Archive | May 2015

The Internet pipes are filling up with video

sandvine 1

According to Cisco’s report published yesterday, 90% of consumer Internet will be video by 2019. And Canadian company Sandvine’s report issued today shows us how we’re getting there. The chart above shows that on North American fixed line networks, 69% of all peak period downstream web traffic is “Real Time Entertainment”, up from 64% a year ago. At five percent per year growth, 90% by 2019 looks like a no brainer.

Some of that real time entertainment traffic is kitten videos, of course. But not the majority: the single largest driver of Internet traffic in prime time is Netflix, representing 36.5% of all bits, up from 34.9% only six months ago. That is a compound annual growth of over nine percent.

sandvine 2

As more people watch over the top video services (Netflix, but others too) for more hours and at higher definitions, the future is clear. The Internet of 2020 won’t be about telemedicine, distance education, or the connected devices.


The whole world isn’t watching


It feels almost trivial to talk about how much TV white North Americans watch versus minority groups. When we have hashtags like #blacklivesmatter in the US, and nearly 1,200 “murdered and missing” aboriginal women in Canada, who’s watching what may seem unimportant.

We know that attitudes around race are largely shaped by the media, and coverage of racial issues by the news media are the key (and sometimes the only) source of information for millions of North Americans. In 2015, the extent to which any medium becomes even more of an echo chamber or filter bubble is important: for society, for educators, for regulators, and for public policy.

In the US, all Americans over the age of two watched 5.5 hours of live or time shifted traditional TV per day in Q4 2014, according to Nielsen. The comparable figure for Black Americans was 7.3 hours, or 33% higher. The number for Hispanic Americans was 4.4 hours, or 20% less than the population as a whole. And the figure for Asian Americans was 3.3 daily hours, or 40% lower, and 55% lower than Black Americans.

Nielsen by race

Race is a breathtakingly sensitive subject, and very few people write about the figures above. But Nielsen is doing a real public service by publishing these figures: I can’t find racial demographics for TV viewing for any other country in the world!

Including Canada. But I do have some data on TV habits by language, from BBM Canada. Did you know that all French Canadian over the age of 18 watched 4.7 hours of live plus time shifted TV per day, compared to 3.9 hours per day for English Canadians, or 20% more in 2012/13 broadcast year? And if I use the latest numbers, my own guesstimate is that the gap has widened for the 2014/15 measurement period, and French Canadians are now over 22% ahead.

Canada TV Viewing English French

It would surprise me if there weren’t similar viewing differences by race in Canada as there are by language…or maybe even wider disparities? The reason I mention this is the decision on Friday by Rogers Media to close down a number of Asian news (and other) broadcasts on their Omni family of stations.

Although not a racial divide, younger Americans are also watching less TV than average: 18-24 year olds watched 42% less live or time shifted TV than all viewers 2+ in 2014.

If I put this all together, I get the following conclusions/questions:

  1. Not everybody watches the same amount of TV: some races, linguistic groups or age groups watch more or less.
  2. Broadcasters are making programming decisions based on those audience figures, as you would expect.
  3. The gaps that I am seeing (across all of race, language and age) are not stable, but are getting wider. In 2011 Asian Americans watched 50% less TV than Black Americans; it is now 55% less. English Canadians watched 7% less TV than French Canadians in 2004/5; it is now 22% less. And American 18-24 year olds watched 27% less TV in 2011; it is now 42% less.
  4. This is not bad for the TV industry. Although there are audience shifts, total viewing hours are pretty stable. The ability for advertisers to reach large and desirable audiences remains high. In fact, because TV is increasingly shifting towards smaller bases that watch more, it is becoming constantly more targeted, which is a good thing.
  5. But it does mean that the era of TV as a “big tent” which had near ubiquitous reach across all ages, races and languages is diminished. I am not talking about the usual “everyone watched the final episode of M*A*S*H” argument. In a 500 channel universe, fragmentation is old news. What I am referring to is a change in the ubiquity of the medium itself: to some growing extent, TV is no longer the fire the tribe gathers around to share information and create a common culture.
  6. Are we moving to new world where young people have one less (big) thing in common with older citizens? Where Black Americans watch TV news with one slant on police violence, and Asians, Hispanics, and non-Hispanic White Americans get an entirely different perspective, because of different media consumption habits?
  7. What does this mean for regulators and educators? And also for public sector broadcasters (PSBs)? I can’t fault a private sector company for moving scarce programming dollars to where the audiences are…but do PSBs have a moral obligation to deliberately try to broaden the TV filter bubble?

In 1968, anti-war demonstrators chanted “The whole world is watching” as they were gassed, beaten and arrested. And they weren’t exaggerating: young and old watched the TV news, black and white, Democrat and Republican.

What happens when the whole world isn’t watching any more? Instead, our own “demographically tailored” news medium shows us only what an algorithm determines we’d most like to see?

Duncan’s Reading List & Book Reviews: April 2015

2015-05-02 08.37.24

It’s a dogpile of books!

No airplane travel this month, and I do most of my reading in transit, so only 2,700 pages accomplished, but a nice and eclectic mix of fiction, genre fiction, and non-fiction.

Ready Player One – Ernest Cline, 2011

Virtually a perfect book for me: a serious science fiction and computer gaming fan who remembers the years around 1980, Family Ties, War Games and the music of Rush (!) really well. But that’s me, and the New York Times warned that the gaming aspects swamped the rest of the book. That didn’t bother me, but I could see it for others. Special shout out to (slight spoilers) some nice use of gender, race and sexuality in characters. If you’re like me, it’s an 11/10, but for a less serious (or younger) geek it might be more like a 6/10.

The Goldfinch – Donna Tartt, 2013

Although the book won a Pulitzer, and many positive reviews, people and reviewers whose judgement means more to me DETESTED this book. Witheringly described as “a children’s book for adults” I was prepared for the worst. It’s not that bad: yes, it relies too heavily on coincidence…but so did most The Luminaries (which I reviewed last month) and critics loved that book. I finished it, and enjoyed some portions. But I have no desire to ever read anything else by the author, and have to say she isn’t as great a writer as she thinks she is, and she needs a better editor. I also found the wallowing in drug addiction and alcohol by the various characters overkill. Less would have been more. 6/10

The Land That Never Was – David Sinclair, 2003

Non-fiction, this is the story of one of the greatest land frauds in history. Scot Gregor MacGregor convinced hundreds of his countrymen to emigrate to the land of Poyais in the 1820s, and thousands of Britons to invest in the scheme. But there was no such country. The story just isn’t that interesting: con men who send people to their deaths are evil, and people are greedy and easily conned. I didn’t feel like I learned anything new, nor was there enough original research. Even the villain wasn’t that compelling. 6/10

Forever Peace – Joe Haldeman, 1997

Last month I read Joe Haldeman’s 1974 novel The Forever War, and loved it. Forever Peace is constantly described as ‘not exactly a sequel.’ Fair warning, but it did win the Hugo and Nebula, so it must be pretty good, right? Nope: in the 23 years between the books, Mr. Haldeman appears to have lost all of his writing ability. I am not exaggerating – the book is confusingly written from two points of view, and the main thrust of the book (the use of remote controlled combat super-robots called soldierboys) turns in an implausible direction halfway through. The lead character is black, but the way society of 2043 reacts to his relationship to an older white woman just doesn’t make sociological sense. Not only did I not enjoy it, but it actually managed to taint my memories of the predecessor book! 2/10

The Long Earth – Terry Pratchett & Stephen Baxter, 2012

Terry was working on a multiple-Earths story at the same time he started writing Discworld books back in 1983. Those took off, and he left the concept alone. But after he was diagnosed with early-onset Alzheimer’s (which he died of last month) he partnered with Baxter to write a five volume series. Was it a bit of a cash grab? Cashing in on Terry’s reputation? Yes, but in the first book there some hints of Pratchett’s sense of humour, and sharp writing. One critic wrote that it was “much more like a Baxter novel than a Pratchett one.” Not without its charms, but 6/10 at best.

The Long War – Terry Pratchett & Stephen Baxter, 2013

At a guess, Terry was sicker and less able to help. Less humour, less imagination, and some gratuitous cruelty near the end that didn’t feel consonant with the normal Pratchett style. 4/10, and I don’t think I will buy Long Mars, which is volume three, and released in 2014.

And I already have 2,200 pages of books picked out for May!

No, they haven’t invented warp drive!


The media is screwing things up again. With headlines about “warp drive”, they are not only getting the physics wrong, they are also misunderstanding Star Trek. Here is the most egregious example I have found so far:

Warp drives that let humans zip around other galaxies may no longer belong purely in the realm of science fiction. NASA is believed to have been quietly testing a revolutionary new method of space travel that could one day allow humans to travel at speeds faster than light.

It is important to note that neither the scientists involved nor NASA have even mentioned faster than light travel. They are looking at a new kind of technology, called electromagnetic (or EM) drive, which may be a breakthrough.

All conventional space drives work through taking something (whether rocket fuel, charged atoms, or even atomic particles), accelerating it, and having your ship be pushed in the opposite direction to the discharge. Classic Newtonian mechanics, but the problem is that to keep moving you need large supplies of propellant.


EM drive claims to offer (at this time) very small thrust without using any propellant at all, which is revolutionary. Frankly, I am skeptical, but even if the drive works, it does NOT allow you accelerate a spacecraft through the speed of light. That limit exists as a result of fundamental limitations of space time, and has nothing to do with the potential virtues of EM drive or any propellant-less technology. But I do know why the media is getting this wrong!

On the various Star Trek TV series, the ships have three ways of moving through space. They have ‘thrusters’ for high precision maneuvers over short distances (which are just rockets), and something called ‘impulse drive’, which accelerates plasma using magnets, and gets up to velocities like half of light speed. But in order to catch a speeding Romulan, they switch over to ‘warp drive.’


Warp drive has two interesting properties in the science fiction world. Most importantly, it creates a ‘warp bubble’ which changes the physics of space time, and allows the ships to go faster than the speed of light. Warp 1 is light speed, and Warp 2 is 8x light speed, while Warp 3 is 27x: the formula is v = w3c. Without that, all of this bopping around the galaxy would take centuries, and interstellar travel wouldn’t fit into hour long TV episodes! 🙂

The second property of the warp drive was that it used tiny amounts of matter and antimatter to power the warp bubble, but did NOT use traditional Newtonian mechanics. In that way, AND THAT WAY ONLY, the proposed EM drive is indeed kind of like a warp drive.

But none of this changes the current view that the speed of light is an absolute limit. EM drive may allow us to get closer to the speed of light, it may be faster or cheaper or more reliable. But nothing in it allows us to go over 299,792,458 metres per second.

Which is too bad: the universe would be a much more interesting place if going from star to star took hours instead of years. That’s why almost all science fiction books, movies and TV shows have to come up with some trick like warp drives to get around the light speed barrier. But just because faster than light drives are a necessity for interesting fiction doesn’t (yet) make them a scientific reality. Despite what the headlines say…