Welcome to the first of the Thought Experiments Essays. In this essay I think about what would happen to a business like mine in light of persistent internet failure, or even destruction. The thought experiment brings to the surface a whole range of interrelated concepts that are not obvious at first. Join me, and let’s see where it goes.
What would happen to an internet-reliant services business in light of internet failure?
This thought experiment begins in a fictional scenario, when my internet became intermittent and slow. It is not something that is unusual, here in Australia. For all that we pay huge fees for being able to use something that is so important, it’s still really bad. Part of the reason is that the country is actually fucking gigantic. Rolling out any kind of communications in Australia is about as different a project from doing it in any European country, as planting a tree is from sitting around watching a horror movie in the dark.
The Australian continent is huge. And what people started to realise is that while the copper cabling was pretty good for telephones, that required huge amounts of maintenance too. Even if you were 30 km from a country town, chances are that your phone connection was dodgy. Every bend, every turn, every damp low point, every rocky high point, causes communication along a cable to slow down. So, when it was proposed that there was some incredibly amazing super fibre being rolled out, those of us with even a remote understanding of the complexity rolled our eyes and laughed at the proposed timeframes.
Balancing the realism of this kind of infrastructure with public expectations of instantaneous access to incredibly fast internet must be super difficult. Hats off to the public relations and communications guys who help teams of politicians manage that expectation. It’s an enormous challenge. Even creating Australia’s skynet of mobile towers is hugely challenging: They have to be certain distances apart; and even then, those in cities assume that internet access – even mobile access – is a given in Australia. Haha! It’s not. Beyond metropolitan and regional centre boundaries, which, let’s be honest, the majority of Australians never venture into, let alone live in any more, it’s a really different story.
In any case, the internet was slow. And then it was patchy. And then it failed. Maybe the router had problems, or my other half had been downloading something particularly gigantic. Or perhaps the internet service provider had problems.
After eliminating the possibility that it was my actual computer (nothing would surprise me; in the past three years I have owned and somehow had fail on me at least four computers, a replacement rate of about one every six to 10 months), I decided that I’d been working for too long anyway and needed a break.
So instead I went and put the kettle on. Leaning over the kitchen bench, as I am wont to do, I grabbed my iPad and went to have a look at Twitter. Oh, right, no internet. Where’s my phone? I grab that instead, switch on the mobile data and go and have a look.
Except, that didn’t work either.
Now I was curious. I sent a text message to my other half (which worked, and I was kinda surprised) telling him that all my internets were down, and were his working?
I had to wait a little while to hear back from him, because he had to go and “have a shit” to check it, working in a factory as he does, but nope his wasn’t working either. Maybe there was a solar flare and the human race is finally at its end, he suggested happily.
Figuring that it couldn’t be down everywhere, I sent a few more messages to a few friends who work in other states, and my sister, and got the same response.
And because there was no internet, nobody could even give me any gossip or rumours from social networks that I’m not in but they are. Usually Twitter would explode at the very idea; rumours would be adrift on Facebook; photographs on Instagram; videos on Snapchat.
They were all silent.
I smiled. That’s kind of awesome.
And then I realised that I had been halfway through a job for someone in the back-end of their website, which I hadn’t mindfully saved for the last half hour. And which, given the internet is now not available, would probably all be lost.
Arrrrrrrrrrr! I exclaimed out loud, and reminded myself for the fifty bajillionth time to save, save, save, save, save the damned work every few minutes.
Given the internet is down, what could I do now?
You could clear out those jobs in your inbox, my mind suggested to me. No internet, I reminded it.
What about…? No, they’re in Google Drive and you can’t access that online, it said. Or you could… or…
I sat down and thought about this. I couldn’t access any documents and continue the jobs I had going, because Google Drive requires the internet, and I didn’t have access to that if the internet is gone away for a holiday.
I couldn’t access any emails or jobs that I’d saved, because that was in Gmail. Same deal.
Or my CRM, which is online and cloud based.
Or my team communications, because same thing.
All my contractors were overseas, and we only communicate via cloud systems or email.
I had already confirmed that my local referring partners were offline too, by SMS. Maybe I should contact all of my clients? Not yet. Wait to see how problematic it is.
I could tweak my sales funnels…. Nooo they’re online. Can’t do that.
I could do all those financial reviews I’d planned for next week, and do my data check-in and see where things are. Goddamn it, they’re all online. Cloud-based accounting, all my analytics, all my reporting capabilities.
Every goddamned one was somewhere else, and I couldn’t access anything up to date while I was offline.
Then the reality of this situation started to unravel itself to me. I couldn’t do any kind of manual financial reconciliation, because all of my bank statements are online; and even though banks have apps, they were all down because the require the internet to function.
I started to make a list of the things that I could do. I could sketch out some things by hand. I could write two or three blogs for my marketing efforts. I could do a bunch of other things.
So for the rest of the day, that’s what I did. I did everything offline that I could, and the result was that I had this ship ready to roll for whenever things sorted themselves out. I had calendars established, plans done, articles written, planned presentations done, slide decks done. I had mapped everything that I had left for that day-that-never-comes-because-you-don’t-schedule-it.
And at the end of all of this, I figured I should probably give my office a clean.
My phone was silent, because my phone line is a Voice Over Internet Phone connection, which means that when there is no internet, there are no calls. Skype, too, relies on the internet to function, so any re-routing wasn’t happening either.
By the end of the day I was so organised that I felt like a new woman.
But what if that happens and you don’t get internet again for a week? Or two? Or longer?
When you have no internet for a day, maybe three days, that’s fine – you can function. But the reality of a full two weeks is damaging. Here’s a quick run-down of what I would be unable to do as a result of there being no internet.
The first is that I can’t check my bank balances: Not online, not by phone, not in person. The reason is because when everyone’s internet is down, that means that the bank’s is, too. All of their records are stored in connected databases, and so if they have no access neither do I. Their apps don’t work, I can’t put money in and I can’t get money out. They have been closed since last Friday and the bank manager told me in confidence that she doesn’t know when they will be able to reopen. Their entire operation relies on internet-based technology.
This means, too, that I have no access to any physical cash. The almighty ATM doesn’t work when there is no connectivity.
Without physical cash, I can’t buy anything. The stores that run on cash can still function. But their ordering is all online. So for those guys who do everything from takeaway to basic necessities are all pretty well fucked now. Even the el-cheapo weekend markets are filled with vendors whose supplies come from major suppliers. Only the growers and producers are able to go to the markets… but because demand was so high they could charge whatever they like. So only those guys with actual money – and a ton of it – were able to get anything. The rest of us are scratching whatever we have in our freezers and pantries and things. We’re starting to see the food that has a use-by date being given away, which is awesome, but it won’t last very long.
The banks and major supply chains will get their shit together pretty quickly, though. It’s not like all communications are down – phone lines still work, texting still works, it’s just the core data exchange that isn’t really functioning. For those who have ledgers and manual methods of recording things, nothing has really changed.
That is to say that the data exchange protocols haven’t been interrupted. If somehow those protocols are interrupted, damaged, or suddenly changed, then no communications would work at all.
Nearly every industry would be at a standstill. No internet means that every single supply chain was down for a while, so once the money thing is worked out, things will be running again. When you can still text and call, and send letters and stuff, life takes a little bit longer but it still runs.
Even medicine, for all of its connectivity and essential databases and things, still works. It just means that people have to use their brains and training a bit more, and keep better records by hand. References formerly used online are being used in hardcopy wherever possible.
So long as data transfer protocols aren’t interrupted, many things would adjust
So, you know, much of life would return to some kind of normal. What we know as life now is not what I grew up with, and I’m only 36. We didn’t have internet in any kind in regional Australia until about 1994, and I was already fourteen. All business ran perfectly well without it for a very, very long time.
And when I say ‘much of life’ would get on pretty well, I mean, every business that is not an internet business. And this means that nearly my entire means of earning anything has suddenly disappeared. Let’s face it: There is not much call for any kind of writer or strategist if there is no demand. No internet means no websites, which means that any kind of strategy or writing for anything in that space is not required. But the underlying skills are applicable just about anywhere.
If you can’t use emails, then you can still exchange files, for example. It just wouldn’t be so fast. We would go back to exchanging discs, USBs, and so on. But it’s still possible, for the work that needs to be done.
It just means that there are a whole lot of things that we can’t do. The premise of my business, and all the web guys out there, is in question. Every method that we have for connecting, publicising events, and getting people together, is gone. Everything that we were reading online is gone. Everything that was the basis of shared knowledge is gone.
However, the internet is gradually permeating everything: From petrol pumps to space rockets. It doesn’t need to, but it does. So in the age of the Internet of Things, downtime stops everything.
The penalty for relying entirely on the internet is potentially enormous, especially when the one thing that you can’t see, you can’t save, you can’t touch, is your entire working life. If every item that drives the internet was suddenly without electricity, in any country, there are whole sectors of life that have been utterly destroyed.
Many of us internet-focused business people live our lives online. We spend more time interacting with binary code than we do with the sunshine and fresh air. We spend time thinking about how we can tweak this, test that, do something more valuable for other people; but that entire concept of value is woven into the nature of this invisible, data-driven net.
The internet is not indestructible, but we like to think that it is
This is a big question, and tons of other writers have attempted to answer the question.
In 2012, Scientific American, attempted to the answer the question, Could the internet ever be destroyed? It came up with some interesting answers. Its conclusion was that you couldn’t really destroy the internet internally to a country via physical damage, because it is so richly interconnected. But you could destroy a country’s capacity to connect to the bigger internet cloud because of the undersea cabling that exists. Doing that would be an act of war, and even if the infrastructure were damaged, the caches that exist in nonvolatile memory mean that copies of the internet exist even without power, and even if you don’t access it. It’s like “keeping” archives. A much greater threat, the article argues, is government regulation.
This is given the nature of the question, which presupposes that there is a person or thing who is actively doing the destroying.
In a similar vein, Sam Biddle wrote an article at Gizmodo titled How to destroy the internet (but please don’t). I can’t tell you the date of that because the internet is tricking us (the Gizmodo article is dated 2015, but this article at io9 references it as being ‘earlier this year’, and that was published in 2012. So… go figure that one out!).
Anyway, Biddle asked a question about how to destroy the internet and got a much more comprehensive, if grittily realistic answer. That answer was to cut the major international cables connecting continents under the sea; then get to the root servers, then the individual data centres. Interestingly, he discovers that you could cut the cables with an axe (they appear from time to time on beaches, as the ocean unburies them), and then smash the servers. But the amount of effort required would be preposterous to the point of making it pretty well impossible.
Following on from Biddle, George Dvorsky addresses more technical notions. For example, the fact that the protocols governing data transfer (IPv4 and IPv6) are at play. These protocols govern other types of communications; your mobile phone won’t function if these protocols are somehow interrupted, for example. Therefore, the fictional situation earlier in this essay – where SMS still functions – would be in appropriate.
One person that Dvorsky talked to, Dwayne Hendricks, argues that the internet, really, is people. While the internet as we know it is a ‘hive mind’ borg, meaning that interruptions are inconveniences and not actual problems, it is like this because it’s managed by people. And that therefore, the only way to destroy the internet is to remove its people.
Then, in 2015, Chris Baraniuk looked at the disastrous events that would break the internet. And one person that Baraniuk interviewed, Vincent Chan at the Massachusetts Institute of Technology, talked about packet protocols. He argues that if you were to introduce noise into that system that disrupt the receipt of either end of the packets, then we would really be in trouble because it would be unreliable. And it would just appear like the network was exceptionally busy.
As Baraniuk writes:
“I think there should be discussions of attack and defence of the internet as an entity,” he says. “That’s never been discussed before adequately.”
But the over-arching narrative, despite call-outs like these, feel a lot like common sense. They say, yes there are only about 30 huge internet exchanges in the world; and only a finite (and much smaller) number of deeps-sea cables. It says, yes there are multiple cables connected countries and islands and continents; and yes there are only a small number of really significant data centres, but because of all the backups and (insert fluffy stuff here) and all the security it’s not possible to kill it.
It seems that people tend to accept this idea that destruction is not possible, even though the internet (as Chan pointed out) ought to be considered an entity. An entity is a thing with distinct and independent existence. Whether or not it is alive requires more intense consideration of the notion of life. It introduces a far deeper and knottier element to the idea of destruction.
Woven into the entity question is the problem that nobody really knows everything that the internet is used for. It’s possible that you’ve heard Danny Hillis talk about this, and more importantly about the fact that because the internet works, nobody wants to talk about it not working:
And in fact, nobody really exactly understands all the things it’s being used for right now. It’s turning into one of these big emergent systems like the financial system, where we’ve designed all the parts but nobody really exactly understands how it operates and all the little details of it and what kinds of emergent behaviors it can have. And so if you hear an expert talking about the Internet and saying it can do this, or it does do this, or it will do that, you should treat it with the same skepticism that you might treat the comments of an economist about the economy or a weatherman about the weather, or something like that. They have an informed opinion, but it’s changing so quickly that even the experts don’t know exactly what’s going on. So if you see one of these maps of the Internet, it’s just somebody’s guess. Nobody really knows what the Internet is right now because it’s different than it was an hour ago. It’s constantly changing. It’s constantly reconfiguring.
And the problem with it is, I think we are setting ourselves up for a kind of disaster like the disaster we had in the financial system, where we take a system that’s basically built on trust, was basically built for a smaller-scale system, and we’ve kind of expanded it way beyond the limits of how it was meant to operate. And so right now, I think it’s literally true that we don’t know what the consequences of an effective denial-of-service attack on the Internet would be, and whatever it would be is going to be worse next year, and worse next year, and so on.
What this leads us to is that speculating about the disruption, interruption, reduction, or elimination of this technology is quite difficult. Therefore, we tend to rely on the common sense notion that all these things are in place and pretty strong, so it’s not possible to interrupt it or destroy it. People like Hillis try to challenge this notion, but anything that challenges a common sense framework is more likely to be ignored, even when it’s critical. (Read Everything is Obvious by Duncan Watts for more about that type of thinking. Yep, that’s an affiliate link.)
An inability to challenge an accepted idea is what contributes to the destruction of societies and empires, because it ignores the simple fact that everything is in a state of impermanence. That impermanence is something to which we have become less and less attuned as time goes by is significant in the context of this essay.
Yes, the internet is a fixed thing that drives stuff but it is extremely unlikely to stay that way. It is the nature of reality that is impermanent. This is not just the fact that people and things die, and that technology evolves. But it is the simple fact that this second is not like the next second; that whatever you are feeling right now is not what you will be feeling in ten minutes’ time; that your inner thoughts change; that even the sensation of an itch or a pain will change the instant that you focus on it. That things we believe are fixed change, and by the time we see that change we are completely baffled by it, and hate it. But that’s the basis of all life. The notion of impermanence tells us that everything changes; and when we pay attention to that idea, we can see that the cycles of impermanence also vary from thing to thing. Therefore, our society is a constant now, but it is not a permanent idea.
Therefore, the internet is not permanent. The idea that your business could be changed, shut down or interrupted permanently or semi-permanently in an instant ought not to be overlooked.
In my own case, the inherent complexity of the cloud systems in which my business is embedded, means that any significant interruption to the ‘supply’ of the internet is a huge concern. As you’ve seen earlier, everything that is core to my business is held by another company, and most of them somewhere else in the world.
This means that even assessing whether or not my business is 100% ethical is a monumental task. And if I needed to have everything held here, where I can control its access and use? It’s not impossible, but it means dramatically changing the software that I use and the business systems on which I rely. It can be done, but it creates a paradox where I am at once more and less able to be mobile: More, because I would not be internet reliant; less, because I would have to carry more things.
The decline of client-side software in favour of ‘cloud’ based software, means that you are reliant on a complexity of moving parts that has a large amount of risk attached. Proponents of the cloud (who sell it on the promise of greater agility and efficiency, and the fact it’s easier for them to make money) conveniently ignore this risk when they sell it to you, but that doesn’t mean that the risk can be ignored.
The first risk is access at your end: Your access has to be continuous and uninterrupted. Therefore, your ISPs, your country’s internet switch, your local and national and international infrastructure, all have to be in place and functioning. They have to be without problems for your business to function without noticing all the parts, which 99.9% of the time is the case. For cloud systems, you have to rely on the security and safety of what’s at the other end – from the infrastructure, to that country’s internet switch, to that country’s laws, to the ISP, to the servers, to the antiviral protections, to the amount of server space they can maintain, to the continuation and profitability of that company… There are a lot of factors that you will assume won’t interrupt your service, but which could.
You also have other things that come into play. Supply of electricity, continued access to technology and the ability to replace technology when it fails (and be able to do so at short notice and often without a budget for it) also play a huge part.
Then, you also have to pay for access to the cloud platform (and your information therefore), often on a monthly basis. So if you can’t afford to do that, then you lose either the platform or your information, or both.
Like I say, there is a phenomenal amount of moving parts. The system, which feels buttery and easy, is highly complex, but it is well designed. Otherwise it wouldn’t have a buttery feel. For it to work for us we need to accept (a) the complexity, (b) the possibility of loss, and (c) the idea that some parts of this complicated system can break. Then, having accepted these things, we have to weigh convenience against the risks and the likelihood of occurrence.
It’s a lot of thinking; the kind of thinking that very few people will do. Give me convenience or give me death, as Biafra would say.
What is the plan, then, if you have such a high level of complexity, and a business that relies on the internet not just to funtion but for its entire premise? What level of thinking do you need to get to, in order to unravel this problem?
Well, let’s have a look. We will look at my own business as a bit of a case study.
For all of the cloud systems I use, much of it is used because it’s convenient. But if I had to, I could replace all the project management tools with a Kanban board on the wall and some cards and pens, and I would immediately have a replacement for JIRA.
Similarly, I could have a notebook, a spreadsheet, or contacts binder and I would immediately have a replacement for Podio.
A hard copy binder already keeps all notes for my clients, and a hardcopy diary already keeps my appointments anyway, so they are not critical path tools; they just make my life easier and less tied to one physical location. I sync them with other tools so I’m not hampered by kilograms of stuff when I’m travelling around the place.
For my accounting, that’s a different story, and so is my banking. All of this critical path financial stuff is cloud-based, and all of it is now electronic. I would need to retain backups and downloads of all accounts, and personal copies of all bank statements in order to mitigate the risk of loss. Even though it’s not something I do at the moment, writing about it is making me realise that it would be a sensible idea! Then, even in the face of internet loss I would still be able to deal with my debtors & receipts. To do it, I would need to keep backups every fortnight, to match the financial cycles of the business.
Backing things up and keeping copies sounds like it would require servers and things. But that’s not true. The technology of SD cards, given the evolution of smaller devices, means that getting stable, unpowered, hard-to-damage storage is cheap. Even having terabytes’ worth would take up less space than my wallet. It’s just doing it and maintaining the habit of creating and replacing backups that is the mental struggle.
For all of the other work that I do – all the writing, for example – that’s easy. I can simply work off-line, like I do for my other stuff, books, essays, that kind of thing. I don’t need Google Apps for that; the only reason I have it is because it makes the second stage of the work, the collaboration, simple and easy. It also keeps everything synced easily between multiple devices.
Most of the services that automatically sync between your devices are problematic too, which many don’t realise. You either have to pay a fee for them, or take the risk that they aren’t going to shut you out. I’ve had the experience of having multiple services shut down at 10 days’ notice, requiring me to find data retention replacements in short-notice, frustrated alarm. One of them even started deleting my material by itself, and then didn’t give me any help to recover it. I have had 8000 files deleted by someone else at another company, because they can.
So, where are we? Right, well the writing stuff can be done without the internet if necessary. But the reason for even getting much of the work is because of the internet! If access to it was disrupted, or a country’s switch was flicked (for example, during a time of either population repression or international conflict, both of which are not inconceivable in the current political climate) then I would actually have to find another way of earning money.
Now, that wouldn’t be too hard. Of all skills, mine are enormously flexible. Offline work is just more difficult to drum up, because you have to do all the leg work manually. If for some reason there was no ability to apply my skills to the internet, then I can apply them in any area. I could develop company manuals, edit students’ work, edit authors’ works. I could work in public relations and communications. I could run leadership training. I could do freelance writing for publications. I could even – because I am also a quality systems auditor – create and/or audit other businesses’ quality systems.
I could also train people to write books, train people to edit books, train people in meditation or collaboration methods. I could create things and sell them at markets.
I have tons of skills; and if I had to use them to earn a living, I could.
But for those that are technology based, it would require that the technology used for the job is able to kept entirely in-house.
Before you scoff at this, let’s consider my dad’s company. He is an auditor and works in major utilities; he has a mobile business, and travels extensively. He lives in a rural area where the internet is not fast or guaranteed lots of the time. And yet he keeps all of his technology client-side and has remarkable data-retention systems that work, and have worked, for more than 20 years. Where my dad has the capacity to find a document he wrote in 1997 for a client, and doesn’t have to spend any money to do so, I struggle to locate things from five years ago. And, being cloud based, my own systems are largely uncontrollable, have more moving parts, and are more expensive. My dad’s systems require some expenditure once a year to replace an old drive, and the application of good administration habits. My expenditure on SaaS systems is in the order of $500/year just to have them.
That’s a whole lot of money for a whole lot of risk. It’s real risk, some of which I have already had to deal with over the past two years – making the likelihood of risk at the high end of the scale.
So what does this all mean, then?
Where are we, having thought about all of this? The thought experiment started out as an idea to examine systems complexity and to consider what happens to an internet-based business in light of internet failure. And where we have come to is the notion that although it seems low risk in terms of outlay and physical things, and less painful in terms of managing data retention, the risk of problems does exist. The internet can (and may well) fail at some point in the future, and in the interim it introduces a lot of moving, uncontrollable parts that we just accept as necessary risk.
It is unlikely that the internet will ever be interrupted, but it is not impossible, and it ought to be accounted for. And if your business had to exist without it for a period of time – from one week to six months – could you do it?
Entrepreneurs spend all their time building financial runways. How many build technology runways or plan for the need for different and alternative capabilities? I’m lucky that my skills and my business don’t require online businesses, but only that its application right now makes it easy. It can be reoriented fairly quickly; and the application of alternative skills (such as in auditing) means that I could go sideways fast, if I had to. There are always benefits to the acquisition of new skills, because it expands your ability to roll with changes, even – in the light of this essay – incredibly dramatic change.
The other point is that bringing everything back in-house (if I had to) is not impossible; it just requires some consideration, time, and creation of good habits.
Moreover, this thought exercise has taught me that backing up the cloud based systems is still something that any diligent business owner ought to be doing. Retaining this data doesn’t take space: SD Cards are cheap, hold a lot of data, and are tiny. But what it does take is time. The question then becomes, how much time are you willing to expend to maintain your business exactly as it needs to be, as against the risk of losing things and what kind of damage would that risk do?
But for the risk of internet outage, that’s a much bigger deal and would require some fast thinking and swift action to morph your business system into something new.
If internet access did fail for people, how long would you take to wait for it to resume before you took action? Two weeks? One month? How long could you survive if you had to rebuild a business to the point where you were getting paid, factoring in the lack of convenience that the internet gives us?
The outcome is perhaps then that when we think about future-proofing our businesses, it lies in understanding what can happen in the ultimate scenario. If the premise of your business is pulled out from underneath you, and you couldn’t rely on technology to solve the problem, what would you do? And if you find that the answer is a deep, black hole, maybe that’s something that you need to consider.
You could argue that this thinking is pointless navel-gazing. Any good risk manager would say that you need to assess the likelihood and the extent of the damage of this risk against any other risk before spending time thinking about it. And there are good reasons to say both of these things.
As business owners, though, we are each of us at risk of thinking of the next thing, of doing the daily work, of living in a future state instead of using what we have right now to assess possibility. We are putting ourselves at risk by not thinking about possibilities, of outcomes, of imagined futures that involve things that we don’t necessarily want to see, or that might cause us problems.
Let’s not be the generation that argues that the internet is so well connected that its destruction is impossible. Doing so stops us from both seeing and interrogating all possibility; instead, it locks us into a dominant way of thinking on a topic that, for all of its apparent complexity, really does just boil down to a country’s switch, a few cables, and some data centres, all of which work thanks to electricity. I would argue that even if this is navel-gazing, experimenting with ideas and outcomes can formulate better ways of doing, and of being, and show us gaps that we would otherwise not see, and which may, at some unpredictable time, enable us to find a new way of moving forwards.