The Future, This Week 22 June 2018

Disclaimer: We'd like to advise that the following program may contain real news, occasional philosophy and ideas that may offend some listeners.

Intro: This is The Future, This Week. On Sydney Business Insightd. I'm Sandra Peter. And I'm Kai Riemer. Every week we get together and look at the news of the week. We discuss technology, the future of business, the weird and the wonderful, and things that changed the world. Okay let's start! Let's start!

Sandra: Today on The Future, This Week: entertaining mergers, long haul economics, and the next social network wants your DNA. I am Sandra Peter and I'm the Director of Sydney Business Insights.

Kai: I'm Kai Riemer, Professor at the Business School and leader of the Digital Disruption Research Group. So Sandra what happened in the future this week?

Sandra: Our first story is actually one that's been all over the news and it has to deal with one of the biggest mergers in the last few years which is between AT&T and Time Warner. The particular story we're focusing on today is one from The Verge titled "The court's decision to let AT&T and Time Warner merge is ridiculously bad." So before we get into the implications of that merger let's just have a really quick look at what actually went on. First of all this merger brings together actually one of the world's largest telecom carriers, AT&T, and also one of the world's largest media organizations Time Warner. And just to remind our listeners a company like AT&T actually has about 150 million mobile subscribers and they've got TV subscribers, they offer broadband, they offer satellite TV subscriptions, while Time Warner, one of the largest media companies owns Warner Brothers and that's Warner Brothers pictures things like New Line Cinema and all the television groups that they are associated, DC Entertainment, a number of TV channels but also owns things like CNN or a Cartoon Network or all of the HBO channels. So this merger actually is a very very big deal.

Kai: Literally. And also in terms of its significance.

Sandra: Yes literally, AT&T is paying 85 billion dollars for Time Warner, and that's a lot of money!

Kai: By anyone's account. But it's also a big deal in terms of its significance because what we're seeing here is a merger between one of the world's leading content producers, a reminder HBO produces and owns TV shows such as Game of Thrones or Westworld, and they are merging now with AT&T, the very infrastructure companies that brings content to consumers, which raises a number of questions regarding competition obviously and this is why the US government actually took the merging companies to court trying to prevent the deal and the article that we are discussing unpacks in detail the judge's decision to allow the merger.

Sandra: So what we won't examine today is the judge's decision because what's happened this week is that AT&T and Time Warner have actually won this court battle and they have convinced the judge that they should be able to merge and this was against the objections of the Department of Justice who brought them up as you said and anti-trust legislation. And what's even more significant is that they will be allowed to do this without any restrictions on what they can pursue once the merger is finalized.

Kai: And what the author of the article does is he points to what he sees as a fundamental flaw in the judge's decision, a flaw that rests on a misunderstanding of the significance of having an Internet provider, an infrastructure company merging with a media company and content producer. So the judge compared this merger with the competition in this market, chiefly Netflix and the like. And he said well those companies are vertically integrated because Netflix not only is a content producer but they also own in the judge's words the delivery because they have their own app that they sell to and consumers directly through the Google Play store, the App Store, Apple TV and other platforms, so in fact the judge likened Netflix to a vertically integrated platform. And therefore Time Warner and AT&T should also be allowed to vertically integrate after all the judge says it's not a horizontal merger where two direct competitors would merge and therefore reduce options for customers and therefore competition, it is a vertical merge. But the author goes on to point out the fundamental flaw in this thinking.

Sandra: So what the judge fails to differentiate is between platform companies and Netflix is one of the examples that he has raised but he has also mentioned companies like Google or Facebook, and Internet providers like AT&T and whilst Netflix or Hulu and Google and Facebook do actually produce a lot of their own content, own their own apps through which that content can be delivered, they still require Internet providers like AT&T to actually be able to distribute that content to the end user, and internet app is not much use without the actual Internet to watch it on.

Kai: So vertical integration does not equal vertical integration in this case when one is merely the delivery via an app and the other owns the very infrastructure you need to actually get the app in front of consumers. So in effect AT&T owns the bottleneck that any one who delivers anything on the Internet has to draw on to reach the consumers and the significance here is that this decision comes a day after the so called net neutrality in the US has ended. Net neutrality prevented an Internet provider such as AT&T from prioritizing certain content over others, so they had to treat everything on the Internet the same. Without this law AT&T is free to monetize traffic to consumers and to for example now prioritize Time Warner HBO all content over Netflix and so the concern is that there will be a severe limitation of competition because AT&T could make it harder or slower in that case for Netflix consumers to consume that content, while prioritizing and featuring quite strongly HBO and other content they it now owns through Time Warner.

Sandra: And indeed it's really important to observe how the judge came to this conclusion and what the assumption was because much of this case has rested on the fact that the judge actually takes a backward looking view at the entertainment industry. The assumption here is that companies like Time Warner or like CNN would be reliant on making a deal with the television cable company to be able to distribute their content and companies like Netflix have an unfair advantage because they own an app. They don’t need a cable company to be able to deliver that to the consumer. And this would be a fair comment if the model going forward rested on most companies making deals with cable companies to distribute the content. But if we take a forward looking view, and if we see the way in which not only the new disruptors to the industries, so companies like Netflix or Hulu have changed the industry but the ways in which companies like Disney for instance have changed the way they distribute their content. Companies like Amazon have created their content. If we look forward we see that most companies now bypass cable companies and distribute their content directly to consumers, either through over the top services or through apps or simply through the Internet. In this world having the internet connection becomes critical and thus become a barrier.

Kai: And that points exactly to the heart of the matter, in that mobile companies telcos such as AT&T have become the new infrastructure for delivering content to consumers. They in fact have become the new cable companies.

Sandra: So the Internet service providers today, being like cable companies actually can create an unfair advantage in the market. So let's summarize what else matters for the future of this deal. First as you've well pointed out with the abolishment of net neutrality, AT&T actually has a loophole for favoring its own content. And that loophole is actually zero rating. That means that they don't even have to slow down other services but just make the consumption of a specific content not metered on their internet connection. So whilst you might be using up your data to watch Netflix, AT&T would allow you to watch HBO without using up any of your data. And this is not the same as net neutrality because I'm not slowing down.

Kai: On top of that they can also bundle right? So when you're an AT&T subscriber and you have got your mobile services to them, they might actually package some of their content and give you Westworld for free for example, whereas Verizon customers, as in their biggest competitor might not get such a deal. So one question I want to ask is where does it leave Netflix? So maybe there's an opportunity you know for Verizon and Netflix to cooperate given that AT&T and HBO are so tightly bundled, which might look good for these companies but might actually be bad for their customers, because now we're in a situation where the choice of mobile connection where you get your Internet infrastructure from already is a decision about which content you get to watch at which cost, which means we have now effectively gotten rid of the idea of a free infrastructure for all that is agnostic to the kind of content that you can consume which is a pretty big change from the kind of Internet that we've become accustomed to.

Sandra: It could also potentially be massively destructive to companies that are content creators, whose entire business model rely on being able to monetize that content across multiple platforms and with very few restrictions, making their content exclusively available on one platform or another would actually destroy a lot of the value that these companies have.

Kai: On top of that, I think recent experience here in Australia also shows that telcos exclusively owning and delivering content is not always guarantor for a great experience and I am of course referring to the giant Optus World Cup debacle where here in Australia the second largest mobile telco Optus bought exclusively the right to screen the entire soccer World Cup on any medium that is, so free TV would only show a few games and all of the games would only be available through streaming on the Optus app. Obviously with a big advantage for existing Optus customers but anyone could buy a monthly 15 dollar subscription to then watch every World Cup game through the Optus app. This and in a big disaster Optus was not able to deliver a stable streaming experience. The content became unwatchable and in the end in what is a big fail, Optus had to surrender all of the pool matches in the first instance now to the free to air network. So a big stuff up, which shows that there's different skills involved in running a mobile network to actually delivering a great app experience.

Sandra: So now that you've gotten that off your chest Kai...

Kai: ...I did. I did. Actually kudos to Apple who overnight after me complaining refunded the 15 dollars for the Optus app, so...

Sandra: So some common sense still prevails in this whole debacle. But let me go back for one second to our big merger, because I think where that leaves us is actually in a very important space where a lot of the recent anti-trust cases that increasingly involve technology companies are still being debated in terms of the anti-competitive elements they would have at a different point in history where the dynamics of those industries were quite different. It is quite a possibility that if the government had chosen to run its case differently, more forward looking, rather than backward looking it would have had a much better foothold to at least impose certain conditions or certain prohibitions on the behaviour of the newly merged AT&T and Time Warner company.

Kai: So a timely reminder how important it is to engage with the future in a forward looking manner and not just as an extension of the past. Well a point we've been making on this podcast time and time again.

Sandra: Well actually the point of this podcast... The last issue to highlight this is that a lot of the legislation we have surrounding these issues is quite out of date. And even if we look at this case, some of the legislation it was debated on is actually not that old, it goes back to the Nineties but technologies changing the dynamics of this market so quickly that actually even legislation from 10 years ago prevents us from actually debating these cases on their merits today.

Kai: And while this might seem like a case that mostly concerns US consumers, the bigger takeaway here is the treatment of Internet infrastructure and access as a product that can be owned by a private company. So I think this ideology to treat infrastructure as just another product that can be bundled with content providers, for example, is extremely dangerous to the public ownership of the Internet as a free space where everyone can get access to any content regardless of their economic stance or the kind of internet provider that they sign up with. I think this is a fundamental principle that is under threat at the moment and that deserves attention in Australia where companies such as Optus for example were able to own the World Cup rights much like in the U.S. with AT&T and Time Warner.

Sandra: And now to something completely different.

Kai: Our next story is from Wired UK. It's titled "Airlines are on the brink of the ultimate ultra-long haul flight"

Sandra: And Australia is right in the middle of this one because it concerns the fact that the flight from Sydney to London is almost within reach. But the economics of such long haul flights make for some very interesting implications. So let's have a look at that.

Kai: So the first flight on what is affectionately known as the Kangaroo Route happened in December in 1947 when Qantas flew for the first time from Sydney to London carrying just 29 passengers and 11 crew. So a lot of attention for each of those passengers, which stops in Darwin, Singapore, Calcutta, Karachi, Cairo and Tripoli. Staying overnight twice in Singapore and Tripoli and taking 93 hours in total.

Sandra: Not only that but it was a bloody expensive flight.

Kai: Yeah it cost the equivalent of 130 weeks’ average pay, and we can see now where the name Kangaroo Route comes from, because it was basically a series of hops from Sydney to London. And these hops are what airlines are now trying to eradicate, so there is only one stop or two hops involved these days with typical stops in Singapore or Dubai.

Sandra: But now the 21-hour flight which is what it would take from Sydney to London, is pretty much within reach and it will be Qantas that's issued the challenge last year to Boeing and Airbus to actually try to make this happen. And right now it seemingly down to the economics rather than the engineering.

Kai: So Airbus has introduced its new A3 350 900 ULR ultra long range, currently operated by Singapore Airlines on its longest route between Singapore and New York. So this plane will be capable of flying non-stop between London and Sydney. And Boeing is about to introduce its version of the triple 7 plane, also capable of doing the distance. But the point that the article makes is that the economics only stack up under certain conditions and on very few such long routes in the world.

Sandra: And probably the Singapore to New York route is actually much better suited to this flight in terms of the economics than the other two longest routes at the moment, one which is Qatar Airways from Auckland to Doha and the other one is our very own Qantas Perth to London route both around 17 hours’ non-stop flights. The problem is with who pays to fly these routes?

Kai: And how do you actually make the tradeoff between packing more fuel for the longer distance at the expense of less passengers’ work? And the solution to this seems to be to go exclusively to the premium segment. For example, on the Singapore flight to New York there is no traditional economy class. It's 90 for premium economy and 67 business class seats because the ratio between weight and the air fare doesn't actually stack up on these planes in a traditional economy configuration.

Sandra: And those numbers are actually much worse if you look at room for instance from New Zealand to London, where only about 10 - 12 percent of the passengers pay premium fares. So even with the technology working the economics just do not allow for this ultra-long haul flights.

Kai: Which raises the question whether this has the potential to become a widespread phenomenon because there's only very few long routes in the world where a significant number of passengers would be prepared to pay premium prices to actually make these flights worthwhile.

Sandra: And when we look to the future of the ultra-long haul flights there is actually another example from history that we can point to where the technology was there. The economics of it just didn't makes sense.

Sandra: So you're obviously referring to the Concordes, the first and also only regular supersonic flights in the history of humankind that happened during the 1990s into the early 2000s. Which is actually another area of innovation in the aviation industry. So while we have solved the long haul challenge, from an engineering point of view and while there's a few routes where this might make economic sense but not in many instances supersonic is kind of a similar problem economics wise. So when we look at the Concorde back in the day, and we all remember the Paris disaster, when a Concorde went up in flames and everyone onboard died, the eventual death of the idea of the Concorde supersonic flights didn't come on the back of engineering challenges. It was essentially again the economics that didn't stack up, because in order to make this work again it had to be a very high priced, high end market product which due to the very nature of the plane didn't really work in this segment because while it was fast to go from London to New York in three and a half hours, the experience on board was everything but comfortable and luxurious. And that has largely to do with engineering challenges because in order to do supersonic flight you have to break through the sound barrier and you can only do this with a plane that has a certain shape, has a pointy nose, and a very narrow fuselage which means there's not much space on board to provide the kind of luxury that passengers have become accustomed to in business or first class.

Sandra: And interestingly it has taken us until now to find a new profitable way of doing supersonic flights because they are back but they are only back in the space where the economics do make sense and that is really small planes.

Kai: Yeah. So much like in long haul travel where slow and steady innovations in materials and engines and fuel efficiency have led to solving that challenge. There's a number of promising US based startups that are on the verge of bringing back supersonic travel, solving engineering challenges in the process, but with much much smaller planes than the Concorde used to be. I want to highlight just one company here a US based startup, aptly named Boom that is working on a supersonic concept. Boom is promising to launch a supersonic plane that looks a fair bit like the old Concorde but is small and only has 55 seats in an all premium configuration and is promising to bring back the flight on a New York to London route in the three hour vicinity. And this company actually has already taken 76 orders for this plane that hasn't been built yet. They're about to launch a prototype and orders from companies such as Richard Branson's Virgin. But there's still a lot of challenges in this space to be overcome. Supersonic flight uses a lot of fuel. Again there's the ratio between how many passengers can I pack? How much fuel do I need? But supersonic flight is also illegal, for example, over the mainland US because of the problems with the supersonic boom when the plane breaks through the sound barrier the noise it makes can be so severe that in the past it has shattered windows in houses, which has led to the US government actually outlaw these flights, so are West Coast flights to London for example is not on the cards with supersonic speed at this point in time.

Sandra: Which is why I guess we're seeing most of the countries that have a lot of airports that are close to the sea looking into this, so we've seen Japan ordering about twenty of them because their routes on which this will be able to fly will be very particular routes.

Kai: So they are all routes that cross various oceans, the Atlantic or the Pacific. So let's round off this story by looking at where innovation in aviation is going. The long haul challenge seems to be solved. We're making progress in supersonic travel with companies such as Boom and other companies that are targeting exclusively business traveller with even smaller planes. Two other areas come to mind: one is alternative fuels and we're not talking concepts of solar or battery powered planes which are far far away from commercial uses but alternative fuel in the sense of bio energy that is mixed into traditional fuel, much like we've seen in cars and a third area of innovation happens inside the plane. Various initiatives of smaller scale innovations regarding smartphone apps for controlling your seating environment, better solutions for disinfecting the lavatories. These kind of things that make the travel inside the plane just a bit more pleasant or safer.

Sandra: So quite a few interesting things there that we might come back to in future podcasts as we see this interplay in the innovations space. Really alternating between the very technical challenges of building these planes and the economics of it. Quite often we tend to think about innovation in the aerospace industry as innovation that is really driven only by engineering and by new materials or new types of planes but we see that being very much kept in check by the economics of this industry which sometimes just do not match with what we can technologically achieve.

Kai: And incidentally that happens to be the same dynamic in space exploration, space travel and making the economics of the space race work where we've discussed in a podcast a few weeks ago, that while in terms of the engineering we've long solved taking people into space, it is now the economics that stand in the way of finding a solution to do this on a sustainable, on a regular basis.

Sandra: And incidentally also what we're seeing play out with the Hyperloop where again we've got the technology now meeting up with the economics of those solutions. So we will be coming back to this for sure. But now for something again completely different...

Kai: "Are genetic testing sites the new social networks?" asks the New York Times.

Sandra: So here we are talking about companies like 23andMe and Ancestry.com which we've mentioned before on the podcasts. So they are companies where you get a kit at home, they get the sample of your saliva, you spit in a tube, you send it back and it will give you insights based on your DNA as to who are the people you might be related to whether that's brothers or sisters you never knew you had, if they've contributed to that DNA database or whether your are more similar to let's say people from Ireland than people from Scandinavia or you might have some Middle Eastern heritage that you never knew had or you might be related to a certain tribe of Native Americans that you never knew you were.

Kai: Yeah. So they give you three different pieces of information. One is your genetic makeup which might predispose you to certain illnesses and diseases. That's an aspect we don't want to discuss today. And the second one is the people you are directly related to who have also done the test and you might not have been aware that you had a third cousin somewhere upstate. And the third one - where does your heritage originate from, in more general terms, so you know 27 percent Scandinavian, 19.2 percent broadly southern European. You name it.

Sandra: So this article from The New York Times caught our attention because it asks the provocative question of whether these might be the new social networks of our time. Because the effect of the results you get from one of these sites is actually quite similar to what used to happen at the very beginnings of companies like Facebook. When you first joined the Facebook community, you found friends you had been out of touch, or family members that you never knew you had, and it gave rise to these dynamics of people seeking each other out, people getting connected. And we are seeing a very similar things happen now on these DNA matching sites where people are starting to form relationships with people that they never knew they were related to or they're starting to travel to places they never knew they belonged to, cultures that they never knew they had a connection to and trying to use that to build more connections.

Kai: So this is one of their major selling points at this point in time, that you don't just get an idea of where your heritage lies but that you can tick a box which says I am happy for my data to be used to connect me to other people also taking the test. And the article reports on a number of interesting cases where people have found long lost relatives or be connected to cousins they never knew they had, meeting these people connecting with these people and building new relationships based on their DNA, in the process very much acting like an early day Facebook which was all about building communities, creating relationships, meeting new people.

Sandra: And the similarities actually don't end there. So we spent a bit of time thinking about this, and actually the similarities go beyond what the article mentions. For better and for worse. First it depends a lot who else is on that specific social network. So much like the actual social networks whether it's Facebook or whether it was MySpace or whether it's Instagram, who you are connected to depends a lot on who else is on the platform. Because remember these DNA tests are comparatively.

Kai: So when you send in your DNA, that string of letters that represents your DNA doesn't tell you whether your Swiss or German or native Australian, that classification comes from the comparison of your DNA with other people's DNA who have previously done the test and disclosed certain information about themselves. So when you sign up you're not only providing your saliva, you're also providing a whole lot of information about yourself and your known family history, which then goes into the data pool and is used to match you up with other people, which means that it works best for people of ethnic groups, for example, where many people before them have already taken the test. So if you are a mid-30s European white person, the results are very good and very detailed but it has been shown that for people of Afro-American descent or indeed Native American descent, there is very little data in the databases of these companies. And so the results can be rather disappointing.

Sandra: So again this data about your genes will be determined by comparing them to the other people in that database. So the tests are not inaccurate because the science is inaccurate.

Kai: No indeed it very much depends on who has taken the test on a particular platform. And interestingly there's been a few people who have sent their samples to various platforms and then compared results. And it turns out they can deviate widely.

Sandra: Which brings us to another similarity with social networks, which means these things get better the more people are on there. So this year, 23andMe actually announced an update to its service that would give you even more insights. So it might not tell you that you're a Scandinavian but it might tell you that you’re in particular Norwegian. So it has been upgraded now to a hundred and twenty different categories up from thirty-one that it used to have previously, simply because of the number of people who have taken 23andMe's test, hence the reference population that the things that they can match your DNA to has improved wildly and the algorithm can now be much more accurate in determining where in the world it has found the match to your background.

Kai: And finally, and probably most scary of all, there might be another similarity with social networking companies such as Facebook and that is the fact that these companies such as Ancestry.com and 23andMe collect a lot of very personal information about their users. The most personal of information which is their genetic make-up, while providing a feel good community and socially oriented frontline service, the data pool that these companies are amassing offers them all kinds of possibilities for future monetization and you can imagine that a DNA pool that is well annotated, that comes with a lot of information that users disclose might be of incredible value for health organizations, insurance companies, commercial advertisers, research companies, biotech startups who want to target new products at people with different genetic make-up...

Sandra: ...Pharmaceutical companies, insurers, the list is endless.

Kai: Absolutely. And it remains to be seen where these companies are going with monetizing their business models. But if valuations of these companies are an indication then investors are clearly banking on the fact that these companies have more potential than just connecting people to each other. And now that we have made you feel bad. You're welcome.

Sandra: So we should end this quite quickly because we're running out of time for future bites and there has been a lot this week. So quick fire around, what's one of your future bites for the week?

Kai: Okay so IBM is building yet another AI. This one it's called Project Debater, an artificial intelligence that is designed to make quote unquote coherent arguments and win debates in debating humans.

Sandra: This has the whiff of bullshit about it. How is it building an argument if it can't understand what it's arguing about? Or it doesn't really want to argue for or against something?

Kai: Absolutely and that's the point the M.I.T. Technology Review article makes. The title is "This AI program could beat you in an argument but it doesn't know what it's saying." Which raises the point why build an AI that can debate? My point is debates are useful where we want to explore a controversial topic, where we want to imagine a future, where we really have a debate in order to make sense of an emerging topic, but this is essentially the kind of areas where AI doesn't really work very well. Deep learning works best in areas that are well understood, where there is a lot of training data to actually train the algorithm, so you can have debates in areas that we already understand and then you can possibly win a debate against a human. But what's the point because those machines will not actually contribute any new understanding because they don't understand anything. So your assessment is right, they're probably just bullshit machines. They are nice demonstration about what you can do in language pattern matching and they bring us forward a little bit in building expert systems or chat bots. But by and large I find this a rather pointless exercise.

Sandra: And it's good you bring up chat bots because that takes me to my quick bite for the week, which comes from The Financial Times. And The Financial Times had a really interesting opinion piece that was called "The digital future is female, but not in a good way"and it was actually a story that seemingly just rehashed an old conversation that we had quite often where Alexa and Siri and Google are all the digital assistants actually have female voices. But this article took that conversation a bit further and it was a very interesting argument. They said well interestingly it's not only those digital assistants that are female, but also a number of car companies, and they singled out by Daimler here, a number of banks said that they were working on digital female assistants on all of these chat bots having a female presence and a female voice, but that this was actually based on years of data that showed that having a female voice for customer support made the interactions more efficient, made client's happier and yielded better results.

Kai: So you're saying we have a bit of a dilemma. While on the one hand we reject the notion that these artificial agents should all be female, the research shows that people actually respond better to female voices.

Sandra: And perceive them as more helpful and more collaborative. So the article was actually quite interesting because it came on the back of London Tech Week, which is a festival all around technology which consistently decries the lack of more women in technology. It comes during the same week that Google's numbers are out and they showed that it's still overwhelmingly male and white and that diversity hasn't improved at all, even though over the past year and a half we had this conversation about diversity and about more women in tech, really at the forefront of the agenda of many of these companies including Google. So whilst we hear that there are not women in tech, it seems that they are over-represented in the digital assistant, chat bot, fem bot space. Which leaves us with an interesting conundrum about how to deal with the lack of diversity there.

Kai: We leave you with this food for thought and that's all we have time for.

Sandra: Thanks for listening.

Kai: Thanks for listening.

Outro: This was The Future, This Week, made possible by the Sydney Business Insights Team and members of the Digital Disruption Research Group. And every week right here with us our sound editor Megan Wedge who makes us sound good and keeps us honest. Our theme music was composed and played live from a set of garden horses by Linsey Pollak. You can subscribe to this podcast on iTunes, Stitcher, Spotify, SoundCloud or wherever you get your podcasts. You can follow us online, on Flipboard, Twitter or sbi.sydney.edu.au. If you have any news that you want us to discuss please send them to sbi@sydney.edu.au.