The Technium

The Myth of Leapfrogging

Hour by hour technology continues to demassify, to shrink in size while expanding in power, and to obsolesce earlier measures of excellence. Through the rear view mirror we spy the manufactured stuff of yesterday and are generally happy to have it recede into the past. Although there are exceptions, today’s technology is more desirable to what came before it. In particular, the products of the first phases of the industrial revolution seem rough, rigid, inert, and harmful compared to products available now. Who would not prefer the solar power panel to the long copper cables strung along on poles, or the efficient compact refrigerator to a clumsy behemoth icebox? To judge from consumer patterns from around the world, apparently no one.

Recent technology, particularly in communications, has disembodied to such an extant, that it has become an almost feathery accessory to the body, rather than a heavy appliance it once was. This liberation births a radical idea: perhaps the billions of people in the world currently living with little advance technology can skip the industrial generation of technology and zoom immediately to the good stuff. Rather than suffer trough the smoky pollution of industrial factories, begin with the best we have right now. The billions in the developing world could jump directly and immediately from the pre-technological era right into the nirvana of six-sigma manufacturing processes, light robotic assembly, and the mind-numbing choice of personalized everything. Because the developing world often lacks established large-scale technological infrastructure, those on the lagging side of the technological divide have a clear opportunity to install a state-of-the art system. They would be starting with a clear slate and could – in theory – populate their entire technosphere with only the latest stuff. This particular choice is not open to developed countries crowded with old systems still in operation and yet to be amortized. What an amazing chance it would be if the very first computer you ever turned on was a supercomputer laptop!! And your first electricity came from solar panels; your first house was super-insulated, your first car was hydrogen powered! In theory, the “rest of the world” could leapfrog right over the backs of the most developed, and land squarely ahead with a the least polluting, most efficient, best personalized and least constrained technology in the world.

In theory. That’s how it would work. The most commonly cited evidence for leapfrogging is the pattern of cell phone adoption in China and other parts of the world. For hundred of millions of people in Africa, Asia, Latin America, their first telephone is a cell phone. But second examples of skipping the industrial revolution are scarcer. I am aware of a small island in the Ganges delta of India and one village in Thailand that installed solar power. Beyond that, the examples of skipping the industrial revolution evaporate on inspection. A few pilot programs here and there, but no real adoption. In fact the closer one looks at the evidence, the more unlikely it seems to me that leapfrogging actually happens.


Part of the problem in trying to demonstrate leapfrogging lies in our definition of it. If we mean simply the broad notion of skipping a generation of technology, then leapfrogging happens all the time, but perhaps not in ways we approve. For instance, in Mongolia today often the first practical wheel technology to arrive in Mongolia is the “leapfrog” technology of Mercedes trucks. With these trucks Mongolians skipped over wheel barrels, carts and bicycles and shot directly to diesel powered lorries. In fact the trucks often come before roads appear. In parts of China the first water pumps a farmer may get may be powered by nuclear generated electricity, rather than say the initial windmills of the west. The first clock in a village in Nepal may be a digital watch. Or the first TV’s in Africa are often satellite driven color TV, skipping over terrestrial black and white technology. In these and many other examples, entire lines of technological evolution are skipped over. But surely leapfrogging must be more than the relative sudden appearance of modern gizmos, even if they skip the previous generation of gizmos.

Intuitively we feel that leapfrogging entails beneficial infrastructure, rather than consumer appliances. It’s not solar battery rechargers that count, but the absence of a huge electrical grid. It’s not cheap commercial jet fares, but the escape of having to building a long-haul highway system. It is not the latest SM mobile phoneset, but the ability to leave millions of mile of telephone wires unstrung. To reckon leapfrogging as significant, it must be seen as a way to skip a complete generation of infrastructure.

We do have examples of this. A large portion of Saudi Arabia and the Gulf States, as well as sections of the Amazon and African frontiers have been modernized with the wholesale importation of the newest systems. An entire 20th century infrastructure has been airlifted onto its soil and wired up – from roads, to power, to communication, housing, and utilities. Look at Dubai, which was not much more than sand and camel tracks a few decades ago. One day there is wilderness; the next there is urbanity, in semi-working order. Most people (except those living there) are horrified by this type of leapfrogging. These instant towns lack any of the smooth depth that ordinarily occurs in organic growth. Because they have skipped over so many steps in the usual evolution of technological accumulation, they are ragged, sharp, obnoxious, untested, and in many ways fragile places. They are certainly not what we would hope upon the rest of the world.

I conclude then, that what we mean by leapfrogging is the skipping over of one particular kind of infrastructure; specifically the kind of large-scale industrial style often imported wholesale into developing countries; and in place of that, the clean installation of a post-industrial more lightweight, less mass-bound, higher-tech system. Mere purchase of post-industrial products is not enough. In other words, leapfrogging means the acquisition of the virtues of a digital livelihood without having to suffer through the vices of an industrial livelihood.


That brings us back to cell phones. Until recently the common complaint of middle-class aspirants throughout the developing world – from the Mid-East to Latin American – was a continuous wail about the inordinate waiting times to get a telephone. This delay – years or decades — almost amounted to a prohibition against phones, and was all the more painful because the centrality of the telephone in modern life was clear to everyone, particularly those without one. The digital economy is primarily an information and communications economy, so unless you were connected, you did not exist. Until the mid 1990s, most of the unwired world did not exist in the modern economy.

Then cellular phones came along. Cheap, portable (hidable), and available now, they were a technological tsunami crashing through the wall of political bureaucracy that had kept the phone locked up for the elites. There was a world-wise sucking sound as anyone with a few dollars grabbed a cell phone. Motorola went to China in 1995 to manufacture cheap cell phones for the US but never managed to export any because all were sold in China itself. The rising curve of phones sold in China was an astounding mirror of rising Chinese standards of living, and of their hopes and ambitions. Doubling every year, cell phone use was explosive, changing the culture in a way nothing had done since Mao. And this mild revolution was accomplished without having to build a million telephone poles, with a million linemen, and a trillion miles of copper. China had leapfrogged over the telegraph era, and flown directly to the digital gold zone. They had gone from unwired to no-need-of wired, and had skipped the wired generation altogether! If they could, then so could the rest of Asia, and Africa and Latin America.

If only that is what had happened. While the exponential increase in cell phones is certainly factual, the assumption that this new technology removes the need for the old is not. The simplest chart of cell phones and land lines per year in China shows the problem. As cell phones acquisition goes go up, so do land lines. Not as fast, but telephone land lines continue to increase in China. They sure aren’t being skipped over.


Perhaps China is an exception? So far the rest of the world is following a similar pattern. India, the other poster child in leapfrogging, is a few steps behind of China but reveals an identical pattern; an explosion of wired phones, trailing an explosion of mobile phones. In the rest of Asia, the mid east, Latin America and Africa wireless phones are proliferating faster than wired ones, but wired (the old generation) continues to expand as well. To be clear, the numbers cell phones in the world tipped past the number of fixed lines about October 2001. That means that three quarters of African countries have more mobile phones than stationary ones. All telecom forecasts for the next five years predict cell phones to continue to eclipse fixed lines. But every forecast also expects land technology to increase as well. The old style grows even as the new accelerates. (The only place in the world I could find a certifiable drop in land lines was in Chile in 2003 when fixed lines dropped to 3.25 million lines from 3.47 millions a year earlier.)

Why would the old generation continue to grow? One study of Latin American telephone adoption suggested that cell phones train people to rely on telephonic communication and thus boost their demand for land lines as well. Cell phones are training wheels for real bandwidth use. So far internet use is primarily confined to land lines. If you want to get online, you need good old copper (or optic) wires, so the real vanguards of society leap past cell phones to get dial-up modems or DSL. One way to interrupt the sizable gains in fixed lines around the world is to see it as internet conquest. Or as a indication of bandwidth. Mobility brings thin and spotty communication; fixed lines brings you deep, constant fat bandwidth. Those relations could change in the future of course, and probably will as technology evolves, but a fiber optic cable is like to beat out wireless for the immediate future.

Since internet connections are booming, might the wired trump the unwired in the realm of TV? What a lovely world it would be if the internet (interactive) leapfrogged over passive TV. But that is not happening either. While one in ten people on earth is an internet user (a success we would not have dreamed about 10 years ago!), the ancient technology of TV is more prevalent than ever: Globally there’s one set for every three people, and one for every household on average. Television adoption has been growing at 13% per year, which is actually faster than telephone line growth.

There is simply no evidence for the “skipping over” part in the one realm that everyone claims demonstrates the dream of leapfrogging.

What we see instead is “racing forward.” We see newer technologies booming on top of a slower boom of old technologies. While cell phones in China are increasing exponentially, China’s use of cement, steam boilers, steel, and all the other ingredients of the industrial age are likewise increasing.

I may be wrong but I believe you can not have a rise in new infrastructure technology without having a rise in old infrastructure. To a degree that is invisible to us, new tech sits on a foundation of old tech. Despite the vital layer of intangible activities which constitute our modern economy, a huge portion of what goes on each day is fairly industrial in scope: moving atoms, rearranging atoms, mining atoms, burning atoms, refining atoms, stacking atoms.

It is no different with our brains. Most of our brain’s activity is spent on primitive processes – like navigating — that we can’t even perceive consciously. Instead we are aware of only a thin, modern, sophisticated layer of “hi-tech” processing that sits on and depends upon the reliable workings of older processes. I don’t think you can do calculus until you do counting; I don’t think you can do cell phones until you do land phones. And I don’t think you can build a digital infrastructure without including an industrial process.

Does this mean that if we were to try to colonize an uninhabited Earth-like planet that we would be required to recapitulate history and start with sharp sticks, smoke signals and mud-brick buildings and then work our way through each era? Would we not try to create a society from scratch using the most sophisticated technology we had?


I think we would try but that it would not work. If we were civilizing Mars, a bulldozer would be as valuable as a radio. Just like the predominance of lower functions in our brain, industrial process predominate the technium, even though they are gilded with informational veneers. The demassification of high technology is at times an illusion. It is not that information technology has no mass, that it lives in an abstract virtual world. Rather, high technology is the embedment of information into materials, the seamless fusion of bits and atoms. It is adding intelligence to industry, rather than removing industry and leaving only information.

In this way, there is no hi-tech without low tech, and no leapfrogging over low tech. We may find times when a new technology races forward, and the velocity of the new outpaces the velocity of the old, but even then we won’t see the myth of leapfrogging.

  • Leadblind

    Hi Kevin,

    I found this to be an interesting article, but I’m afraid I have a few issues with it.

    The first issue is whether your examples against leapfrogging (landlines vs. cellphones, internet vs. TV) are really examples of new technologies failing to displace old technologies because “leapfrogging is a myth” or whether they co-exist because the older technology continues to have some advantages that the new technology simply hasn’t yet rendered irrelevant. Landlines as you mention, still lead in reliability and speed, whereas TV keeps you in sync with other TV watchers in your society (avoiding the “meme isolation” that can be caused by individual browsing choices). Broadcast media generally also removes the need to actively select and manage your content, which is a difficult proposition if you are not already quite educated and media saavy. None of these reasons imply that leapfrogging is a myth, just that the displacing technology hasn’t matured enough for a leapfrog to occur. Perhaps this is what you mean by “racing forward” rather than “leaping forward.” But I believe this is a questionable distinction when you consider it from the perspective of the citizen in the developing country. For an agrarian society which, less than a decade ago literally had nothing but plants and dirt, the installation of landlines and TV is nothing short of a massive leap forward in technology. In terms of changes to lifestyle and understanding, it is as large a leap forward as going from a medieval England to London circa 1960. Yes, it doesn’t hurl everyone into 2007 all at once, but wow, what a jump.

    The second issue I take with your article stems from the analogy you draw between learning to count before doing calculus and the need to implement “low tech” industrial infrastructure before you can implement “high tech” digital ones. I feel that this is misleading because you are comparing the infrastructure required for the production and movement of objects with the infrastructure required for the production and movement of information. I don’t think it is fair to compare the two, and certainly not to label one as “lower tech” than the other. Which is more advanced? The fact that we can make 5 million spooons an hour to a hundredth of a millimeter, or that we can instant message our sisters while we ride the bus? Why can’t my 802.11N router pave my driveway worth a damn? Clearly these are different technologies for different purposes, and to place them on a single continuum of “low tech” to “high tech” based on when they were first achieved isn’t fair. Maybe we put the making of spoons at a higher priority for a few hundred years?

    In contrast, if you look at the industrial infrastructure by itself, leapfrogging occurs in the same way as it does with digital technologies. Obsolete tools, methods (and yes, supporting infrastructure–no big aqueducts being ordered lately) are passed over in favor of the latest and most modern ones. Current industrial methods and tools, including bulldozers, are the human state of the art in making and manipulating the physical environment. The persistence of certain tools and technologies in the face of advancement in other areas does not show an inescapable reliance on “low” technologies but rather the limit of our knowledge in those areas and/or the excellence of existing methods. None of this implies that a developing country (or expedition to Mars) would select anything but the most effective and up-to-date solution available for a given problem, be it a bulldozer, a down jacket or an 802.11N router. To say that the presence of the first two indicates the persistence of “low technology” and using that to disprove leapfrogging does not follow in my opinion. What could they possibly leapfrog to when these are already the cutting edge technologies for their respective functions? If, on the other hand, Chinese construction is currently based on picks and shovels, and bulldozers become widespread, is that not a leap forward?

    Thirdly, I have a nitpick about certain examples you cite (solar energy, hydrogen cars) as indications of superior, demassified, etc. technologies that have failed to show evidence of leapfrogging. I would suggest that nobody so far has gotten these things to work cost-effectively on a large scale, so there is no reason to expect that a developing country would switch to them. This does not mean leapfrogging is a myth, only that leapfrogging cannot transport a society into the unrealized future on the back of unrealized ideas. Instead I would suggest that what leapfrogging does is rapidly bring a society into the present, and position it to succeed (due to a low burden of existing infrastructure and a high cultural acceptange of change). To put it another way, the idea that China or India have not transformed themselves into solar or hydrogen economies may say more about the current/future feasibility of solar and hydrogen than it does about whether leapfrogging in fact occurs.

    Finally, I would like to offer some historical examples of leapfrogging in the past as evidence that leapfrogging is no myth. These include the rise of the United States (relative to a more advanced Europe), post-war Germany and Japan (relative to dominant Allied countries), post(korean)war South Korea (relative to the west), Taiwan (relative to the west), and of course China, India, and Vietnam today. In each instance, the country began at an agrarian level (with the exception of Germany and Japan which were basically bombed into a relatively blank slate), received an investment of capital, embraced cutting edge technology, and had the chance to build new. In each case the phenomenon of leapfrogging brought these societies quickly close enough to the technical level of their national competitors that their cheaper labor and exchange rates were enough to compete. Based on the strength of their exports they then, one after another, moved from being high-volume, low-cost producers to increasingly high-tech and “knowledge” based consumer economies.

    I apologize for the length of this post but you did say this is a work in progess for a book, and I felt compelled to answer fully. Good luck with your writing.

  • Monte Davis

    I’ve been struck by the prevalence of the “why can’t [technology X] advance as fast as IT?” meme. It may be as popular as “if we can put men on the moon, why can’t we..?” a generation ago. (And, of course, Wired has done more than a little to propagate it.)

    Whenever I think of Moore’s law, I’m reminded of an obvious but often forgotten point Ralph Gomory used to make when he headed IBM research: IT is fundamentally unlike most other technologies in that it manipulates *patterns*, which can be instantiated at almost arbitrarily small scales and energy levels. Couple that with photolithography (for fabrication at very low marginal cost per copy) and network effects, and you have advantages that are very hard to match for any technology that has to manipulate matter or energy in macroscopic quantities.

    I’m working on a book about space, and I don’t think it’s coincidence that the entrepreneurs and angels include Musk, Allen, Bezos, Carmack, Anderson, and others who made their pile in IT. They bring much to the table in management and design philosophy, but beneath that I suspect there’s a tacit, ingrained expectation that their next rocket will inevitably do twice as much for half the cost. I wish them well, but the reality is likely to fall considerably short.

  • Thomas Bratt

    “New informational organizations are layered upon the old without displacement, just as in biological evolution.” – Just a note that readers interested in this may want to look up Rodney Brook’s subsumption architectures, used in robotics.

  • The problem with your argument is that history may not be a reliable guide.

    What you say is true, so long as centralized infrastructure and large-scale, capital-intensive, corporate-owned industry remain viable and accessible.

    But leapfrogging to decentralized, ephemeral systems may be the only viable option if adopting the older style of centralized infrastructure and capital-intensive production becomes unsustainable.

    I expect that we’re in the early stages of a number of dovetailing crises that will render the old kind of large-scale, industrial capitalism unsustainable: Peak Oil; assorted other input crises resulting from the inevitable economic law that demand for subsidized inputs (like long-distance shipping) outstrips the state’s fiscal resources for providing them; and the realization crisis facing capitalism, as described by Michel Bauwens, resulting from the growing unsustainability of IP-based business models.

    I believe state capitalism is hitting the wall of resource and input crises at the very time small-scale production technology of the kind Borsodi, Mumford and Bookchin envisioned is approaching a singularity. The latter singularity will mean the feasibility of shifting ever greater amounts of production from wage labor to the informal and household economies, which may be extremely welcome as a safety net when the old industrial economy stagnates and runs out of gas.

    The old centralized industry, to put it bluntly, may not be there. And the “small is beautiful” stuff may be all that’s left.

    • @Kevin Carson: As a student of Small is Beautiful and ecotopian visions for many years (I started out as an editor of the Whole Earth Catalogs), I hope you are right. But so far, there is no actual data as evidence for the vision to date. I am eager to see some if you can point me to it. I don’t mean all the negative reasons why it has to happen, I mean evidence of leapfrogging taking place.

  • Tony Fisk

    I think that landlines should be viewed as a separate technology that provides a different service (high bandwidth internet connectivity as opposed to real time personal communication)

    Are the landlines in China and India copper or optical?
    If copper, then you have a case.

    It might be a bit clearer if a high bandwidth, high coverage wireless technology became available (eg lasers on high altitude platforms)

  • Hi Monte!

    Kevin – you do have a point, but your conclusion: ” no hi-tech without low tech, and no leapfrogging over low tech” doesn’t meant that all low-tech is necessary. None of our modern society in the industrialized world depends on anybody using stone-age tools, there’s no dependence on animal labor except for those eccentrics who insist on it (the Amish for instance). However, we do still eat meat and have a dependence on animals being raised and slaughtered, a process that dates back thousands of years.

    So in a healthy society we have a very wide mix of economic activities. If you plotted activity in the US on a chart with one axis representing capital-intensity and a second representing technology-level, you’d see things across the entire range. Low-capital low-tech items like basic agricultural labor and simple services; low-capital high-tech various online services like things depending on Google adwords; high-capital low-tech like mining, high-capital high-tech like aerospace… we need all of that for a healthy modern economy, millions or even tens of millions of different actors playing their parts.

    But that doesn’t mean we would return to “sharp sticks, smoke signals and mud-brick buildings” on a new planet – barring nano-replicators (if you read Ray Kurzweil, they’re just on the horizon…) we’d have a smaller range of capital and technology level in local production, but there’s a base of knowledge available that does allow some real “leapfrogging” – and high-tech light-weight items could presumably be imported from Earth anyway.

  • Kevin Kelly


    For all the reasons you mentioned, I agree with you that old technologies are not inferior, and that many new ones like solar are not yet perfected. The only reason I mention solar and cell phones is that they are the examples usually given for evidence of leapfrogging.

    And so far they are the only ones. Your examples of the US “leapfrogging Europe” is a case of one country developing faster than the previous fastest, and is not — at least as far as I can tell — what anyone who makes the leapfrogging argument is trying to claim since everyone agrees with that.

  • is an article I wrote for WorldChanging which is directly relevant to this.

    “In particular, I’ll be covering what can be leapfrogged, what cannot be leapfrogged, some thinking about the embedded capital base of the first world, and “invisible leapfrogging” – the leapfrogging which happened so fast, nobody called it that!”

    Specifically, I’d suggest that consumer goods, landrovers and various other technologies have leapfrogged so effectively that nobody calls it that.

  • Old technology often has lower requirements (including required capex), meaning that without somehow finding extra resources poor countries have no choice but to “go retro”. That doesn’t mean leapfrogging is impossible or is a bad strategy (if you can afford it).

  • David Liddle

    Come on, Kevin. Recursion has a well-defined mathematical meaning, and shouldn’t be subject to literary license. Not everything that is self-referential is recursive, by a long shot. Certainly not “science looking at itself.” To quote L. Peter Deutsch, perhaps the best programmier of the 1970’s, “To iterate is human; to recurse divine.” Very few functions can authentically be rendered recursively. Don’t cheapen the elegance of that rare skill by a post-hoc popular application of the term to a phenomenon from the philosophy of science.

    Great book!


  • hi there,

    I think there is a typo here:
    One way to interrupt [should be “interpret”] the sizable gains in fixed lines around the world is to see it as internet conquest.



  • Bob Coppock

    I heard you rtalk Friday. If science is going to change radically in the next 50 years, is there any opportunity for slow science?

    That is, will there ever be studies that take decades or centuries, like the Framinham study? Many ecological experiments will take a long time to produce results. Much science now is limited to the time it takes to produce a Ph.D. thesis or the term of a post doc. AAfter all, you were speaking to the Long Now Foundaiton, which is supposed to think in terms of thousands of years.

  • Thanks for the good myth-busting screed, Kevin. One question I ask those who make way too much of leapfrogging: So what? From this side of the digital divide, leapfrogging seems a miracle, something akin to a person jumping into the 21st century from the hunter-gatherer era. But inside the developing world, it’s really just using what’s there. So lots of people in Africa have mobile phones who used to have no phones. Is that a revolution? Or just a convenient way to SMS friends? There’s nothing wrong with it, but it hasn’t fundamentally raised standards of living or pulled people out of poverty or challenged the corruption and exploitation that is the norm in many Southern tier countries. As in Mongolia, an old-fashioned Mercedes truck does more (as long as you can afford to buy one and pay for gas).

    My suspicion is that the amount of ink and hot air spent trumpeting leapfrogging is based simply on this: a Masai warrior is exotic and unapproachable whereas that same Masai warrior with a cell phone suddenly seems ‘just like us.’ Leapfrogging makes people feel that our technical know-how is somehow ‘civilizing’ the rest of the world.

    There’s nothing wrong with a high plains herdsman having a cell phone. Let’s just not pretend it is changing the underlying power relationships or in any structural way making for a better world.

  • Kevin Kelly

    Yes, Thomas, readers should read Rod Brook’s ideas about subsumption architecture, which I happily appropriate here.

    Tony, as far as i can tell the landlines I am counting are all copper lines — to the end user — and not trunk lines, which may or may not be fiber.

    Arthur, you slightly mistunderstand my point. I don’t think low tech is necessary for high tech to work – it is only necessary to reach high tech. The opposable thumb may have been needed for the evolution of human intelligence; it it was, it is certainly not needed for thinking now — we can imagine thumbless thought. I am suggesting low tech is scaffolding innovation like the opposable thumb for thinking.

  • PK Ramos

    As long as knowledge becomes available to all, it is evident to me that new, emerging societies do not have to go through the same tortuous path of trial and error invention that developed societies had to go through in order to advance and try to catch up. A society that does not have locomotives, for instance, does not have to go through the steam version to get to the diesel version to the diesel-electric and finally to some Mag-Lev technology if the final one can be learned, or acquired directly.

    A country that is going to build their first airplane EVER would probably not build a bi-plane, but cut directly to a turbo-prop small aircraft at once. That is leapfrogging.

  • Kevin, when you say “I don’t think low tech is necessary for high tech to work – it is only necessary to reach high tech” as an argument against leapfrogging, are you suggesting that every society has to somehow recapitulate the same steps to reach high-tech levels?

    If we assume that humanity’s accumulated knowledge is available, at least in some form, to every society, then my argument (perhaps not very clearly stated before…) is that SOME low-tech is necessary for the society to function well, but some low-tech is not. We still need police forces and sewer systems, we still need mining and agriculture, we still need “cement, steam boilers, [and] steel” but we don’t need candles or horse-drawn wagons or steam ships or quill pens (other than as luxury items for amusement).

    So some low-tech is essential, some is superfluous, for the modern world. There IS an issue that, to a degree, geography and local culture can require adaptation of the essential pieces; just importing them wholesale may not work or may not fit well. That leaves things in the sort of fragile state you mention in the Arab world.

    The view that leapfrogging means moving away from manipulating atoms is probably rather US-centric: we’ve farmed out a large part of our industrial production to other nations (China in particular). But modern manufacturing is itself very high-tech, with extensive use of robotics, I wouldn’t call most of the atom manipulation in the world these days low-tech work at all! Even mining and agriculture are highly dependent on computers and high-tech equipment, in the western world.

    There is also the issue of simple economics. Putting solar panels in African villages isn’t so much “leapfrogging” as unsustainable experimentation. If solar panels actually provided something sufficiently useful relative to their cost to the average African, you would see them taken up with the same speed as cell phones have been around the world. But if there are better things they can do with their cash, solar panels probably won’t be high on the list. In fact this whole push to get away from the centralized (efficient and cost-effective) power of the electric grid is a bit of a fantasy – but that’s a subject for another discussion…

  • Kevin Kelly


    I basically agree with you. Some, but not all, low-tech is needed to get to high tech, and some, but not all, low tech will underpin high tech once you get there. What we don’t know is what are “candles” and what’s “cement.” Are land phones dispensible or essential?

    I very unfashionably do believe some technologies are inevitable, in the sense that they are a narrow bottleneck that any civilization — at least one based on earth — must go through. In this way I think electricity on wires and the web — for two examples — are inevitable inventions and that they must come in the order they did.

  • Manish

    Kevin, excellent post, but I disagree with the statement with your “I don’t think you can do cell phones until you do land phones.” There is a significant population in India (and I would talk from Indian perspective) that is ‘doing’ cellphone without being trained to do a land phone. And I mean ‘untrained’ and illiterate population, not someone trained to the ways of modern society. Because for such a person a cellphone is another tool, even if he/she hasn’t used a land phone.

    But coming back to the bigger point, yes indeed it is true that old infrastructure would be needed as a foundation for new one. In case of telecom infra in India, I can say that there are regions of India where a landline never reached, and now with mobile coverage, would not do so in future (let’s leave the internet argument aside, for a while). But that is only specific to some regions of the country/society. If you look at India as a whole, the mobile telecom infra is built on the physical and knowledge base of the traditional infra.

    And the ‘knowledge base’ part of the argument is more important, because in India mobile telecom has grown exponentially due to the already present expertise and understanding of telecom in general. The early adopters, developers and promoters of mobile telecom came from the traditional telecom. They are still the ones (both in terms of institutions and personnel) that control, guide and lead the way for mobile telecom growth.

    And a similar logic applies in the internet infra growth as well. Just replace mobile telecom with wireless internet infra (WiMAX etc.) and we see the same arguments, the same growth patterns evolving in this area as well. I am sure a majority of broadband consumers in India (even in big cities) would have their first internet connection via wireless advanced technologies.

  • Kevin Kelly

    Manish, I agree. My error in language. When I said “you can’t do cell phones unless you do land lines” I did not mean individuals, who as you correctly note, have no trouble at all leapfrogging. By “you” I meant a society’s infrastructure.

    Individual users have an uncanny ability to leapfrog over centuries of development. From the jungle to the web without blinking.