like Vikings. It worked well in my 20s. It worked well in my 30s. It started to unravel when I was 40. I couldn’t shut it off. All of a sudden, there was no bottle of wine good enough for me. I’m drinking, like, literally the finest wines of the world. Foie gras is not exciting. Truffles are meh. I don’t want lobster; I had it yesterday. What am I looking for, eating and drinking like this every day?

Fortnite Concert

People have gathered in virtual worlds for decades. People have attended virtual concerts for years. Yet the Fortnite event represented something different by many orders of magnitude. By one (unsubstantiated) estimate, 10M concurrent users attended the show in the game’s “Showtime” mode. In other words, this was something much more than a concert. It was a peek, albeit a short one, at what an AR- and VR-suffused future looks like: connected congregations of embodied avatars, in mass-scale events that still manage to feel personal.

Micromobility

NYC DOT data mines Uber and Lyft Trips

New York will start clawing in the same kind of data from the ride-hailing companies that have stormed its streets in recent years. If Uber, Lyft, Via, and Juno want to keep operating in the city, they’ll have to provide the TLC with even more finely detailed data than they do now: the date, time, and location of pickups and drop-offs (at least down to the intersection), the vehicle’s license number, the trip mileage, itemized trip fare, route (including whether the vehicle entered traffic-choked Midtown), and how much the driver was paid. The city intends to use all this data to learn more about what’s happening on the streets, and to plan. It will ponder how to beat traffic and improve road safety.

2019-02-05: Massive growth in bikeshare

The rate of the industry’s growth is no joke. The global explosion of shared bikes and scooters in the past few years amounts to the “the fastest technological adoption in history,” as the event’s website noted. And it’s just getting started. “Logarithmic is the way to go with everything” explained Micromobility Conference founder Horace Dediu. He pointed to charts showing the expected rise in adoption of shared electric scooters and their 1-to-3-wheeled brethren. “If you’re not measuring in logarithmic, you’re in the wrong business.” In the meantime, the mini-vehicle gold rush will likely keep beckoning new prospectors. 2 former employees of the traditional docked bikeshare operator Motivate, now both working at 2 new-mobility startups, reminisced about old times maintaining New York City’s hardy CitiBike fleet. They complained that all the sudden interest from software startups was leaving critical hardware overlooked. A former manager pointed to the apparently fragile wheel hubs on a competitor’s shared bike model. “We tried those. They’re just not going to last.”

2019-05-08: The segway was too early but was otherwise correct

This transformation is one that Dean Kamen trumpeted when introducing the Segway in 2001 — the Segway will “sweep over the world and change lives, cities, and ways of thinking” — but the Segway was too early and expensive and now e-scooters and e-bikes are actually set to deliver on that promise

2019-08-12: Tiny cars

Minimobility vehicles will use more energy and road space than scooters, but still much less than full sized cars and transit systems. Expect to see energy numbers perhaps 33%-50%, though it is possible to do even less. Minimobility vehicles will take 50-25% the space on roads of human driven sedans. Many are only 2.5m long and thus fit perpendicular in parallel parking spaces as motorcycles do. That’s cool, but even better is the ability of computer-driven cars to valet park, fitting 5 in the space used by 1 human driven sedan. (You can also fit 10s of small micromobility scooters in a sedan’s parking space.)

2019-08-24: Similarly in LA

Los Angeles will deploy a massive surveillance dragnet targeting the less than menacing threat posed by…bikes and scooters. That’s right, a city-wide, real-time tracking network, a veritable Orwellian surveillance state, targeting the same sort of scooters popular with middle schoolers. In a perplexing technical document posted by the Los Angeles Department of Transportation to the open-source software development site GitHub, the city laid-out its vision for an unprecedented tracking system. Buried between technical jargon like “All MDS compatible provider APIs must expose a public GBFS feed…” is a terrifying vision of the future. Under these guidelines, every scooter and bike will need to report the exact GPS coordinates of each device with military precision every 5 seconds. Even worse, the devices will have to report this data in real time.

2022-12-14: More on tiny cars

Why do minicars offer so much societal upside? Let’s walk through the reasons.
To start, they pose far less danger than automobiles to pedestrians and cyclists, whose deaths have been soaring in the US. It’s easy to understand why: If forced to choose, you’d rather be hit by a 500kg golf cart puttering along at 30 kmh than a 3500kg Chevrolet Silverado going 70kmh.
Minicars are better for the planet, too, devouring much less energy and resources for propulsion and manufacture. Their batteries are a fraction of those underpinning electric SUVs and trucks. A Hummer EV’s battery weighs 1500kg, 30x the size of Nimbus’s battery pack. Smaller batteries require less lithium, cobalt and nickel — essential minerals whose shortages have hobbled electric vehicle manufacturing. Even a modest shift toward miniature EVs would free up precious material, catalyzing the transition from fossil fuels.
Minicars also expand mobility access, in two ways. Most obviously, their lower prices put them within reach of those who struggle to afford a full-sized car. But slower minicars, such as golf carts, offer another key advantage. Because most operators don’t need to pedal (or, in some states, possess a driver’s license), they offer mobility to those physically unable to drive a car or ride a bike. That’s an especially valuable capability in a country with soaring numbers of older people.

0 To 1

A short review can’t fully do justice to the book’s treatment of monopolies. Gwern’s look at commoditizing your complement almost does. But the basic economic argument goes like this: In a normal industry (eg restaurant ownership) competition should drive profit margins close to 0. Want to open an Indian restaurant in Mountain View? There will be another on the same street, and 2 more just down the way. If you automate every process that can be automated, mercilessly pursue efficiency, and work yourself and your employees to the bone – then you can just barely compete on price. You can earn enough money to live, and to not immediately give up in disgust and go into another line of business (after all, if you didn’t earn that much, your competitors would already have given up in disgust and gone into another line of business, and your task would be easier). But the average Indian restaurant is in an economic state of nature, and its life will be nasty, brutish, and short.

The BuzzFeed Lesson

The core problem for BuzzFeed is that never really happened: Instant Articles relied on the Facebook Audience Network, not Facebook’s core News Feed ad product, and nearly all of Facebook’s energy went into the latter. Companies that embraced Instant Articles — and, in the case of BuzzFeed, built their business models around them — were left earning pennies, mostly on programmatic advertising.