Thursday, October 13, 2016

The government role in shaping the Internet in China and the US

The US government invested $124.5 million in building and demonstrating the feasibility and value of the ARPANet, TCP/IP internetworking protocol and subsequent deployment to education and research organizations with the establishment of NSFNet and the NSF developing nations program. That was a pretty good investment. (Government procurement was also important -- for example, lessons learned and programmers trained in building the SAGE early warning defense system played a key role in the progress of networking and computer science in the US).

Internet development timeline

The Internet is
dumb by design.
NSFNet was at first the backbone for the international Internet, but the government stepped back once the initial research and development was completed, phasing out the NSFNet subsidies over a four year period. Private telephone and cable companies with local monopolies or oligopolies became Internet service providers and owners of our Internet infrastructure.

In contrast, the market for Internet applications and services was competitive from the start. (Note that the Internet was designed to be "dumb" -- to move packets of data as quickly as possible between "smart" computers that would connect to it and host innovative applications and services).

Private capital has financed and developed our Internet startups and ecosystems of organizations to support them -- incubators, accelerators, co-working spaces, investors, consultants and hackerspaces. The first startup ecosystems were in the Silicon Valley and along Massachusets Route 128, but many, including my home town, Los Angeles, have followed their lead.

1,113 tech startups in Los Angeles

535 startup support organizations

The Chinese have taken a different approach to Internet infrastructure. They established an academic network in 1987, linked it to Stanford University in 1993 and the following year established a full Internet connection. Like all other nations, they began with a slow link to a small academic network, but within a few years, the Chinese had realized the importance of the Internet and had established competing, government-owned national backbone networks.

Unlike the United States, the Chinese government is also playing an active role in support of Internet application and service companies. In his 2016 Report on the Work of the Government, Chinese premier Li Keqiang stated that
Further progress was made in implementing the strategy of innovation-driven development, the penetration of the Internet into all industries picked up pace, and emerging industries grew rapidly.
The Chinese are funding startup ecosystem organizations like those created by the private sector in the US. For example, 710 subsidized startups are clustered in the Dream Town district of Hangzhou.

Chinese workers remodeling Dream Town spaces

How has it worked out?

The US had a significant lead over the entire world when a small Chinese academic network joined the Internet. One cannot directly compare the US and China -- there are many confounding differences, but we can compare China with India, a nation with some similarities. India's academic networks joined the Internet a little before China, but they essentially started at the same time. By 2002, the Chinese Internet was significantly more advance than that of India.

The International Telecommunication Union's Information and Communication Technology development index (IDI) provides a recent indicator of Chinese progress. The IDI is a function of eleven indicators measuring ICT access, use and skills in a nation. In 2015, China ranked 82nd on the IDI while India ranked 131st. In 2010, China ranked 87th and India ranked 125th.

While China has done better than India to date, their economy is slowing and their program of support for startups can lead to misallocation of resources and cronyism -- as happened in the Chinese construction industry:

The US started at the top, so our infrastructure had no where to go but down relative to other nations. The US ranks 21st in fiber deployment among the 34 OECD nations and our average Internet connection speed ranks 15th among the 74 nations served by Akamai.

Percent of broadband connections with fiber,
OECD, December 2015

Connection speeds, Akamai, Q1 2016

While the US has faltered in infrastructure deployment, we have retained the lead in Internet applications and services. As shown here, the US had 39 of the most popular 100 Web sites as ranked by Alexa in September 2016 while China had only 10.

These rankings should be taken with a grain of salt – SimilarWeb, an Israeli company, rates the US substantially higher with 40 sites in the top 100 as opposed to 10 for China. Chinese sites also focus heavily on China whereas the US sites generally seek a global audience.

It seems that the US private sector has done well in Internet applications and services and not as well in deploying infrastructure. Private companies seek to maximize corporate profit rather than social goals regarding education, the economy, etc. This has contributed to other nations surpassing our infrastructure in spite of our pioneering role.

These are complex issues and I cannot reach definite, generalizable conclusions (nor can anyone else), but, if I had to guess, I would think that China would be better off cutting back on subsidies for Internet application and service companies and the US would be better off with a more competitive Internet service provider market -- discouraging consolidation and "gentleman's agreements" among companies and encouraging ownership of infrastructure by local governments and customers.

Update 10/16/2016

For teachers or others who might like to present this topic, I have prepared (and used) a nine-slide, one-video annotated PowerPoint presentation.

Tuesday, October 04, 2016

Vision precedes engineering prototypes which precede products

Google unveiled voice controlled "intelligent" assistants today -- in a phone and a home listening device.

The vision of a voice-controlled intelligent assistant in which the manufacturer made tightly integrated hardware and software was shown in Apple's 1987 Knowledge Navigator concept video.

We'll see if Google has pulled it off.

Monday, October 03, 2016

Google Hangouts on Air is a useful, evolving teaching medium

Google Hangouts on Air is a useful tool for teaching today and it will be different and improved tomorrow.

Google Hangouts on Air (HoA) is a free service within YouTube. It allows one to hold a live teleconference with up to ten people. During the teleconference, an unlimited number of people can watch the live stream and it is recorded. When the teleconference ends, the video is automatically archived on YouTube. (See this presentation for some examples and instructions on how to create an HoA).

I've experimented with HoA for teaching three times. The first time, my class met online when I was out of town in 2013. It was a complete failure -- the Internet speed and student hardware were not up to the task.

The second time, the results were good, but we made and learned from some mistakes:

The third time -- a couple weeks ago -- we avoided most of those mistakes (and made a couple new ones). I am ready to use HoA more frequently, even when I am not out of town.

I polled the students the week after our last HoA session. Eight students were in the live hangout, 2 watched the live stream, but were not in the live hangout, 8 others only watched the archived video later and two did not participate. (A couple were also absent the night I did the poll). Here are the results:

This is a small, anecdotal sample but it is consistent with the assessment of the students last year. We also talked about the experience in class the following week and the students confirmed that they were batter able to focus in the hangout than in the classroom and were more relaxed and willing to participate. (Most also saved travel time by participating from home). The students who were in the live hangout had a generally better experience than those who only watched the video.

We are entering an era of live-streaming on the Internet -- Twitter is streaming football games and a number of Web sites streamed the presidential debate. Google, Facebook and others are offering streaming services and, like all other media (including textbooks), live streaming will evolve and improve. Google HoA is a useful tool for teaching today and it will be different and improved tomorrow.

Thursday, September 29, 2016

Hybrid fiber-wireless tests by Google and others

Hybrid fiber-wireless technology will improve our access to the Internet and alter infrastructure ownership and business models.

Google Fiber installations run fiber all the way to the premises, but that is costly and the subscription rate has been disappointing, causing Google to slow installation and cut staff. It seems they are looking for a hybrid fiber-wireless connectivity solution.

In an earlier post, I noted that Google was experimenting with high-frequency wireless for possibly providing "last kilometer" links from homes and offices to Google Fiber. In a recent FCC filing, they requested approval for high-speed wireless testing in 24 US cities.

Even more interesting is their recent acquisition of Webpass, a boutique high-speed Internet service provider that has been experimenting with "PCells" an experimental wireless technology developed by Steve Perlman of Artemis Networks. I first heard about PCells when I saw a video of a talk and demo that Perlman gave at Columbia University, but you should start by watching this shorter, more recent demo:

There are other videos and tech papers on the Artemis Web site.

Perlman's demos are impressive -- computers, phones and tablets that are inches apart receive full-speed wireless connectivity and they freely move around without losing contact. Transmission speed slows a bit as they are moved, but full speed resumes as soon as they stop moving.

PCell access points are small and distributed compared to conventional cell towers and, if they turn out to be simple and effective, they could be installed and owned by individual users as well as companies --forming next-generation "street nets," connected to Google's, or anyone else's, fiber.

The jury is still out on PCells, but it is safe to say that they or other fifth generation wireless technology will improve our access to the Internet and alter infrastructure ownership and business models.

A PCell access point

Monday, September 26, 2016

Live streaming of football games and presidential debates, version 1.0

I would also be willing to pay for interaction -- perhaps asking a question or making a comment of my own.

There has been a lot of discussion of the appropriate role of moderators in the presidential debates -- should they challenge the debater's statements if they are false or simply ask questions and let the candidates challenge each other? That will become a moot point in the era of live-streaming on the Internet.

Twitter is streaming Thursday night football games -- you see the game as well as the stream of tweets. I reviewed the experience and concluded that there were too many commercials and that I wanted to be able to filter the tweets. I am not interested in the comments of a million football fans, but would like to be able to follow the live-stream comments of experts like a group of professional football players, sports writers, Las Vegas odds-makers, etc.

Now, let's apply that to the live streaming of a presidential debate. (Note that Twitter is streaming the debates).

There will be plenty of commentary in the tweet stream accompanying Twitter's video and the audience will fact-check the debaters regardless of what the moderator does. But, as in a football game, I don't want to see tweets from every partisan viewer, I want to see informed, factual commentary.

If I could watch the streaming comments of a relatively small group of experts along with the video stream, I would not care whether the moderator challenges the debater's statements or just reads questions and keeps the speakers on schedule. I would also be willing to pay for interaction -- perhaps asking a question or making a comment of my own.

The asynchronous Internet has changed political campaigns just as newspapers, radio and television did. Now we have a new medium -- live streaming over the Internet.

Early movies, like those shown below, were made by filming stage plays. Similarly, today's live coverage of football and presidential debates is just a starting point. Adding expert or crowd-sourced fact checking and commentary would only be a small variation on what Twitter has already done with football games. There will be more changes as the medium co-evolves along with our culture and education system.

Sunday, September 25, 2016

Has the Internet enabled lying, crooked Donald?

We are in the early days of the Internet as a political medium and hopefully it will co-evolve along with our society and education system.

Last June, Donald Trump began calling Hillary Clinton "lying, crooked Hillary" and established a Web site of the same name. Leaving Trump's coarseness aside, is the allegation fair? (His coarseness calls for a separate post).

Politifact is a fact-checking Web site run by a Florida newspaper. They rate political statements on a six-level scale ranging from True to Pants on fire:

True – The statement is accurate and there’s nothing significant missing.
Mostly true – The statement is accurate but needs clarification or additional information.
Half true – The statement is partially accurate but leaves out important details or takes things out of context.
Mostly false – The statement contains an element of truth but ignores critical facts that would give a different impression.
False – The statement is not accurate.
Pants on fire – The statement is not accurate and makes a ridiculous claim.
They justify their ratings with reasoned, sourced analysis and have been awarded a Pulitzer Prize. (You can read the details on the rating rubric here).

The following are summaries of the Politifact ratings of statements by President Obama, Hillary Clinton and Donald Trump. (Click the image to enlarge it).

As you see, Clinton is a bit more honest than President Obama and lies much less frequently than Donald Trump. The ratings of Obama and Clinton have changed little since January. Trump is telling the truth a little more frequently, but over half of his statements were found to be lies.

I retrieved the September ratings from Politifact this morning and retrieved the January ratings using the Internet Archive Wayback Machine.

I guess all politicians lie, but few, if any, lie as frequently as lying, crooked Donald. (The "crooked" part calls for yet another post on his business dealings).

The Internet has changed political campaigns just as newspapers, radio and television did. Candidate's statements are archived and Politifact and others can analyze them, but the Internet also enables the dissemination of lies like this faked image of Hillary Clinton and Osama bin Laden, which can be found on many Web sites:

(To be fair, a lot of democrats shared a fake image showing President Bush holding a picture book upside down at the time he was informed of the 9/11 attacks).

The Internet also increases the odds that we will see lies we might "like." As Eli Pariser points out in his book The Filter Bubble, ad-driven sites like Facebook have an incentive to send us things we agree with to keep us on their sites longer.

The Internet enables us to easily create and disseminate lies and it also enables us to discover and expose them, but does that matter? Has the Internet brought us to what William Davies calls the age of post-truth politics? After all, Politifact shows that over half of Donald Trump's statements are lies, yet millions of Americans are willing to vote for him. While Hillary Clinton and President Obama lie less than Trump, they also have millions of supporters who are ignorant of or indifferent to their lies.

That is discouraging, but remember that we are in the early days of the Internet as a political medium and it may co-evolve along with our society and education system to bring us something better. For perspective, check out this early use of television in a political campaign:

Update 10/23/2016

The Internet facilitates the repetition of lies and exaggerations and we tend to become more polarized as search engines and news services show us things we are likely to agree with. On the other hand, the Internet facilitates fact-checking services and Duke University tracks over 100 such sites:

Interactive map of over 100 fact-check sites

The Internet, like other technologies, can be used for good and bad -- the Internet doesn't tell lies, people tell lies.

Friday, September 23, 2016

Football streaming on Twitter -- too many commercials and need to be able to filter tweets -- but like all new media, it will improve.

I watched a bit of Thursday Night Football on Twitter last night. You could watch a small screen with tweets as shown above or go full screen and lose the tweets. I watched it on a laptop with a large, high resolution screen and on a Mac with a 21-inch display and the video was smooth and looked good on both. That was the good news.

The bad news was the commercials. I am not a football fan, so do not know how many commercials a typical broadcast game has, but it seemed like Twitter spent more time on commercials than the game. I would be curious to see statistics on the number of minutes spent on commercials, commentary and game action on Twitter versus broadcast television.

I “cut the cord” years ago, so am used to paying Netflix and others for streamed content without commercials. If I am typical, Twitter will fail with this commercial-based business model. (The Motley Fool Web site says the ads did not pay off in the first game, which was streamed last week).

For a while, I watched both the TV broadcast and the Twitter stream. The Twitter stream was relatively delayed, but the lag time varied and they did not have the same commercials. I wonder how the commercial sales and revenue are handled.

Turning to the user interface – I did not time it, but it seemed like there about 20 tweets every thirty seconds. With that many tweets coming in, I think the best way to watch would be to go full screen during the live play and mute the audio and read tweets during the commercials -- not a good deal for advertisers.

I did not notice any obviously malicious tweets, so I assume there is some automatic or human filtering, but it would be better if they would let the user control the filtering. For example, to let one see only tweets from a selected group of friends or a selected group of experts like professional football players, sports analysts or professional gamblers.

But, lest I seem too negative -- this is their first try at streaming sports. All new media stumble at first, often copying what came before. The first movies were made by filming stage plays and there are many other examples from radio, TV, textbooks, online learning, etc.

Twitter is streaming the US presidential debates next -- let's see how they do on that.

Thursday, September 22, 2016

Verizon wouldn't lie to sell phones -- would they?

In previous posts, I have been unkind to my local-monopoly Internet service provider Time Warner cable and the US ISP industry in general. I've also been unkind to Verizon FiOS in their battles with Netflix and criticized their "gentleman's agreement" to abandon fiber to focus on wireless connectivity, leaving me at the mercy Time Warner Cable.

Those stories all had to do with landlines -- what about Verizon mobile? My wife had an unlimited account with a reseller of Verizon mobile service (a "mobile virtual network operator" or MVNO). She no longer needed the unlimited account, so decided to switch to Verizon.

This was shortly before the new iPhones came out, so she wanted to keep using her old phone for a month or so and, since she had been using it on the Verizon network, she assumed it would work after shifting her account from the MVNO.

To be safe, I went online and had the following chat with a Verizon salesman named Brandon:

Chat transcript -- click image to enlarge

As you see, I gave him the phone's mobile equipment ID (MEID) number and he said it was incompatible with the Verizon network and offered to sell me a new phone. I pointed out that the old phone had worked on the Verizon network for years and he suggested that it may have been blacklisted or have the wrong antenna -- like AM versus FM radio. He elaborated, saying it might be compatible with some, but not all, of their network or perhaps my wife had been roaming for five years.

Maybe the phone would not work somewhere on Earth, but it has worked everywhere my wife has been in the United States and abroad for the past five years. (She is not an early adopter :-). Here are the specs (MEID 990001106522642):

I am not a mobile phone geek -- Is there something that would render the phone incompatible with Verizon mobile service?

We ignored Brandon's warning and opened a Verizon account -- the phone worked fine (in southern California) until it was replaced with a new iPhone.

My guess is that Brandon was just telling me what he saw when he queried Verizon's database, so he was not lying. But is the Verizon compatibility database accurate and, if not, is Verizon lying in order to sell new phones? (This reminds me of the Volkswagen smog check shenanigans).

Update 10/10/2016

Let's talk about customer service.

In spite of Verizon saying it was incompatible with their network, my wife has been using her old iPhone 4 on their network without problems. When she transferred her account to Verizon, they said she had to sign up for a billed (post-pay) account, but could then switch to an automatically-charged, pre-paid account (to get a higher data cap) whenever she wanted to.

She did that two days ago. Doing so required that she speak on the phone with three different Verizon employees -- a support person, a sales person and an account verification person. There were long waits on hold before each of those conversations and she had to explain the situation to each of the people -- there was evidently no transfer of information between them. The entire process took well over an hour.

I am posting this for two reasons:
  1. Verizon has sold their FiOS landline service, but it was the lowest ranked company in the lowest-ranked industry (Internet service providers, ISPs) on the American Customer Satisfaction Index before the sale. I've not seen a ranking for mobile ISPs, but based on this experience, I would expect Verizon to look bad.
  2. One frequently hears complaints about poor service from bureaucratic government agencies. Government agencies are large organizations with monopoly power. ISPs are also large organizations in monopoly or oligopoly markets. A large, private organization in a monopoly or oligopoly market is nearly as likely to provide poor service as a government agency.

Update 10/15/2016

The customer-service saga continues.

I made two payments on my wife's pre-paid account. Perhaps because she changed her account from post-paid to pre-paid, Verizon failed to credit one of the payments and does not seem to realize that she is now post-paid and should be getting a higher data cap.

I tried to clear it up by going to the "My Plan" page on the Verizon Web site, but the page was temporarily down:

That was 25 hours ago. The page is still temporarily down.

I also tried to clear this up with a phone call or chat, but could only find sales pages on their Web site. (I got the phone number last week from a Best Buy store clerk).

Verizon is making the folks like the IRS or Department of Motor Vehicles look good.

Thursday, September 15, 2016

The Internet revolution in perspective

Google, Facebook, self-driving cars, iPhones, etc. are changing our lives, organizations and society, but this is not unprecedented. Similar scientific and technical disruption occurred about 100 years ago. Guglielmo Marconi invented and commercialized electronic wireless communication at that time and, in his review of a book on the life of Marconi, Paul Kennedy writes that in the first decade of the 20th century:

Breakthroughs in science and technology occurred so often that it would be brash to claim that any one of them “changed the world” (which doesn’t stop proponents from doing so). The Wright brothers’ success in aviation in 1903 led to national air forces being created only a few years later. The automobile was becoming reliable, standardized and produced in such numbers as to change urban landscapes. Giant trans-Atlantic liners altered oceanic travel. Electric power was coming to houses and oil-fueled propulsion replacing coal-fired engines. The Dreadnought battleship (1906) made all other warships obsolete.
In his biography of Albert Einstein, Walter Isaacson writes that:
In 1915 Einstein wrested from nature his crowning glory, one of the most beautiful theories in all of science, the general theory of relativity ... His fingerprints are all over today's technologies. Photoelectric cell and lasers, nuclear power, fiber optics, space travel and even semiconductors all trace back to his theories.
This is not intended to diminish the impact of the Internet on our lives, organizations and society, but to lend perspective. The Internet is disruptive, but so were the printing press, number systems, phonetic writing, agriculture, the recognition of natural cycles, spoken language, etc. What else?

Wednesday, August 10, 2016

NBC streaming the Olympics on the Internet -- quantity, not quality

NBC Sports is allowing many (not all) cable TV subscribers to stream Olympic Games events live and to watch archived copies after they finish. There are several ways to watch the streams -- in a Windows 10 app, on a TV set using the Roku app and on the Web. I checked them out and was generally disappointed.

With the Roku app, watching a live event is like watching a broadcast TV program -- you sit back and watch, but have no control. When you are watching the archived recording of a completed event, you can pause and fast forward/reverse, but that is all.

The event navigation interface is also lame. There are five linear menus: Features, Live and upcoming, Highlights, Full event replay and Sports. You scroll through a list of thumbnail images to find the one you want to watch. That is fine for the 13 features and 34 sports, but not so handy for the 300 Highlights, 771 Live and upcoming or 538 Full event replays. (These numbers will vary of course). Search is sorely missing.

There are also frequent delays for updating content and, if you try to watch a live event, you are often informed that it has concluded or will begin shortly -- they do not update menus at the time an event starts or ends.

The Windows 10 live-event user interface is a little better, but nothing to write home about. As you see below, there is a volume control and buttons to pause/resume, go full screen and to turn on/off closed captions.

NBC Sports app user interface

The navigation interface, shown below, is similar to that of the Roku. There are five linear menus: Features, Live and upcoming, Replay, Highlights and Sports. There are no scroll bars, so you use the cursor control keys to move through the selections.

Windows 10 app navigation interface

As with the Roku, the "live" events are frequently either completed or not yet begun.

The Web user interface is a bit better. As you see below, it includes the handy "15-second rewind" button and rewind and fast-forward buttons.

Web user interface

The Web navigation is also more flexible and complete than with the Roku or Windows 10 app. The bad news is that there are ads and you cannot go full screen. (The others also have interspersed ads).

There is also an Android app, but I am not going to bother installing it -- I prefer watching sports on a TV set and it will doubtless have some of the limitations discussed above.

As I said at the start, I am generally disappointed with NBC's streaming coverage. Part of that is because my expectations were raised by NBC's excellent coverage of the Tour de France. The mobile and desktop user interfaces were clean and powerful and they presented real time data during the race and inciteful analysis after each stage was completed. The business model was also different. You pay for Tour de France access, but do not have to watch ads (which are often followed by a screen saying the event you want to watch has already concluded).

I realize that the Olympics are a tougher event to cover -- many venues and many elimination levels -- but I hope that by 2020, the NBC Olympic team will sit down with the Tour de France team and re-design their coverage. I hope they also offer an ad-free option.

Thursday, July 28, 2016

The digital divide has persisted over the life of the Internet.

National Bandwidth Potential, a novel Internet diffusion metric indicating application feasibility, shows a persistent digital access divide.

People have been trying to measure the global diffusion of the Internet and the digital divide between rich and poor nation for twenty five years. The first to do so was Larry Landweber, who noted whether or not a nation had an Internet (or other) connection. It was a binary metric -- yes or no -- and it was suitable to its time because there were only a handful of users who were restricted to teaching and research, using a few applications like email, file transfer, news groups and remote login.

1991 Internet diffusion (purple)

Five years later, the Internet had many more users and applications in commerce, government, entertainment, etc., so my colleagues and I developed a multidimensional Internet diffusion framework. One of our dimensions was pervasiveness, based on the number of users of the Internet per capita.

That made sense in 1995 since there were relatively few applications available for the slow dial-up connections of the time. A few people had faster ISDN or DSL connections and an organization might connect over a faster digital link, but most users were running the same few applications over analog phone lines.

Today, users per capita is pretty well meaningless. A Cuban who accesses email using a 2G cell phone and a Google Fiber user who has symmetric gigabit access to multiple computers and devices on a home LAN are clearly not equal.

To some degree, we anticipated this sort of thing via the connectivity infrastructure dimension in our framework. It considered international and intranational backbone bandwidth, Internet exchange points and last-mile access methods, but it was an imprecise measure -- mapping a nation into five levels -- and data was not readily available. (Our case studies typically required two weeks of in-country interviews).

Skipping ahead twenty years, a paper by Martin Hilbert uses an interesting diffusion metric -- nationally installed bandwidth potential (BP), which is a function of the number of telecommunication subscriptions (fixed and mobile), the kind of access technology per subscription (cable, DSL, GSM, etc) and the corresponding bandwidth per access technology. Their estimation of the latter is quite complex, taking factors like data type, upload/download speed, compression, etc. into consideration. The methodology is described in a ten page supplement to the paper. (It is behind a paywall -- let me know if you would like a copy).

Hilbert computed the BP of 172 countries from 1986 to 2014 and observed that the digital access divide is persistent. It is true that wireless connectivity is relatively inexpensive and mobile Internet use is growing rapidly in developing nations, but it is just as clear that many applications are precluded by the speed and form factor of mobile devices. A WhatsApp chat with a friend is not the equivalent of watching a high-resolution movie on a large screen TV and I am confident that Hilbert did not conduct his research or write the paper I read on a mobile phone. Even reading this blog post, following its links to other documents and taking notes on it would be tedious on a phone.

I expect this imbalance to persist because improved technology is costly and it enables ever more complex, demanding applications. The only trend I see that may in part reduce this feasible-application gap is the move to server-side processing for big data and AI applications, but even then interaction and the display of results will require bandwidth.

Hilbert's data also shows global shifts in application feasibility. As shown below, BP dominance has shifted from the US in the early, NSFNET days to China today. Korea has joined the top ten and the shares of Japan and Western Europe have dropped. The share of the bottom 162 countries rose slightly in 2001, but had fallen below the 1986 level by 2014.

Ten countries with most installed bandwidth potential

Income differences explain much of the persistence of the digital divide, but policies regarding Internet infrastructure ownership and regulation are also important. For example Estonia ranks 40th in the world in GDP per capita, but is ranked 20th on the International Telecommunication Union ICT Development Index.

Policy choices may play an even larger role among the top ten nations. The US ranks 9th in GDP per capita and Korea is 30th, but my son, who lives in Korea, pays $22 per month for symmetric, 100 mbps connectivity and has a choice of several competing Internet service providers. I live in the US and pay considerably more than he does for considerably slower service and have no ISP choice -- I am stuck with Time Warner Cable.

While we are waiting for enlightened policies, we can hope for technical change like the OneWeb and Spacex satellite Internet projects.

Tuesday, July 12, 2016

Coverage of the 2016 Tour de France -- big data

Gathering real time data on each rider enables a clean video user interface, real time presentation of the race status and post race data analysis.

For several years, I wrote posts on streaming coverage of the Tour de France, Olympic Games and the Tour de California. Those posts focused on topics like user interface, ads, video quality and comparison of NBC's coverage with that of the BBC.

I missed last year due to travel, but am watching the current Tour de France, and there have been significant changes for the better.

For a start, NBC now bundles coverage of the Tour de France with several other races, so one purchases an annual subscription. That means cycling fans can see more races and, presumably, that the archive footage will remain accessible at least during the year.

(In the past, both NBC and the BBC have deleted their archives some time after the end of the Tour. I believe they have an information stewardship obligation and should maintain the archives of important events for analysis by journalists, scholars, fans, remixers, etc. The cost of doing so would be low and, if they were not behind a paywall, they could be found by search engines.)

The video quality is also better than I recall -- a consistent 2.2 mbps stream with none of the dropouts we saw during 2014.

The user interface has been simplified since 2014 when it had five modes -- live video, standings, stages, riders and more:

2014 Five viewer modes

and you spent most of your time in the four-frame Live Video mode:

2014 four-frame Live Video user interface

By contrast, the live video UI this year is simple, with a small race status indicators like the time between the race leader and peleton in the screenshot below, popping up from time to time on a full video screen with customary controls at the bottom:

2016 live video user interface

At first stripping out ancillary information might seem a step backward (or forward if you are an Apple minion), but it is not. Much more ancillary information is available this year and it is accessed through a "Tour Tracker" site. The Tour Tracker allows you to see in-depth information for each stage, with tabs for Teams, Stages, Standings, Results, Recaps, Replays and Photos and a link to the live video window shown above.

2016 Tour Tracker user interface

All of that data is available because the race is now very well instrumented. Each bike has a small GPS transponder affixed to the seat.

GPS transponder

The data from the transponders is uploaded to the mobile data center of Tour partner (and team sponsor) Dimension Data, enabling them to provide live data during race -- check out the following video (2m 50s).

This data collection enables Dimension Data to provide real time status of the race, individual riders, teams, etc. In the example below, we see the speed of several riders, the time gaps between them and the distance from the leader to the finish superimposed on the live video window.

Status update on Live Video viewer

In addition to real-time status statistics, Dimension Data is able to analyze data after a stage is complete. For example, the following image shows that the stage 6 sprint winner, Mark Cavendish, accelerated a little bit later than the second and third place finishers. I would expect that this sort of data is helpful to the racers and their managers. (The teams receive some information that is not available to the general public).

Post stage analysis

GoPro cameras are another source of Tour data. Since 2015, GoPro has been a Tour partner and they had cameras on cameras on official cars and motorbikes, team cars, mechanics and selected bikes. Fellow Tour enthusiast Jim Rea spotted some live GoPro footage during stage 1, but has not seen any subsequently. That being said, you can see archived video after the stages are complete by searching on Google for "GoPro: Tour de France 2016 - Stage n Highlights", where n is the stage number.

The following video is not from the Tour de France, but it shows what it is like to be in the sprint at the end of a race.

And the following video shows a crash from a mechanic's point of view:

(A 360 degree virtual reality versions would be cool).

The NBC package also includes a free mobile app. The live video on the Android app has a simplified user interface with only a pause/play toggle. The data view offers the options shown below, but it is not as complete as that of the Tour Tracker Web site.

I watch The Tour on a wide-screen laptop with a 3200 by 1800 pixel display and toggle back and forth between the Tour Tracker and video windows. An alternative would be to run the mobile app on a smart phone, casting it to a TV set and using a laptop or tablet for the Tour Tracker.

The bottom line is that race coverage had improved significantly since I last watched The Tour. The video quality has improved noticeably and the addition of real-time GPS data has added to the experience. I didn't even bother trying a VPN tunnel to check the BBC coverage.

Tuesday, May 31, 2016

Brick and mortar stores -- Apple, Microsoft and Google?

Dell, HP and others now have relatively upscale Chromebooks that approach, and in some features surpass, the high end Google Pixel and Google just announced that Chromebooks will be running Android apps in the future. At first, those apps might not be optimized for the Chromebook form factor, but many will look good in phone or tablet-size windows and I bet we see Chromebook-friendly Android apps in the future.

Given all that, I thought I might like to get one, so I headed over to the closest thing I know of to a Google store -- the Google section of my local Best Buy.

It's a total Fail.

As shown here, all they had was half a dozen low-end machines. That might work for a Chromebook for a school child, but it is not sufficient for someone thinking of spending $700 or more.

But, it gets worse.

There were two, sweet, young sales people wearing Google shirts next to the Chromebooks, so I asked if they had other machines -- perhaps a Pixel -- somewhere else in the store. It turned out they didn't know what I meant by "Google Pixel." I explained what a Google Pixel was and one of them went off to inquire. When she came back, she said they did not have them.

Since I was there, I asked about the six machines they had on display and discovered that they were confused about the difference between memory and storage. None of the machines on display had more than 2GB of memory, but they assured me that that was no problem because you could attach a large external hard drive.

(In the early days of personal computers, there was a joke that the difference between computer store sales people and car sales people was that the car sales folks knew they were lying).

I don't know if these kids were Google or BestBuy employees, but they were wearing Google shirts and that surely cheapens the top-notch "Googler" brand.

If Google hopes to sell and support high-end hardware, they will have to do much better than this, and that will be expensive.

A little while ago, I had been in a shopping mall near my home and dropped in on the Apple and Microsoft stores, which are just a few stores apart.

It was the middle of the week, but the Apple store was quite crowded. Customers were talking with sales people, playing around with machines, getting help from Apple "geniuses," etc. Apple runs classes in the stores, offers walk-in customer support and the employees are knowledgeable and helpful. I snapped this picture just before the man in the foreground told me to stop taking pictures:

I walked over to the Microsoft store and found it to be pretty well empty -- the store employees outnumbered the customers. They had a wide range of computers on display -- from both Microsoft and OEMs. They also offered service and classes and the workers were as knowledgeable and friendly as those in the Apple store. There was no pressure and no problem playing around for as long as I wanted to and they were happy to have me take pictures.

I had visited the same Microsoft and Apple stores two days after Christmas in 2014 and, while both were more crowded post Christmas, the Apple store was totally jam packed and the Microsoft store still fairly empty.

I personally don't see much difference between the Microsoft and Apple stores and can't figure out why one is so much more popular than the other, but, I can tell you for sure that Google will have to be creative and spend a lot of money if they want to sell us high end hardware. They will also have to step up customer support. You can sell a $35 Chromecast in a BestBuy store or online, but not a $1,300 Pixel Chromebook.

Monday, May 02, 2016

Two teaching experiments with Google Hangous on Air

I teach a class on the applications, implications and technology of the Internet and a major theme running through the course is the use of the Internet as a tool for collaboration. As such, we tried two teaching experiments using Google Hangouts on Air (GHoA).

(GHoA is a free video conferencing application for up to ten people. It differs from other video conferencing services in two ways -- an audience of unlimited size can watch the video conference while it is live and it is automatically recorded and stored on YouTube when it ends).

Student "office hours"

Our first experiment was having students who had done well on the midterm hold "office hours" online using GHoA. I did not participate in any of the sessions, but reviewed the videos afterward.

Students holding "office hours" online

Since I was not "present," the students were generally unguarded and light hearted, talking more freely than in class. Their discussion revealed a couple of content misconceptions, which I corrected the following week.

They also discussed the class itself. One group agreed that it was harder than they had expected and one group felt free to criticize the class. That gave me the opportunity to bring their criticism to the entire class, discuss the point they made and to give the most critical student extra credit for speaking his mind.

They talked about their study habits and how to do well in the class. In doing so, one group came up with the idea of using our weekly quizzes as a “study guide” and answer/discuss questions online. (I don't give them the answers).

They also got to see and hear themselves in an online conference and learned some practical things about microphone adjustment, camera location when using a phone or tablet, microphone positioning, speaker feedback, etc.

The sessions were not mandatory, but I gave those who participated extra credit for convening or attending a session. Many students chose not to participate and I polled them, asking why. Schedule conflicts at the time sessions were convened was the most frequently cited reason.

An online class meeting

The second experiment was to conduct a class session using GHoA instead of in the classroom. (We met at the usual class time, so schedule conflicts were not an issue).

I begin each week with a presentation of misconceptions I saw in their homework assignments and quiz answers from the previous week and current events relevant to our class. Since the goal of the class is to introduce the "skills and concepts needed for success as a student and after graduation as a professional and a citizen," that is followed by presentations focused on a couple of concepts and on a skill, for example, how to use GHoA, an image editor, etc.

I followed the usual in-class format during the GHoA session. The first nine students who "came" to the GHoA session joined the live video conference and those who logged in later joined the viewing "audience."

This was the first time I had run a GHoA class, and it was a learning experience for me. As shown below, I made a number of technical errors. It also felt strange to be presenting material without seeing the audience -- it made me appreciate radio announcers. I suspect one could get used to it.

Mistakes due to my inexperience

I also made the mistake of not preparing the students well enough. They only had one presentation and one assignment with GHoA before we ran our experiments.

After the session, I polled the students on their experience during the live hangout and their use of the recorded video. Here are the poll results for three of the questions on the live hangout:

Selected responses regarding the live class session

And two of the questions about their use of the recorded video after the session:

Selected responses regarding the session recording

As you see, they said they were more comfortable viewing the session at home than in class, their minds were less likely to wander and they generally thought it was as good or better than the classroom as a learning experience. The majority went back and watched at least a portion of the session recording, but there was an inconsistency in their reporting.

The last four questions asked about their overall preference and solicited comments. When asked whether they preferred meeting in class or meeting in a GHoA, 53% preferred the GHoA, 13% the classroom and 33% were indifferent. When I asked them in class what they thought was the best way to offer the course next semester, the consensus was that the first few meetings should be in the classroom and about half of the remaining meetings should be online.

(There were 19 questions in the entire questionnaire and you can see the full poll results (including their comments) here).

This was my first try at using GHoA and I made several mistakes which could be corrected. If others have used GHoA as a collaborative teaching tool, please share your experience.

Friday, April 01, 2016

The Tesla Model 3 reminds me of the original Macintosh, but Elon Musk does not remind me of Steve Jobs.

The Mac and the Tesla Model 3 have a lot in common. For one, the Model 3 was not the first electric car or Tesla's first electric car and the Macintosh was not the first computer with a graphical user interface (GUI) or Apple's first GUI computer, but both came out at just the right time.

Workstations, the Xerox Star and the Apple Lisa all had GUIs before the Mac, but they were too expensive and remained niche products. When the Mac came out, technology had just improved to the point where a consumer computer with a GUI could gain a foothold and catch on. The Mac, with its proprietary hardware and software, was the only GUI game in town for several years until technology improved to the point that commodity hardware could support a GUI and Microsoft brought out Windows 3 and 3.1.

Similarly, electric cars, including Tesla's, preceded the Model 3, but they were too expensive and inconvenient to grow beyond a niche market. Battery, material and other technologies have now improved to the point where the $35,000 Tesla will appeal to the mainstream. One does not have to care about the environment or global warming to like it. It's sensors, safety features, comfort, size and increased battery capacity (coupled with more charging stations and home chargers) and the ability to be upgraded via software download will appeal to a wide market, including owners of gasoline-powered cars.

As it was for the Mac, the timing is right for the Model 3. it's not too soon and not too late, but just right, like Goldilocks.

As technology improved, personal computers with GUIs became ubiquitous. The transition away from gasoline will take longer than the transition from command lines to GUIs because of the longer replacement cycle for cars, but the tipping point has been reached.

There are also financial parallels between the two. Tesla bankrolled the Model 3 from sales of the Roadster and Models S and X and Apple bankrolled the Mac with sales of the Apple II. Both were fading when the new machines came out. Tesla's stock is now 60% above its price on February 12 and the Apple II was running out of gas when the Mac was delivered.

Both companies developed comprehensive proprietary designs. Apple built the hardware and software for the Mac and Tesla is making the car and the batteries.

But that is where the similarities end. Apple holds on to their hardware and software innovations, protecting them with patents and law suits. Not Tesla. On June 12th 2014 Tesla released all of their 249 patents, saying they would not sue anyone for using their technology in "good faith." As shown below, they took down the plaques on their "wall of patents" after releasing them, replacing them with an image and the slogan "OEMS all our patent are belong to you." (I think Yoda wrote that for them).

Tesla's "wall of patents" before and after (image source)

It seems that Elon Musk sees other car and battery manufacturers as collaborators in the effort to replace gasoline-powered cars rather than competitors.

Disclaimer -- I am kind of an Elon Musk fan boy and make my students watch these videos of Musk being interviewed by Sol Khan, announcing the formation of Tesla Energy and recruiting engineers for the SpaceX satellite Internet project.

Update 4/7/2016

In a blog post entitled The Week that Electric Vehicles Went Mainstream, Tesla says they received 325,000 reservations for the Model 3 in the first week. They also say that translates into about $14.5 billion in sales if all the reserved cars are purchased.

The base price of the car is $35,000, but these figures average out to nearly $45,000 per car. Elon Musk indicated that the base price included all of the sensors and software, but they clearly expect to sell relatively expensive accessories like a second motor for four-wheel drive and additional batteries for extended mileage per charge.

Update 10/20/2016

Elon Musk has announced that all Teslas shipped from now on will have the hardware needed to support fully autonomous driving, which the National Highway Traffic Safety Administration defines as:
The vehicle is designed to perform all safety-critical driving functions and monitor roadway conditions for an entire trip. Such a design anticipates that the driver will provide destination or navigation input, but is not expected to be available for control at any time during the trip. This includes both occupied and unoccupied vehicles.
The new Tesla processor has 40 times the computing power of the previous generation and runs neural net vision, sonar and radar processing software. The following video shows a test drive and the Tesla sensors.

Update 10/24/2016

The Tesla is also similar to the Macintosh in its attention to design and appearance. The new Teslas ship with eight cameras with 360-degree viewing at up to 820 feet of distance, 12 ultrasonic sensors that can detect both hard and soft objects and a forward-facing radar that helps see through rain, fog, and dust.

That is 21 devices, but you have to look carefully to see them -- they are small and unobtrusive, as shown here:

Friday, March 11, 2016

The Khan Academy -- on the Internet or a LAN near you

The Khan Academy began when hedge fund analyst Sal Khan started posting short, conversational videos on YouTube to help his cousin with her math class. The videos went viral. Today there are 23 courses in math, 7 in science, 4 in economics and finance, 25 in the arts and humanities, 3 in computing and preparation for 8 tests like the SAT along with content from 25 high-profile partners.

The Khan Academy is a non-profit organization that promises to provide a world-class education that is "free for everyone forever," and their open source software is available on GitHub. Over 39 million "learners" have used the material and it is being translated into 40 languages.

As shown below, the courses are comprised of fine-grained modules focused on a single concept and each module includes a test of mastery. The modules are arranged hierarchically, and a student has not completed the course until he or she has mastered a module -- they encourage experimentation and failure, but expect mastery. (Getting a C in a typical college course means the student understood only about half of the material and will do poorly in classes for which the course is a prerequisite -- an effect that compounds throughout college and into the workplace).

Portion of the beginning arithmetic course knowledge graph

In addition to the teaching content, the Khan Academy software presents a "dashboard" that enables a teacher, parent or other "coach" to monitor the progress of a student or class. The red bar shown in the dashboard view below indicates that a student is stuck on a given concept. The teacher can then help him or her or, better yet, have a student who has already mastered the concept tutor the one who is stuck. (Research shows that the tutor will benefit as well as the tutee -- "to teach is to learn twice").

Dashboard with fine-grained progress reports

The dashboard enables a coach to adapt to the strengths and weaknesses of each student and spot learning gaps. They understand that students may be blocked by one simple concept, and sprint ahead once it is mastered. (I recall sitting in freshman calculus class, and being totally lost for half the term, until I figured out what the teacher meant when he said "is a function of" and the class snapped into focus).

Confusion on a single concept "blocked" this student.

The third major component facilitates community discussion among the students taking a given class, allowing for questions, answers, comments and tips & thanks.

Tracking student participation in the course community

But, what if you don't have Internet access?

Learning Equality grew out of a project to port the Khan Academy software to a local area network at the University of California at San Diego. Their version, KA-Lite, can be customized for an individual learner, classroom or school running on a Linux, Mac or Windows PC as small as a $35 Raspberry Pi.

KA-Lite is three years old and has been used in 160 nations by over 2 million learners from above the Arctic Circle to the tip of Chile and translations are under way into 17 languages. The following shows organizations that are deploying it and installations.

There is an interactive version of this map online.
To learn more, visit their Web site and contribute to their Indiegogo campaign.

See this companion post on MIT's Open Courseware, which is also available off line.

For the history, pedagogical philosophy, accomplishments and future of the Khan Academy along with a video collage showing examples of their content, see this 20-minute talk by Sal Khan:

Real Time Analytics