North America

Robots and NHL Expansion

Winnipeg and Las Vegas, the two newest NHL franchises since Minnesota and Columbus joined the league in 2000, have one thing in common: nobody lives near them. Apart from much larger, regional capitals, like New York City or Phoenix, both Winnipeg and Las Vegas account for a far bigger share of their state or province’s total population than do any of the other cities with NHL teams.


In other words, both Winnipeg and Las Vegas are located pretty much in the middle of nowhere.

In spite of this, Winnipeg and Vegas represent opposing strategies to adding new teams to the NHL. Returning a team to Winnipeg was an example of what we will call a short-distance strategy. It was (to state the obvious) intended to capitalize on hockey fans, a.k.a. Canadians, who live in Winnipeg. As Winnipeg had been the largest Canadian city without a team, and Manitoba the largest province without a team, bringing the Jets back was an obvious decision for the NHL to make.

The league does not, however, expect many people at Jets games to have come from afar. Even outside of Winnipeg’s metro area, most Manitobans live not far from the city. Winnipeg’s neighbours, moreover, are distant andd sparsely populated. Saskatchewan has just 1.1 million people; its largest city, Saskatoon, is 710 km away from Winnipeg. Calgary and Edmonton are 1200 km from Winnipeg. Fargo is 330 km to Winnipeg’s south, Minneapolis 615 km. And almost nobody lives in northwestern Ontario. Toronto and Ottawa are more than 1700 km away. Manitoba cannot rely much on its neighbours to buy hockey tickets.

Las Vegas is following the opposite strategy: a long-distance strategy. It hopes to attract fans (aka gamblers, tourists) from hundreds or thousands of kilometres away: from Canada, the rest of the United States, and even overseas.

Even the Vegas locals, who the NHL hopes to convert into hockey fans, are dependent on long-distance tourism. Without tourism, Las Vegas’ economy would dry up and force many of the locals to leave (or at least, to spend less money on hockey tickets). This the Las Vegas Golden Knights would not be able to afford. Once the Oakland Raiders move to Las Vegas in 2019, Nevada will have the smallest population per each of its major sports franchises of any state or province—with only one exception: Manitoba.


Pittsburgh and Nashville  

This year’s Stanley Cup contenders, Pittsburgh and Nashville, are very different than Winnipeg and Vegas. For one thing, neither are the largest cities in their states. The Greater Nashville metro area is home to only an estimated 27 percent of the population of Tennessee; Pittsburgh’s metro area is home to just 13 percent of Pennsylvania’s population. Pittsburgh was fifth from the bottom on both of the blue graphs above.

Nashville's Nearest Neighbours.png

The Nashville Predators, and its nearest fellow teams in every direction—plus the Thrashers, which left Atlanta to become the Jets in 2011. 

For Nashville, not only are the Predators the only team in Tennessee, they are also surrounded by five states with no NHL teams: Georgia, Mississippi, Arkansas, Alabama, and Kentucky. Tens of millions of people live within a few hundred km of Nashville, and none of them have their own teams.

Pennsylvania Teams.png

Pittsburgh too is in a region with a large population yet relatively few hockey teams—albeit not nearly to the same extent as Nashville. This region includes Ohio, which has no team north of Columbus, and the Virginias, which have no teams at all. Most importantly, it includes part of southwestern Ontario. Pittsburgh is located closer to the Canadian border than any other American team apart from Buffalo or Detroit.

Pennsylvania is also one of just two states that has exactly two NHL teams. (The other, Florida, relies on tourists and snowbirds, like Las Vegas will). This is a useful arrangement, creating an intrastate rivlary in which western Pennsylvania can cheer for the Penguins and the east for the Flyers.

Golden Knights or Goldilocks? 

Pittsburgh and Nashville are both examples of a medium-distance strategy for NHL expansion. Whereas Vegas will rely on fans jetting in from thousands of km away, and Winnipeg relies on Manitobans keeping the seats full, the Predators and Penguins can both — in theory, at least — attract fans or ticket-buyers who live within tens or hundreds of km of their arenas.

The question is however: which strategy is best?

The reason I bring this up is, as the title of this article indicated, robots. If Sillicon Valley is right, and technologies like autonomous cars really are coming just around the corner, might this make a medium-distance strategy wiser? Would it make the recent expansions to Winnipeg or Las Vegas ill-advised? After all, an autonomous vehicle could make driving tens or even hundreds of km to come home from a game—at night, in the winter, on a rural highway, after having drunk a beer or two earlier—safe and easy. This might increase dramatically the distance that fans are prepared to travel to go to a game.

A medium-distance strategy for future NHL expansion should, in general, prioritize cities that are in Canada or near the Canadian border. Such a team would allow Canadian hockey fans could come to games without having to travel too far a distance. Such cities might include Quebec City, Hamilton, Cleveland, Seattle, Milwaukee, or perhaps even Halifax, Saskatoon, London (in Ontario), Portland, or Toledo.

Most of these cities could not support a team without some new major advance in transportation technology, such as autonomous cars: the number of hockey fans who live in them is simply too small. Some may not be able to support a team even with robot cars. Halifax, for example, has a mere 400,000 inhabitants. It would need to draw in many fans from other Maritime cities to become viable.

While Quebec City and Hamilton are arguably the most sensible additions the NHL could make if following a medium-distance strategy, Seattle is I think the most intriguing one. Seattle is of course a sizeable city in its own right; it accounts for 50 percent (by metro area) and 9 percent (by municipality) of Washington’ population—middle-of-the-pack figures for cities that have NHL teams. Yet Washington as a state has only two major sports franchises (the Seattle Seahawaks and Seattle Mariners) for 7.2 million people. Along with neighbouring Oregon, Idaho, and British Columbia, the Pacific Northwest has only four teams (Seahawks, Mariners, Trailblaizers, and Canucks) for 17.5 million people. It used to have six, but the Supersonics and Grizzlies moved away.

quebec expansion

One reason the Pacific Northwest has so few sports teams per capita is its mountainous terrain. The mountains make land expensive, raising the cost of an arena. They also make driving tricky and limit the number of highways available, creating traffic. This makes it difficult for fans from other cities to drive to and from Seattle, Portland, or Vancouver to take in a game. For a Seattle NHL team this would be an especially important challenge, as the team would want hockey fans to visit from Canada. It is 236 km from Vancouver to Seattle, and 278 from Portland to Seattle. It is just 115 km from Victoria, BC’s capital city (with a population of 368,000), but only as the crow flies.

Pacific Northwest .png

Autonomous cars could, perhaps, help the Pacific Northwest overcome these challenges. They might do so by allowing an arena to be built further from Seattle’s expensive downtown core, or by allowing an arena to have much smaller parking lots (and therefore to occupy less expensive real estate) or by making it easier to drive hundreds of km through the region’s rugged and rainy terrain.

On the other hand…

Of course, it is easy for me to just say “autonomous cars” and then try to make up a cool-sounding argument around it. But that does not mean in any way that my argument is a good one.

In this case, it may instead be that a short-distance or long-distance approach, of the Winnipeg or Las Vegas variety, really will be better than a medium-distance one. This is something that the league should, I think, try to determine for itself.

If a short-distance strategy is determined to be best, then the obvious choice for expansion would be to put a second team in Toronto. Even with two teams, the municipality of Toronto would have approximately 1.4 inhabitants per NHL team and 700,000 inhabitants per “Big 4” sports franchise. In contrast, the municipalities of of Hamilton and Quebec City are home to only around 500,000 people each.

The Greater Toronto Area (not even including nearby Hamilton or Kitchener-Waterloo) would have 3.2 million people per NHL team and 1.6 million per Big 4 team were it to add a second NHL franchise. The Greater Montreal Area would have just 2 million people per team were it to do so.

A long-distance strategy, on the other hand, might focus on cities in the south, where hockey-loving snowbirds could flock. This could mean a first NHL team in one of the southern states without any, or a second team in Texas, or a third attempt at a team in Atlanta, or maybe even a fourth team in California.

Indeed, the most recent round of NHL expansion in southern cities was during the 1990s, when, perhaps not incidentally, the cost of travel was cheap and the Canadian dollar was weak, as oil prices were at an all-time low. San Jose, Anaheim, Miami, Tampa Bay, and Dallas all got teams during 1991-1993 (Ottawa also got a team in 1992), while Phoenix, Denver, Raleigh, Nashville, and Atlanta got teams during 1995-1999.

In contrast, the three since then have been northern: Columbus and Minneapolis in 2000, then Winnipeg in 2011. A long-distance approach, however, might be less friendly towards northern cities—particularly far-northern cities, such as Winnipeg, Quebec City, or Saskatoon. It might worry that too many Canadians will flee the cold and dark of winter to seek the bright sun of the south.

North America

The Witching Hour: How To Fix Traffic in 3 Easy Steps, Without Resorting To Autonomous Cars


1. Allow autonomous cars during “the Witching Hour”: from 4 am-5 am. They can drive slowly in order to be safe and quiet; say, at no more than 10 km per hour when in residential neighbourhoods. Even at these slow speeds, this will allow car-sharing  cars to be delivered to peoples’ homes for use the following morning. (In fact, the cars themselves do not even necessarily need to have an autonomous capability. They could instead just hitch a ride on top of slow-moving road roombas). In the case of electric cars, this will also allow them to drive themselves to and from battery-charing stations at night, when electricity tends to be cheap and road-traffic sparse.

2. On main streets, have both an express LRT lane — with stops very far apart from one another — and a non-express bus lane. On narrower streets, have the non-express busses share a lane with regular car traffic.

3. Next to many of the LRT stops, as well as next to train stations, construct “take a car, leave a car” vertical parking lots. These will be “valet” lots: you drive a car-sharing car to the lot’s entrance, then get out of the car and have it drive itself (or be carried by a road roomba) into the lot to park. This will not only save drivers time in parking, but will also allow the lot to hold far more cars than any traditional vertical parking lot could, since without humans it can have much shorter ceilings, more tightly winding ramps to get cars up or down floors, and many more parking spots per floor. It will allow easy pick-up or drop-off of car-sharing cars. Along with the Witching Hour, this will overcome the “first mile-last mile” problems that otherwise tend to limit public transit’s effectiveness and appeal.

…So, there you have it. Three easy steps! With the Witching Hour, and car-sharing, and vertical parking lots, we can finally help to get rid of our cities’ spooky traffic problems.

Middle East, South Asia

Robots and the Middle East 

It used to be, roughly speaking, that cheap labour + energy = industrial output = military power.

This made Iran the natural power in the Middle East. Iran had far more energy than countries like Turkey, Egypt, Israel or Pakistan, and far more labour than the Gulf Arab countries or Libya.

Robots in the Middle East, graph 2.png

X-axis is population size, Y-axis is the value of energy reserves as estimated by Business Insider in 2014


Robots in the Middle East, graph 1.png

X-axis is population size, Y-axis is energy production in 2014 as estimated by Shift Data Portal. Saudi Arabia leads Iran in energy production, but the Saudis have had the benefit of having an alliance rather than a rivalry  with the United States

The Gulf Arab monarchies have tried to overcome their relative deficit in labour by importing workers from Asia. However there are limits to such immigration, not only because of the fear that the immigrants could cause political  instability (they are mostly men, and tend to be poorly treated), but also because it is not cheap to provide food and water in the desert.

UAE 2014 .png

The US has often worked to influence or contain this Iranian potential. It pressured the Soviets and British to withdraw from Iran following World War Two, helped to overthrow Iran’s Prime Minister in the 1950s, helped to unleash the Arabs on Iran in the Iran-Iraq War of the 1980s, and in recent years has opposed the Iranian-allied regime in Syria and played Good Cop-Bad Cop along with the Israelis in threatening to carry out airstrikes on Iran’s nuclear and military infrastructure.

Going forward, however, the traditional relationships between labour, industry, and military power may be breaking down.

Labour may no longer be so important to industry, as industrial labour will in many cases be replaced by machines or, in the case of skilled labourers like engineers and computer programmers, may sometimes be outsourced using high-quality digital communications. Having a large labour force may perhaps even limit the ability to industrialize, since countries with large populations could have to deal with robot-caused unemployment, competition for energy between the residential, commercial, and industrial sectors of the economy, and protests against local industrial pollution. This could put the Gulf Arab states at less of an industrial disadvantage when compared to countries with larger labour forces, like Iran.

Meanwhile, industrial output and labour are both less likely to translate into military power than they once were. In past wars, like the Iran-Iraq War or the Israeli-Arab Wars, wars were fought by giving lots of soldiers lots of weapons. This is, for example, one reason why the Israeli-Arab wars never lasted long. Israel did not have a large enough population to run its military and factories at full capacity simultaneously, so it had to end wars quickly in order to avoid running short on supplies. Otherwise it would risk becoming too dependent on US supply lines, as occured when it was attacked by Egypt and Syria during the 1973 Yom Kippur War.

In the future, though, human soldiers may be replaced by machines in some cases, rendering population size less important in war. In addition, the quantity of weapons produced could continue to become less significant than their quality, given that weapons can now be destroyed by precise satellite-guided missiles. As a result, if the Gulf Arab states and Iran were to use their oil and natural gas reserves to become industrial powers, it would not have to translate into their becoming military powers. This could make existing Middle Eastern military powers like the US, Turkey, or Israel more likely to tolerate the industrialization of the Gulf.

The machine-driven industrialization of the Gulf Arab states and Iran could make sense for a number of other reasons as well:

— The Gulf region is even less populous than it may seem at first glance. This is because the vast majority of the populations of Iran, Saudi Arabia, and to a lesser extent Iraq live hundreds of kilometers away from the Gulf, and separated from the Gulf by mountains and desert. If you count only the provinces of Iran, Saudi Arabia, and Iraq that border the Gulf, the entire population of the Gulf region (including the smaller Gulf Arab states) is only around 30-35 million. It is also very hot there during the day, making physical labour difficult.


(Population Density of Saudi Arabia and Iran)


Topography of the Gulf (Iran is on the right-hand side); many of these mountains are as high as the Alps

— The Gulf region is rich not only in oil, but also in natural gas. Iran in particular is the world’s third largest producer of natural gas, and is thought to have the second largest reserves (the largest by far if you do not count Siberia). Iran also directly borders landlocked Turkmenistan, which is thought by some to have the world’s fourth largest gas reserves.

Natural gas, however, is very difficult to transport long distances or to store up in large quantities. Qatar has managed to become rich from exporting its gas in liquified (LNG) form, but this only works because Qatar has a tiny population (2.3 million people) and as LNG prices in Asia and Europe have been high. For the rest of the Gulf’s gas, it would be difficult to replicate Qatar’s success. It might make more sense, then, for the gas to be kept within the Gulf, used there to produce energy for industrialization.

Shift Data gas.png

shift data oil

Gas-fired power plants tend to be the cheapest types of power plants to build, and have efficiencies considerably higher than coal plants. Gas is also a feedstock in petrochemical industries like plastics. The Middle East already accounts for an estimated seven percent of global petrochemicals production.

Saudi Arabia

— The Gulf region is located at the centre of Eurasian and African trade routes, both by land and by sea. While all of the Gulf’s trade routes are politically fragile, this may actually make industrialization sensible, because it is easier to stockpile large amounts of manufactured goods or industrial raw materials for use during a crisis than it is to stockpile oil or especially gas. Industrialization would also give the region more economic autarky, which would be useful if its trade routes to the outside world were ever imperilled or cut. And the Gulf already possesses large industrial port areas as a result of its energy exports.

— Industrial areas which use machines can be clustered in ways that traditional, labour-based industrial areas cannot. Machines, after all, do not need lodgings, and are not bothered by pollution. Moreover, the Gulf itself could be a mega-industrial cluster, given that it has the world’s largest concentration  of cheap-to-produce oil and gas by far. In other words, you might be able to have a bunch of local industrial clusters forming a huge, region-wide industrial cluster.


Industrial clustering could have a number of advantages. First, it could be easier to defend militarily, which, given the enormous expense and difficulty of missile defence systems,
could be significant. In the Gulf, the energy fields, power plants, factories, and ports could all be concentrated in a fairly small, defended area. If you ignore coal, there is nowhere else in the world that comes even close to being able to have this. (Texas is probably the closest…).

oil costs.png

Second, clustering could perhaps help to allow for carbon capture and storage. Carbon capture and storage has thus far proven to be far from economical in most cases, and yet it is also necessary if the world wants to limit carbon emissions without ending consumption of fossil fuels. It could be that the way to make it economical is to cluster many power plants together in order to allow for economies of scale to form. Moreover, the only type of carbon capture and storage that has proven economical thus far is when carbon is used in Enhanced Oil Recovery (EOR), being sent down oil and gas wells in order to increase oil and gas productivity. The Gulf already uses EOR in some cases, and has huge EOR potential as a result of the size of its oil and gas fields.

US Oil Demand

One factor that could help spur the industrialization of the Gulf would be if people in the rich world, particularly in North America, would stop driving alone in their cars so much. By switching to alternative forms of transport — whether carpooling, taking public transport, UberPool-ing, e-commuting, shopping online, taking a self-driving bus or self-driving electric car, etc. etc. — it may leave Gulf oil available for industrial use within the Gulf.

Regional Political Triangle

This article has been a bit of a theoretical ramble, I know. But if we assume for a moment that the Gulf does manage to use its energy resources and robots to become a industrial powerhouse, what would happen to the politics of the wider region? India and Pakistan, after all, are located very close to the Gulf, putting a billion and a half people within its reach.

I think you might see something like a US-Gulf-South Asia political triangle emerge. The US has already been allies with the Gulf Arab states and Pakistan for many decades now (though of course, it is often a tortured alliance), so this would require only bringing Iran and India into the fold as well. India is already well on its way in this regard; Iran may perhaps be following behind it, given the recent warming of relations between the US and Iran during Obama’s second term.

In such a scenario, the Gulf would provide goods and oil to South Asia, South Asia would provide skilled outsourced labour to the Gulf, the US would provide food and military protection to both the Gulf and South Asia, and the Gulf and South Asia would help the US to put a lid on regional chaos and contain any European, East Asian, or Eurasian powers that may emerge over time.

Then again, maybe none of this will end up happening. The Gulf may be ludicrously rich in easy-flowing oil and gas, but translating that energy wealth into industrial success will be no easy feat, with or without robots.