The Technium

Two Kinds of Generativity

[Translations: Japanese]

The nature of every invention is to start out vague, incomplete, and open to change. A new invention or device is born primed for hacking, and for re-definition. The new half-baked thing is many things to many people. At this stage, the device is in the hands of tinkerers, nerds, fans, and hacks who will make it do all kinds of things no one had thought of. This skeletal generality enables generativity — the ability to generate new goods and services and ideas — because the tool itself is being invented, which is the highest kind of genesis. In this mode a newborn device or invention or idea is thrilling. Its naked open-ended potential appeals to a set of early adopters (who are often called an elite) who explore the glorious incompleteness of a new tool in many ways.

But for many others (who are often called the masses), this very openness, this ill-defined thing requires too much expertise, or control, or knowledge, or care, or time, to use productively. To them its skeletal generality is a turn off. The fact that the invention is incomplete and has not discovered what it wants to be is a negative. One has only to think of the early days of automobiles, or windmills, or radios, or the internet, or cell phones, and how unappealing most people found their incompleteness (“they hardly work and are hard to use”).

But the masses usually don’t have to wait long for the natural history of an invention to kick in. A device become more specialized and “complete” as it evolves. As it does, it becomes more specific in what it does, more closed in its identity, more clear in what it is. It becomes more powerful in evolving its identity. As it matures it becomes more completed, more approachable, more understandable, more able to do things for more people.

We see this in all walks of technology. For instance the first cameras gave great latitude to the first nerdy photographers. Since you had to make your own film, you could create it do all kinds of things — favor the infrared spectrum, or embed it in fabric, or make it three feet wide. The hacker-photographer had great latitude. But as the outlines of popular photography became clearer, the camera honed in one a certain specific design: film was manufactured to an emerging standard, and equipment more certain what it was for. The result were cameras that anyone, not just photo geeks, could use, and the result was an incredible generativity as millions and eventually billions of people started photographing EVERYTHING. As the craft became more specific, it became more ubiquitous, and the levels of creativity unleashed by this easy to use device vastly exceeded the amazing creativity of its founders.

Almost every technology follows this lifecycle, starting out with fantastic hackability at first and maturing into dependability. Many of the qualities that early adopters love — the way it can be modified, tweaked, owned, and directed into all manner of directions — and its unlimited potential — are also the same reasons why many others shy from trying to use it. In the same way, the very hardening and convergence that draw the masses to a technology also turn off those who like its earlier kind of generativity, when things were lose.

We’ve seen this shift happen many times in the past, such as in early personal computers when 12=year olds in their bedrooms were writing programs (hurray!), that often did not work well (boo), but later became huge projects that took hundreds of folks to code (boo), but that actually were easy to use and worked all the time (yeah!). Then the web came along, and it was newborn enough that it was easy for kids to make stuff, but later standards and expectations evolved and it took a lot more to keep folks returning. Then apps came a long and anyone could play, but sooner or later, these too will mature and not be so hackable. Ditto for genetics, robotics, and so on.

Occasionally these mature products enable entirely new inventions. These new inventions are again open to the first kind of generativity, and the cycle of technology continues. Naturally hacker and nerds and tinkerers flock to the immature zone, where they can help define the new new thing.


So, the same story is told over and over again. Once upon a time the early adopters made their own electrical parts — capacitors, resistors, crystals, etc. — from which they cobbled together radios and other equipment. But once it was clear what a crystal and capacitor were, the frontier moved on to making the radio — but not with the old guard complaining about the loss of the joy of making your capacitors. Then as radios became more defined, hams did not make radios, they bought them in hundreds of varieties. So they began making their own Altair computers. For a while. The same migration ensued. What you don’t make your own computer? No, but I wrote the operating system. What you don’t write your own OS? No, but I wrote my own program. What you don’t write your own programs? No, but I code my own website. What you don’t code your own website? No, but I write my own apps. What you don’t write your own apps? No, but I weave my own lifestreams…..

There is a natural arc by which each invention moves from the generative openness in a new-born to the refined generativity of a well defined idea. Some folks mistakenly believe that the modern regime of manufacturing and consumerism inevitably closes off all cool inventions to the first kind of generativity, but this maturity has always happened, long before the industrial age. Technology’s natural cycle is merely being accelerated now.

New-borns with infinite potential but low-productivity become middle-agers generating great productivity and unleashing fantastic creativity; in turn the mature keep the frontiers expanding by generating more new borns. I speak here, of course, of ideas and devices.

Each new unformed, hackable, potential invention is quickly refined by use, and this use makes a technology more specific, more conditional, more open to use by know-nothings. Therefore each technology eventually becomes less malleable, less powerful in undefined ways but more powerful in defined ways. It moves from the margins to the center.

The good news is that there will always be a margin where things are uncertain, ill-defined, and totally open, and even better news for hackers: the margin is constantly expanding relative to the center.

Update: Here’s a confirming bit from Steven Levy’s article on Hackers Revisted in Wired:

Unlike the original hackers, Zuckerberg’s generation didn’t have to start from scratch to get control of their machines. “I never wanted to take apart my computer,” he says. As a budding hacker in the late ’90s, Zuckerberg tinkered with the higher-level languages, allowing him to concentrate on systems rather than machines.

For instance, when he played with his beloved Teenage Mutant Ninja Turtles, Zuckerberg wouldn’t act out wars with them, like most other kids. He would build societies and pretend the Turtles were interacting with one another. “I was just interested in how systems work,” he says. Similarly, when he began playing with computers, he didn’t hack motherboards or telephones but entire communities — manipulating system bugs to kick his friends off AOL Instant Messenger, for instance.

Big business may stumble upon and commodify their breakthroughs, but hackers will simply move on to unexplored frontiers. “It’s like that line in Last Tango in Paris,” O’Reilly says, “where Marlon Brando says, ‘It’s over, and then it begins again.’”

  • Nick Carr

    Human beings have always been tool-makers, and recently in their history they have also become profit-seekers. Both these characteristics generate the kind of technological progression you describe, ie, the refinement of tools through which the tools become more specialized and more broadly used – and hence more broadly sold. (The essential force behind this progression is human nature; it’s not any kind of force inherent in tools themselves.) I think, in order to provide a full description of this process, you need to also bring in an account of the tool-producer’s (or, more accurately these days, the tool-marketer’s) economic motivations. By reducing what you call the first kind of generativity, the marketer transforms the user of a device into a consumer of the device and thereby increases the consumer’s dependency on the marketer’s products and services. The marketer doesn’t want the consumer to make his own film (to draw on one of your examples); the marketer wants the consumer to purchase readymade film, thus opening up a new stream of profits. Similarly, the marketer doesn’t want the consumer to repair or otherwise modify a computer; the marketer wants to be responsible for all repairs and modifications (and to reap the profits that flow from those services). In the end, the marketer contrives to make the tool so self-contained, and impervious to repair, that it becomes disposable, which serves to encourage the fast repetition of purchases. (Apple has become very adept at this; it’s one of the reasons it doesn’t want its customers to have easy access to its devices’ batteries.) So while you’re right that the basic pattern of tool progression that you describe has always been around, what’s happened recently is more than just an acceleration of this process. There are also strong new market forces at work.

    • Kevin Kelly

      @Nick Carr: Yes, marketplace dynamics are in sync with this cycle. Money can be made selling more refined things, and to your point, there is a natural lockin of the consumer; the more consolidated the technology becomes, the more dependent the consumer becomes on the closed, or defined, technology. Once the tech is lockedin, the technology becomes “successful” but also subsumed, nearly invisible, almost unchangeable (think cars). Radical innovation moves elsewhere.

  • Nick Skelton

    Did you write this in response to Cory Doctorow and others who protest about the iPad, seeing it as a closed ungenerative device?

    I think you explain why such concerns are misplaced (or perhaps why they are correct but not very important).

  • Chris Warner-Carey

    I wonder if there is an analogy with philosophies, ideas, religions and how they develop. Replace the word “invention” in the first paragraph with “religion”. It seems that religions(I’m thinking of the three Abrahamic faiths)develop from the general (broad,nearly universal teachings)to the specific (dogma, “one true religion”), from a certain level of ambiguity to concrete rules and proscribed actions and beliefs(operating systems). In my experience (disclosure-I’m a pastor)most people, especially when they are young, and/or new to a faith are uncomfortable with too much openness, too much ambiguity–they want and need concrete operating instructions, because life is complicated. Like new technologies and new devices, which are seen as too open, there is a drive to consolidate new ideas and teachings as well. A wild leap here:perhaps the rise of fundamentalisms of all kinds is, in part,a reaction to the growing complexity of the world, and a desire for certainty. “I just want my computer to work. I just want the universe to make sense!”

  • Kent Schnake

    Excellent post. I’ll be 60 in August. The advantage of a few decades under the belt is that I have seen some ugly duckling new “inventions” gradually refined into ultra cool products. An example that comes to mind is the Osborne Portable Computer. Pretty clunky and puny. Back then I could hardly imagine netbooks, iPads, etc. etc. Hopefully I have learned a bit of patience and jettisoned a bit of cynicism about new products.

    Thanks Kevin.

  • Jeremy Foote

    Great post (as usual).

    It’s interesting that the information about a technology is stored more and more in the tool, and less and less in our brains.

    In the early days of photography, many people understood the principles of exposure, as well as the effects of alternative chemicals, alternative film mediums, etc. Even though photography has become ubiquitous, I think that there are probably fewer people who understand this level of photography. All of the knowledge about the chemicals, etc. is stored in the technology itself.

    This transfer of knowledge is not lossless, however. As these technologies become abstracted, and become invisible, we are empowered to do more with these tools, but we also lose the ability to alter, and maybe even to truly understand, the underlying decisions about why they are the way they are.