Science / Tech, Top Stories

Is the Internet Complete?

In 2013, a debate was held between friends Peter Thiel and Marc Andreessen, the thrust of which was to determine whether we are living through an innovation golden age, or whether innovation was in fact stalling. Thiel, of course, played the innovation sceptic, and it is interesting now with five years remove to look back on the debate to see how history has vindicated his position. In short all of those things that were ‘just around the corner’ in 2013 are, sure enough, still ‘just around the corner.’

One strand of Thiel’s argument at the time (and since) was that the ostentatious progress made in computing in the last 15 years has blinded us to the lack of technological progress made elsewhere. We can hardly have failed to notice the internet revolution, and thus we map that progress onto everything, assuming that innovation is a cosmic force rather than something which happens on a piecemeal basis.

Certainly, this argument has gained more traction since 2013. However, in this piece I’d like to add an extra layer to it. Is it possible that innovation is not only stalling in non-tech areas, but in tech itself? Could we make an argument to say that the internet itself is, in fact, complete?

The driving logic for this argument is easy to dismiss—namely that all of the big ‘possible’ ideas associated with the internet have been taken. One might say that companies like Google, Facebook, and Amazon were all inevitabilities from the moment computers around the world started to link up, and that once these roles were filled, innovation started to dry up as there was fundamentally ‘nothing left to do.’

The first counter to this is that it’s easy to say in hindsight. Sure Amazon—or a company like it—seems like an inevitability now, but there was once a time when people were highly sceptical of the idea that anyone would want to conduct any type of financial transaction over the internet. The second counter, proved by the first, is to say that we can’t possibly know what might be coming over the horizon at any given time. The next Google might be just about to break, and if it were to do so then it would make a mockery of such defeatism.

Both of these arguments are fair and true. However they simply refute the idea that the internet is finished at this moment, rather than the more fundamental idea that it’s possible for the internet to be finished at all. It is this second idea—or at least the theoretical possibility of it—that I want to illustrate here.

Let’s compare the internet to another world changing innovation—the car. The car started as a ‘base concept’; a motorised chassis to transport you from point A to point B. That was the car on ‘day one,’ and this underlying concept has remained true up to this present day. However, that does not mean that the idea was complete on ‘day one.’ Over time, the car was innovated upon and developed. We added passenger seats so you could take people with you. We added a roof, so it wasn’t only suitable for fair weather. We added air conditioning to keep us comfortable, and a radio to keep us entertained. And, of course, we dramatically improved its performance and reliability. All in all, it probably took about 60 years for the car to go from ‘base concept’ to ‘finished article,’ from which point all cars have remained, on the whole, the same. Sure, a car from 2018 is far more advanced than a car from 1965, but it isn’t fundamentally different. It’s just a more polished version of the same thing. The 1965 car is, however, quite a lot different from an 1895 car, because that was the period of true innovation that fleshed out the idea.

We can say, therefore, that the car—as a concept—is ‘finished.’ Now, that isn’t to say of course that there has been no innovation since 1965, and that there won’t be any innovation in the future. Far from it. But it is to say that this innovation has been, on the whole, mere improvement on a static idea. Cars are cars, TVs are TVs, washing machines are washing machines. Once the idea is complete, we merely fiddle in the edges.

In spite of this precedent, we don’t see the internet in the same way. We don’t see the internet as a ‘base concept’ (i.e. a vast directory of information), which is gradually being shaped and polished into a finished article, from which point it will just tick along. Why not? I would suggest it’s because of the business structure. With the car, you had competing businesses each turning out their own version of the idea. Ford versus Mercedes versus Nissan. However, with the internet, you don’t have different ‘competing internets,’ you just have one—and business’s role within it is to look after the component pieces.

It’s a bit like there had only ever been one car, and different brands had each brought a new addition to the table to create the final useful thing. Facebook came along and put in the seats, Google the driving interface, YouTube the radio, and so on until the car was finished.

Seeing the internet this way, we might speculate that we have come to the end of the initial shaping of the idea, and that from this point on we shall merely be optimising it. We have on our hands the internet equivalent of a 58 Chevy—there’s a long way to go, but fundamentally it does what we want it to do.

Now, maybe this isn’t correct. Perhaps we haven’t reached that point yet. Perhaps another huge innovation is about to break. But the point is that by this logic we should expect such a slow-down at some point. Just because the internet isn’t something we can take in our hands, that doesn’t mean that it shouldn’t conform to the same development cycle as every other human invention ever created.

Innovation is not a linear process. It comes in fits and starts, propelled by the introduction and subsequent optimisation of these ‘base concepts.’ If we deny this idea, and if we expect the internet to become this self-perpetuating innovation spring, then quite simply we will cease to innovate. We will focus all of our energy on endlessly tweaking this one creation we have, and in the meantime will forget to create the next one. We will create the Homer car from The Simpsons—as advanced and complex as it is pointless. And, in the meantime, the grounds of real innovation, the source of the next internet, will remain fallow.


Alex Smith is a strategist specialising in the underlying nature of complex systems and companies.  He is founder of Basic Arts and you can follow him on Twitter @smithesq


  1. Tim says

    Cars are back to being not finished. Their current ability to second-guess drivers (lane keeping, automatic braking) is new. There’s work currently ongoing to do away with the need for a driver completely.

    • Hi Tim, also let’s not forget renewed investment in “flying cars” using the multi-copter drone architecture.

      • markbul says

        So you want the same a-holes who cut you off and run stop signs flying overhead? I think that’s called a non-starter.

        • @markbul

          A: not the point; development is underway empirically, your opinion is irrelevant

          B: pair flying cars with autonomous piloting and your imagined fears evaporate anyway

  2. Andrew says

    I think the author has fundamentally misunderstood what the Internet is. The Internet is not a technology, it is a medium. For example, there is an important distinction between a television set, which is a piece of technology that can be improved by increasing its resolution, compactness, sound fidelity, and so on, and television itself, which is the process of transmitting pre-recorded or live audio and video from a central broadcasting location to a large number of receiver sets.

    The internet can only be superseded by a superior medium, which will probably require different accompanying technology.

    • Jeremy says

      Andrew mostly said it a little better than I was going to say. Although I probably would use a different analogy than TV. I would say that the internet is much more broad – like roads. Roads have existed for thousands of years but we only got railroads a couple centuries ago and freeways a few decades ago. And only now are those roads used to carry goods imported from China or transport people via Uber services. The internet transports information like roads transport physical goods. The possibilities are very wide.

      • markbul says

        Roads are still roads – bad analogy for your case. And internet that delivers 3-D porn is the same internet. The internet is a communication medium. What’s missing? Maybe real-time video. But then that’s already possible. And do you really want the trolls you argue with in comment sections to see you? And do you want to see what they’ll show you? Ewwww….

  3. markbul says

    If you think that the internet isn’t done, then what is it missing? And no, you can’t answer ‘we don’t know yet.’ We’ve had time now to imagine the possibilities, and people with money to burn have been burning it for years now trying to come up with something new. Amazon was a logical outcome of the technology – it’s just an electronic retail catalog business. Twitter is a party line telephone. Paypal is a credit card. I think we can safely assume that the internet won’t be giving you a BJ any time soon – nor will it transport you to the deck of the Enterprise.

    • Ivan says

      If you can say PayPal is a credit card, then some websites are… giving you a “BJ”.

  4. The internet is not complete, cannot be, ever. And it will ever more closely approximate it, just as science (at its best) is continually becoming less wrong (though unlikely to ever be “right” – just a more useful approximation at some level).

    The internet is a network communications infrastructure.

    It’s basic protocol (TCP/IP) was developed by DARPA (Defense Advanced Research Projects Agency) to allow reliable battlefield communications as long as any possible communications path remained in tact on the battlefield (ie, when paths were coming and going frequently).

    Aspects of today’s infrastructure are highly centralised for a variety of levels of “economic” and “security” reasons.
    Alternatives to the existing “internet” are possible, even probable, and certainly under development.

    One of the major issues, that happens with all complex systems, is that as the complexity of the system increases, the impact of any innovation on any particular part of that complexity decreases, so that we see smaller and smaller improvements coming from any particular innovation (when looking at the system as a whole). When that system is embedded in an economic system that has to show a return on investment, then the rate of implementation of innovation decreases, as the return on investment slows. Couple that with the need to maintain backward compatibility to older, but still functioning, devices out there in the real world, and we see what we have.

    Those of us who have been in the coding world for a few decades know that most of the banking infrastructure is still on a CoBOL code base. “If it aint broke – don’t fix it” rules – particularly in the more conservative institutions.

    Cars are also changing, rapidly.
    Self driving cars are real, if not yet common. I saw my first Tesla Model S on a rural New Zealand road two weeks ago.

    Flying cars are unlikely to ever be common – the combination of risk profile, required energy density. and weight restrictions imposes extreme complexity (itself a risk factor) for those of us that value life highly.

    As we move from fossil fuels to solar power, with very high speed internets, then we may see surface vehicles that are much slower and lighter taking us places extremely safely, extremely reliably, extremely efficiently, while we immerse ourselves in whatever communication medium takes our interest. Risk of serious injury in vehicles traveling less than 20 miles per hour is almost zero. If all goods vehicles are unmanned, what does it matter that they take a few more minutes or hours to get where they are going.

    High speed travel is likely to be on underground maglev trains in evacuated tubes – no air resistance – 1g acceleration to half way, then down again to destination. With local feeders to major nodes no place on the earth would be much more than an hour’s travel from the closest local node – which would likely be within a few minutes travel of a 20mph surface vehicle. (For those who are willing to accept the slightly higher risks of high speed travel).

    Fully automated systems are changing things exponentially.
    That rate of change is increasing.

    When we achieve fully automated systems capable of making a copy of themselves using local rock and solar power, then if it takes the first two weeks to make the second, then within 2 years there can be one for every person on the planet, and the very notion of value in exchange (market value) is not simply redundant, but becomes the single greatest source of existential level risk to us all (with all the conceptual “drag” that comes with it). {The energy requirements of the last few replications mean that the process would probably best be done in orbit using lunar mass, but that is a another story.}

    And that is the sort of transformation speed that is possible – poverty eliminated within 4 years.
    Every individual on the planet with practical access to the sort of life choices available only to the most wealthy today, and full recycling and ecological responsibility in a post market “economy”. A literal global village.

    Not to say that it will happen, and it is most certainly a technological possibility.

  5. I think the author has fundamentally misunderstood what the Internet is. The Internet is not a technology, it is a medium. For example, there is an important distinction between a television set, which is a piece of technology that can be improved by increasing its resolution, compactness, sound fidelity, and so on, and television itself, which is the process of transmitting pre-recorded or live audio and video from a central broadcasting location to a large number of receiver sets.

    I’d add another dimension, which is the progress from morally simplistic and formulaic cop shows (which still persist, of course) to complex, long-form or novelistic dramas like The Wire or Breaking Bad.

  6. Clinton says

    It’s silly to talk about the internet as a thing in-and-of itself, instead of talking about computers directly. Computers and software are the fundamental building-blocks of our communication networks now that they’ve swallowed telephony and facsimile technology. To even ascribe business as a primary factor is to mistake socio-economic content for the medium itself. The internet is a bunch of connected computers, which are universal machines. The internet we have today is an infinitesimally tiny manifestation of the potential state thah all those computers could theoretically be in. Furthermore, with Free Software users control their computers, not corporations.

    Once people who are ignorant about how computers work stop hogging the microphone with their uninformed, 20th century thinking about them then things are going to get really weird, really fast.

  7. I think it’s fascinating that the comments section is effectively focused entirely on what the title statement means rather than whether it is true. I think this reflects a key point of clarification that the article fails to make as the most important point is what do you mean by “is the internet complete”. However, the author isn’t really claiming the Internet is definitely complete. The real focus is to claim that there is a point where it is complete and then wonder if we are there yet. Within that context there is a big question about what the “Internet” is. From a pedantic point of view the Internet is merely the IP based network spanning the globe over which data is transferred but from a non-pedantic colloquial point of view most people would use the word Internet to mean the whole world wide web infrastructure and everything related to it.

    The proper question is “what can we use the Internet to do that wasn’t done before?” and the follow up question of “have all the key opportunities been filled”. The Author is really asking whether there are going to be any more big internet based companies exploiting new ideas that create game changing situations in the economy. As far as I can see the core capability created by the Internet and related communications infrastructure are opportunities for disintermediation, the removal of middle-men or replacing of existing intermediates with direct contact or much cheaper intermediates. Amazon, Ebay and so many other direct sales organisations are the obvious examples of this in retail and retail is the one area where this explanation is most obviously true. All the social networks/apps are an obviously similar process in social interactions. This brings to the point where we need to raise the issue that we aren’t really talking about the traditional internet, we are talking about a broader infrastructure including all of our data communications capability and the infrastructure that uses it.

    If we look at this from the wider sense then we could claim that we have now created a global data communications infrastructure where all the key devices – servers, fixed network, mobile network, home computer, tablet, mobile phone, wireless light bulbs (etc) all exist and there is nothing fundamentally new to be added, we just make these devices better in the same way that we progress from early motorcycles, cars and trucks to modern equivalents. Nobody innovated a whole new vehicle class (go on, prove me wrong), they just made them better.

    The real question is does an explanation of this sort tell us anything useful about the future. My view would be that the author definitely has a point and that we could very easily now be into the phase of incremental improvement rather than basic innovation. The last basic innovation really was the wholesale integration of mobile phones with GPS into the overall network leading to a number of genuinely new things. Are we now in a period of market optimisation where new internet based companies are exploiting much more marginal gains and therefore the big “boom” is over or is there something else fundamental to come. And, in line with the author, even if there is another thing there will come a point when there isn’t.

  8. Pingback: Is the internet complete? - Basic Arts

  9. Seems pretty flimsy to argue lack of new internet developments are systemic technological issues. Especially during the same time period that just so happens to have seen the run up to, and imposition of the most sweeping regulations on internet infrastructure and business models in history.

    In fact, net neutrality was a direct response to the spectre of ‘smart pipes’ and ‘zero rating’, radical innovations in both pricing and data flow models that would result in better, faster, cheaper internet. But because it *might* not be completely fair to people who can’t afford bandwidth (how is the status quo though?) They had to implement nn. Q.e.d. it was regulation that stopped the march of innovation in the internet.

  10. Damien Van Brink says

    I agree partially with Theil, in that yes we have given too much airtime to the tech evangelists and their latest new thing that you definitely don’t need. Innovation may have slowed but I don’t think its BECAUSE of the computing/tech sector. A number of smart people have suggested multiple economic policy choices that discourage innovation: Low wages/Stagnant wages – as why would you invest in technology, or even look for other solutions if you can just pay somebody $10 to do it. Poor welfare state – why would you have a go at a new business if failure can mean living on the street. There were others but only those two stuck.

    I’m not sure looking at technological innovation in the frame presented in the article is the best. A car, for instance, is fundamentally the alloying of a combustion engine and wheels. Wheels have been around for quite a while, and engines over 100 years, so you could argue the car is nothing noteworthy. Production efficiency over time increased to the point where you could conceive of create consumer products using these formerly expensive technologies. Same is happening with genetics now, where a sequence formerly cost billions of dollars, it now costs thousands. I would suggest that a whole array of products and services will flow from this once the price comes down and scientists, engineers and entrepreneurs can get creative with it.

  11. Jared Stuart says

    In my opinion the internet isn’t quite complete yet because it still lacks some information. For example, if you search up some pass event way back in history, the internet will only give you bits and pieces of the information whereas a history book will give more. When machinery was invented, everyone taught that that was the end but up to this day we’re still finding ways to improve. Not everything can be found on the internet, but most things can and at the same time it might not always be free. Moreover, the internet is relaying on human input. The Mind of the computers are based on all the information feed by human. The internet will always continue because every day there is a new set of information being put out there.

  12. Pingback: Quillette – Internet complete? | Ted Howard NZ's Blog

Comments are closed.