Posts Tagged ‘generativity’

There is an enormous wave of interest in the role that mobile phones play in development. The widespread access to increasingly sophisticated mobile phones have opened up an information and communication platform that is improving, among other goals, livelihood, health and stability. Two new articles remind us that they come with a heavy price.

First, Ethan Zuckerman writes about the implications for innovation and activism that result due to centralization:

Creating novel functionality on a mobile phone network is much harder [than on the Internet]. Truly revolutionary applications like mobile money transfer have generally been deployed in tight collaboration with network operators – M-PESA was not an independent startup, but an initiative of Vodaphone/Safaricom with support from IBM and DFID. It is unclear whether Safaricom would permit a rival mobile banking system to develop expanded functionality and deploy on the same network.

Mobile applications in the developing world generally focus on providing services via short message services (SMS). This is due in part to the need to provide services on a wide range of devices, and in part to the comparative ease of deploying SMS gateways without cooperation from network operators. Voice-based services (IVR – interactive voice response) would often be a better technology for the needs of low-literacy users, but it’s difficult to deploy at scale without co-locating equipment with network operators… Unlike with the Internet’s decentralized DNS system, assignment of shortcodes is generally centrally controlled, giving operators control over the promotion of platforms by refusing to issue easy-to-remember codes. (Imagine if Skype had needed permission from AT&T or France Telecom to register skype.com.)

Because mobile phone networks are centralized, they are more easily controlled by governments than the Internet. Filtering and censoring the Internet has proved a frustrating cat and mouse game for both governments and activists. Despite millions of dollars spent to filter the Chinese internet, hundreds of thousands of Chinese users access and publish banned content. By contrast, Ethiopia simply turned off SMS services in June 2005 over fears that students were using the technology to organize protests against rigged elections – and services remained turned off for more than two years.

Secondly, Yochai Benkler notes that the expense of traditional computers puts them out of the reach of individuals in the developing world, thus limiting the radical redistribution of capital that has occurred in the West. Although mobile phones are affordable and increasingly powerful, Benkler reckons they may be a Faustian bargain that “comes at the expense of a truly open, neutral network.”

As we think of ICTs for development, we must understand that the challenge is a focus on widespread distribution of high-capacity devices, in the hands of a highly skilled population, over open networks running simple and non-proprietary standards.  Devices must be cheap enough to be widely distributed as basic background features, owned by individuals in a pattern uncorrelated with pre-existing power relations.  Devices must be accompanied with skills training in the use of the device and the open network, so that the difficulty of use does not continue to drive people to the simpler devices that deliver the more predictable, controlled, and “safe” applications.  In the near future, this may mean programs focused on women, much as micro-lending has been, or youths and children.  In the longer term, it must mean an emphasis on cheap computers from the lineage of the personal computer, not souped-up mobile phones.  Or, in the alternative, it means that we need a heavier focus on regulatory interventions that will require mobile phones and phone networks to be more open and flexible—although this is a harder row to hoe.  And in all events it means devices coupled with training.

The networked information economy and society promises a radical shift in power and capabilities from industrial, centralized forms to decentralized forms that counterbalance market dynamics more effectively with social dynamics.  To achieve this, a highly distributed physical and human capital structure is necessary.  Understanding this requires that our focus on ICT for development should be on achieving the radical, decentralized distribution of flexible, open, physical capital throughout the population, coupled with the necessary training to harness the wisdom, insight, and creativity that is already there.

This is not to say that mobiles for development should be sidelined. I believe quite the opposite, in fact. But these reminders of the properties and structure of the mobile phone ecosystem should weigh heavily on those working in ICT4D.

[P.S. These are from a new collection of articles posted by Berkman in preparation for the high-level Sept. 23rd discussion on communication and human development.]

Update: Steve Song has posted his recent presentation on the Village Telco project that seeks to create a better network.

In recent weeks, the iPhone has made quite a stir because of the regulatory decisions made by Apple. Jonathan Zittrain raised this worry in his book, The Future of the Internet, where he cautioned that generativity – the nature of systems to accept input from everyone – was being traded for sterile appliances – devices which do only simple tasks (GPS, TiVo).

The iPhone has led a new way, called contingent generativity, that makes generativity dependent upon an intermediary. Apple gets to decide whose Apps are available for download and though Steve Jobs had claimed that they would only block apps that were malicious, pornographic, bandwidth hogs, illegal or threats to privacy, that hasn’t proven true in practice. As I noted at Techdirt, Apple is becoming a Soviet ministry price-setting intermediary that decided the “I Am Rich” application wasn’t allowable even though it didn’t seem to break any rules. “I Am Rich” isn’t alone; other apps which provide additional functionality have been pulled with little to no explanation.

But being an ex-ante regulator isn’t enough. Apple, which is famously closed in character, also has the ability to regulate apps already on a user’s iPhone or iPod Touch. The so-called kill switch was not disclosed to the public until a curious user uncovered the capability. Only then did Steve Jobs admit the functionality existed, saying Apple needed the capability but “Hopefully we never have to pull that lever, but we would be irresponsible not to have a lever like that to pull.”

This position raises a number of questions, many well articulated around the web, not the least of which is why Apple thinks it needs a kill switch an the iPhone and not it’s Mac computers. The issues raised and trend shown by the iPhone’s kill switch is worrying and, as you might expect, some clever engineers have found a way to disable it for jailbroken iPhones, but a thread on the Free Culture mailing list got me wondering if there was a better way to solve this conundrum.

I think there is and I think it should draw on the scholarship of Cass Sunstein and Richard Thaler in their book I recently reviewed, Nudge. For the uninitiated, Nudge is a book about “libertarian paternalism” which aims to create situations where it is easier to make the best choice while not limiting other options. Through architecting designs that enable better decisions, or nudges, libertarian paternalism provides a middle ground between freedom and mandates.

Apple has the opportunity to do so with the iPhone kill switch. The intentions of the regulatory function are good: many users are, for whatever reason, unable to avoid or fix security compromises. Apple has experts who can help these users, but a mandatory kill switch is not the best option. It treats all users the same and removes their ability to run applications they desire, regardless of potential hazards. Asheesh Laroia suggested that Apple allow users to permanently opt-out of the system.

I would go one step further towards openness and make the kill switch an opt-in feature. Call it AppleCare Pro for iPhone or something less awkward. Heck, Apple could even charge for it! Make it a prominent decision in the set-up process and allow users to revisit the option when they desire. Provide nudges towards it when the user downloads an App which might be dangerous (similar to how Google warns searchers they may be entering a nasty page).

This would give the worried or non-experts the ability to have Apple’s paternalistic reach extend to their phones without compromising the autonomy of those who want independence. Parker Higgins worries that those who need Apple’s protection are those likely to ignore the warnings, but I think Apple could architect a system where they are nudged towards better decision-making without a presumption of ignorance.

In doing this all, Apple should remain aware that openness and honesty is the best option. The fact that they hid the kill switch until outsiders found it is reminiscent of Comcast’s deceptive practices regarding BitTorrent throttling.

One of the questions I’ve been kicking around since getting back from Berkman@10 has been one raised by Jonathan Zittrain’s book, The Future of the Internet. As I noted back when I reviewed it, JZ’s premise is that increasingly, vendors are selling locked-down, sterile devices to willing consumers who are fearful of the negative effects of generativity – spam, spyware, viruses.

In parallel with this shift towards sterility has been a push-back by those with the technical skills to hack their devices for additional functionality. With Linux or Wikipedia, the user is encouraged to hack, edit and create. With the iPhone or TiVo, the opposite is true and we are expected to accept what Steve Jobs or another executive decides. This has not been accepted happily by all and many are unlocking their iPhones or other sterile devices.

This “forced generativity” might seem like a protection against sterility – no device can be fully locked down and users are bound to open them incrementally. When I asked JZ this during a Washington Post discussion, his answer pointed out an important qualification: this forced generativity is increasingly contingent upon avoiding centrally controlled updates which can re-sterilize freed iPhones or TiVos.

Zittrain also laments that not everyone has the technical skill to force generativity, saying “I don’t want a world where only the hackers get Get Out of Jail Free cards, and where everyone else risks serious crossfire to break out of a sterile platform.” I think this is an important point, but not necessarily the entire truth. Of course, technically-inclined folks are more likely to be able to avoid sterility, but aren’t they the people who can take advantage of generativity in the first place? Hackers are the ones adding to Linux and using the neutral Internet to create new web services.

However, anecdotal evidence would suggest that more than just hackers force sterility. For example, Jan Chipchase (NYT profile), a cultural anthropologist who studies the use of mobile phones in the developing world, has written about “cultures of repair.” If you travel around the developing world (and to a lesser extent, richer nations), you will encounter a massive industry dedicated to mobile phone augmentation. Chipchase writes, “Aside from the scale of what’s on sale there is a thriving market for device repair services ranging from swapping out components to re-soldering circuit boards to reflashing phones in a language of your choice , naturally.” These are not just repairs; at times they come very close to doing what American hackers do with the iPhone – they add functionality and opportunity. And it isn’t just the technically minded, its the poorest of the poor:

“The informal repair services that are offered are quite simply driven by necessity – highly price sensitive customers cannot afford to go through more expensive official customer care centers and even if they could their phones are unlikely to be covered by warrantee – having been bought through grey market channels, been sent as gifts from friends and relatives abroad, or were locally bought used, second or third+ ownership.”

What is happening is a broad effort to force generativity upon an industry which is woefully sterile. The warrantees, contracts, networks and devices of the mobile telecommunications sector reek of sterility. People, and not just hackers, are pushing back.

Perhaps the real conclusion from this small quibble is that people do want and are capable of utilizing generativity. Efforts to control too much will be rejected or augmented.

Jonathan Zittrain, the co-founder of the Berkman Center and professor at Oxford University, has a new book entitled “The Future of the Internet – And How to Stop It” in which he solidifies himself as one of the leading thinkers in Internet studies.

JZ provides a simple framework for understanding digital technologies whose implications are profound. For JZ, what make the Internet (and to a lesser extend, other technologies) great is its ‘generative’ nature. “Generativity is a system’s capacity to produce unanticipated change through unfiltered contributions from broad and varied audiences.” The generative nature of the net has enabled a whole new era of innovation spawning Google, Wikipedia, YouTube and more. However, with these welcome advances has come another effect of generativity – badware. The generative state of the net also allows malicious exploits like spam, viruses and spyware.

These massive enterprises pose an existential threat to the Internet as it has functioned and prospered thus far. As a result, a new form of digital devices and networks have arisen. JZ calls these sterile and TiVo, most cell phones and other locked down devices are prime examples. A middle ground is exemplified by the iPhone whose SDK provides a form of “contingent generativity” – you can generate new innovations as long as Steve Jobs approves. The rise of intermediaries is something about which I have written before.

What Zittrain has done in The Future of the Internet is fire a warning shot past the bow of all the tech geeks who relish their TiVos and iPhones. He openly admits that these devices are wonderful accomplishments of technology and design, but the book forces the reader to come to grips with the ideological and practical implications of these “safe devices.”

The implications are manifold and often unwelcome. Sterile devices tend to make innovation more difficult as freedom is limited. Wikipedia would not have taken off and have millions of articles in a model of control. In fact, its predecessor, Nupedia, was a failed attempt at controlled encyclopedia creation.

Contingency provides government or other would-be oppressors an easy means to surveil or censor. Through his work with the OpenNet Initiative JZ has studied the dozens of nations who actively censor the Internet, in effect making the generative network a sterile one. Similarly, the shift to cloud computing or software as a service means that data centers are controlling huge amounts of other people’s businesses and have the ability to stop things they would prefer not occur or are more easily open to regulations which may stifle creativity.

The obvious and easy answer that governments, corporations and users are embracing to avoid the many negative effects of generativity (badware) is the shift to sterility. However, JZ thinks a more norm-based approach will be able to save the benefits of generativity while vastly limiting the downsides. A prime example is robots.txt, an optional but widely accepted standard which allows people to not be included in search engines. Google and Yahoo have no compulsion to follow webmasters’ requests, but they do. Likewise, Wikipedians by and large seek consensus and a neutral point of view, even though they are free to not do so. JZ’s plea is for “netizens” to vote with their processors and bandwidth for solutions which embrace the ethos of generativity. The Berkman Center’s StopBadware.org is an example and has worked with Google to stop people visiting websites known for disseminating dangerous code.