Future Tech 2024: An Interview with John Lindsay (Part 1)
Broadband in 2024, Shifting Usage Patterns, Cloud Services, Privacy and Data Harvesting
Welcome to Market Clarity’s Future Tech series where we will be speaking with luminaries in research labs, vendor labs, telcos, futurists and others, about cutting edge research insights into the wide range of gadgets, technologies, and trends that will impact telecom services (and everyone that uses these services), including: wearable devices, sensors, new form factors for phones + tablets, drones, the Internet of Things, security and privacy, cloud services, broadband access technologies, the future of BYOD and much more.
In our inaugural interview we speak with John Lindsay — one of the Australian telecom industry’s most prominent technologists.
John started his first ISP in the early 90s, was a founder of iiNet South Australia, general manager of Chariot Internet, Network Operations Manager for iiNet Ltd, Carrier and Regulatory Manager and later CTO for Internode before being appointed Chief Technology Officer for iiNet Ltd after the sale of Internode to iiNet in 2011. John is now helping early and growth stage technology and Internet businesses survive, grow and thrive in his new company Lindsay Strategic Advisory.
Shara Evans (SE): Welcome John, and thank you for taking part in the Future Tech series. Before we talk about the future, please take a moment to briefly describe your new venture.
John Lindsay (JL): Thanks, Shara. I’ve hung up my shingle as a consultant (Lindsay Strategic Advisory). I’m mostly helping early-stage businesses and growth businesses with their plans, their networks, and with funding. I have taken on Simon Hackett as my first major client, which is ironic, given that I’ve spent so much of the last 14 years working for Simon. It’s actually been really quite exciting. He’s got a lot of business interests outside of the Internet area, and it’s been fun to look at businesses that aren’t about providing broadband access to people.
SE: That would be a lot of fun. Simon is such an innovator.
I’d now like to turn to the topic of what the telecom services world is going to look like in 10 years, in the year 2024. You’ve been in this business for many decades and working with one of the larger service providers that grew up from a little service provider into one of the four largest in the country. What are your views on the kinds of broadband access technologies that we’re going to see both in Australia and around the world in ten years’ time?
Broadband in 2024
JL: I guess we should really bear in mind that 10 years is half the life of the commercial Internet in Australia. We’ve had pretty much 20 years of commercial Internet.
I set up my first ISP, Byron, on the Australian Public Access Network Association (APANA) in a spare room of my suburban house around about 1994. That morphed into the business that was known as iiNet SA, where I had as business partners Leigh Hart in Adelaide, Michael O’Reilly and Michael Malone from iiNet. We created a dial-up ISP pretty much from scratch, using the iiNet branding. We created that in South Australia. We built the business up, and eventually we sold it because there were some people from Sydney who looked like they might be the future — consolidating Internet service providers. Even back in the ‘90s, the idea of growth from acquisition, and consolidation of the industry was there right from the very beginning.
Over the last 20 years, we have seen the bandwidth available to consumers rise from 14 Kbps modems, 28, 33.6, the 56 Kbps modems that were driven by ISDN-based access servers. There was a brief time where 128 Kbps ISDN was the leading edge for home users. I had that in my home in the very late ‘90s. Then ASDL came along in 2000/2001, along with wholesale access from Telstra in 2001.
That available bandwidth has kept growing. Now we have NBN fibre to the home. We have NBN wireless. We have 3G and 4G mobile, various other wireless technologies — accommodating a 2.5% per-user per-month growth in traffic. You can find graphs based on actual customer data inside companies like iiNet that show you this, plotted back over 20 years. When Michael Malone delivered the Charles Todd Oration a few months back, there was a graph in his slide pack that we put together based on actual data that we gleaned from company records over 20 years. (Although that the slide didn’t make it into the final edit of the presentation.)
I can’t see a compelling reason why that growth will stop, which means that the typical user will be consuming 20 times as much data in a decade as they do today. It’s not too hard to imagine things that could drive that, not least of which is that we are in an on-demand society. We expect to be able to find a hamburger at three o’clock in the morning, even if in Sydney we can’t find a drink at three o’clock in the morning.
The days of sitting in front of a television looking for entertainment that’s being shared with all the other viewers as it goes free to air or delivered by the monopoly Pay TV provider — they are really now becoming the past, that the future is about video on-demand and that video is high-definition. That means that it’s consuming several gigabytes an hour.
I appreciate the argument that says, “Why are we spending billions of dollars building a national broadband network just to distribute TV, distribute video,” but at the end of the day we’re building a national broadband network to let people use it to communicate however they want to. It’s really not for us to judge what people are using it for today, and nor is it really for us to have to dream up plausible possible future uses. If you provide the connectivity people will find interesting things to do.
SE: So in 10 years, do you think a typical home would have a 100-meg connection, a gigabit connection, something else? What do you think will actually be required to provide the kind of service that gives you good-quality data services for the things we can imagine today? Home automation, and the Internet and Things, and all these other devices that will be connecting with things or services around the world — what kind of access feed do you think a home will really need to do that?
JL: The average broadband household using fixed-line service, on average, is using half a megabit on DSL services, averaged across 24 hours. On NBN, I know from my background that that number is more like one megabit. I don’t know what NBN Co said today during their first ever results presentation, but if they fessed up how much CVC has been sold and how much is being used, you could probably infer something that looks like a megabit (Mbps) or two—certainly well above the 75 Kbps, which was in the original design of the NBN. If they had actually paid attention back in 2007 to this 2.5% per-month compound growth and had actually built that into the model, then they would’ve been in a slightly better place over the last couple of years in terms of how that works. If the typical broadband user is using about an average of 1 Mbps today, in 10 years’ time, they’ll probably be using an average of about 20 Mbps.
SE: That low? Even with all kinds of new devices that are not even imagined today?
JL: We have the issue of peak versus average. There are an awful lot of people in Australia who have one-and-a-half Mbps ADSL today and who think that’s actually quite reasonable. You know, that’s a three-to-one ratio of the peak to the average. The average ADSL2+ user has about a 12 Mbps line sync, so they’re at 24 times average, and that is in fact probably quite a reasonable amount of bandwidth.
I think that in 10 years, there will be a lot of people using 50- and 100 Mbps connections. There’ll also be a lot of people using 25 Mbps or even 12 Mbps connections.
JL: Today, there are tens of thousands of people in Australia still using dial-up modems for e-mail.
SE: I wonder how many of those who are using dial-up do so because there’s no other choice where they live, as opposed to using dial-up because they don’t want to migrate to DSL.
JL: Well, they’ve had ample opportunity in the last decade to acquire satellite-based services. There’ve been all sorts of government subsidies that have meant that people who wanted such a service could get it for a metro equivalent to DSL price.
I look at people like my mother who sends and receives a handful of e-mails off her iPad, and she uses her Internet connection for cheap Voice over IP phone calls to Europe and to keep her TiVo on, fed with its electronic program guide. In 10 years from now, I expect that her broadband consumption habits will be broadly similar. She occasionally uses iView on her iPad to watch an ABC show if she misses it and didn’t manage to record it on her TiVo.
Shifting Usage Patterns: Cloud Services
SE: So there’ll be some people who just don’t take advantage of the technology. But if we look at the majority of people, do you see that usage patterns will shift from largely being download-driven to symmetrical, where people are, say, using cloud services and sending information upstream just as much as they are getting information like watching videos?
JL: I think that it’s desirable to enable that because one of the obvious uses of a broadband service today is to create a reliable offsite backup of your data. When your life is in the cloud, when every photograph that you’ve ever taken is in the cloud… You know, my eldest child is now a teenager. Every photo since she was three years old that has been taken of her is digital. Every photo that she has ever taken in her entire life is digital, and she has managed to amass tens of gigabytes of photos that she has taken or that she is in, in her short life. I think that being able to keep all of that as a life record online and backed up is one of the coolest things that you can do with broadband access.
Likewise, what the future looks like in terms of where data is, how you get to your own data, where the server is that’s processing the data, how it’s being displayed to you — packets just want to flow. It’s really a historic accident that ADSL is asymmetric, that it satisfied some user requirement and some technical requirements from the 1990s when it was invented. The demand is shifting.
Privacy and Data Harvesting
SE: Well, it’s interesting, John, how you talk about the shifting demand caused by cloud services. Certainly, that’s going to, at least in my view, change the paradigm of traffic flow, but it raises a whole range of other questions about cloud services, in particular because of the recent revelations by Edward Snowden about NSA snooping on information that’s stored in the cloud, on conversations and many other things. Do you see these kinds of revelations and the increasing use of cloud services changing in the future? What do you see in terms of security and privacy implications for putting so much of our life history up in the cloud, and relying on it always being available to us in the cloud, but also being private unless we choose to share it with someone?
JL: I think that the biggest issue with data and communication today is encryption. It’s interesting to note that services like Google — for instance, the Google search — now flips into an HTTPS connection as soon as it can. So if you browsed to google.com, it will flip into an HTTPS connection immediately. The reason it’s doing that is so that anybody in the data path between you the user and Google servers can’t actually see your query, your search, and the results. You might assume that this was due to some sort of paranoia, or you could perhaps more accurately realise that it’s not paranoia. Just because you’re paranoid doesn’t mean that everybody is out to get you or not out to get you.
There are numerous organisations now who are making it their business literally to spy on users, to observe users, to observe what they do online, where they go, what they search for, what the results are that come back from their searches even, and build that up into a huge database, and then that data is sold.There are so many small interactions that people have online, and they think of that interaction as being inconsequential, but when lots of websites are actually selling the data, what they sort of think of as the waste data, selling it to these big data warehousing companies, an overall pattern emerges from what the website operators have been convinced is anonymised. You don’t need a username because all we need is the IP address. Well, with anonymising — there’s an IP address, which is up there with saying, “we’ll anonymise it with your driver’s license number.”
SE: Yes, or your mobile phone number.
JL: Yes, yes, even better. So increasingly canny operators of services on the Internet are encrypting the data. Now you would think, for instance, that Microsoft, with their Watson tool that gathers up information about processes that have crashed on Windows machines — you would think that they would encrypt that data from your server or your desktop computer on its way to Microsoft servers, but in fact security researchers have discovered that they don’t and that monitoring that data stream provides an incredibly useful source of data about threats and exploits on the Internet.
JL: Yes, whatever it happens to be that’s in memory at the time when the application crashes. There are so many examples right now of applications — particularly on mobile platforms — where the connection between that mobile client and the cloud service is not encrypted.
More worrying is when you sit in an airline lounge and you connect to the free WiFi servers, if you then go and collect your email, chances are, unless you’ve taken the steps to configure your mail to use SSL, that you just broadcast your email username and password in clear text over WiFi. And chances are that the quiet person in the corner with a modest-looking thin Ultrabook is just snorting, trawling with AirSnort, snorting up all of this data where it’s being carefully indexed and filed away for later reference.
But, there are already apps such as Threema that support encrypted mobile messaging.
We’ll be publishing Part 2 of the Interview with John Lindsay on Thursday 20 March – where we will discuss encryption, the end of privacy, sensors, Australian Internet services in 2024 and the role of dark fibre.