Eric Brewer is Professor of Computer Science at the University of California Berkeley where he leads the Technology and Infrastructure for Emerging Regions (TIER) research group.

Eric spoke about the role for technology in effective development strategies at the base of the pyramid.

The history of development to date has been characterized by large agencies funding big projects with strings attached, usually in the form of debt or demands for political allegiance. These kinds of projects are hampered by their scale and the requirement to work with corrupt governments. They typically include little role for new technology as projects move slowly and lack the expertise to facilitate this.

Outside the sphere of traditional development, technology is having a major impact on economic prosperity. The mobile phone revolution, driven by bottom up demand, provides enormous advantages to any worker operating in a large radius. A taxi driver given a mobile phone, for example, will increase his revenue by 60% on average. Other bottom-up businesses have seen major success. The Village Phone scheme, which runs as a franchise model with capital coming from microfinance, now covers the majority of Bangladeshi villages. A village phone lady will make on average two times the income she would have done from farming.

However, the mobile phone remains a largely urban phenomenon since cellular networks require a certain density of users before they can economically justify the installation of a base station. The availability of an internet connection is crucial for the viability of businesses and services in rural areas.  WiFi-based Long Distance networks (WiLDNet) are emerging as a potential low-cost alternative to traditional connectivity solutions for rural regions. Unlike mesh networks, which use omni-directional antennas to cater to short ranges, WiLD networks are comprised of point-to-point wireless links that use directional antennas with line of sight over long distances.

Eric's Berkeley research team has partnered with Aravind Eye Hospital in Theni in the southern India state of Tamil Nadu to use this technology to address the problem of blindness in the region, 70% of which is treatable. The long-distance wireless network they have installed is allowing eye specialists to interview and examine patients in five remote clinics via high-quality video conferencing. 25,000 patients have recovered sight using this system and it is set to expand to 50 centers covering 2.5 million people.

Eric's team has also worked on software that addresses local educational needs in developing regions. In poorly resourced schools, students will often be sharing a mouse and computer screen with a group of others. Metamouse gives each student their own mouse to use; when answering questions all users must agree on a location before progressing. This encourages collaboration between students and has had impressive results in boys in particular, with a 50% improvement in scoring compared to each user having their own PC.

Barbara van Schewick, Assistant Professor at the Stanford Law School, introduced the current debate about net neutrality and explored the implications for diversity and freedom of expression online.

Network providers were at one time ‘application blind' - they were unable to see what was contained in the data packets that allow information to be transmitted online. Now that this is no longer the case, a debate has emerged about the role for regulation in controlling the ability of network providers to block or interfere with applications. What was drawn up as a voluntary policy statement is now being considered and revised by the FCC's Open Internet Proceeding.

Blocking of applications is problematic on several counts. First, there may be incentives for network providers to block applications that threaten their own profitability (for example, Skype). This leads to a situation where the success of applications is no longer decided on user criteria and the overall value created for society diminishes. Second, the great promise of the internet is that it removes traditional gatekeepers (such as mass media outlets) to speech. This is undermined if network providers have the ability to control what content users see. This is particularly problematic since users cannot easily switch to another provider as they could if a particular store did not carry a product they wanted. The cost of switching makes this impractical and in places without a choice of providers, this is not an option.

In drawing up regulation against blocking the FCC is debating a number of related issues:

Discrimination: Even if blocking is prohibited, discriminating between levels of service can still allow network providers to slow down an application to the extent that it becomes un-useable. This is actually a more effective tool than blocking since it is much harder to detect. Users may attribute slow speeds to poor design and potentially useful applications will fail to get traction.                                                                                                                                                                                                                                                                                                                                                                                                                              

Charges for different levels of service: Even if we agree network providers should not discriminate between the services they provide in an arbitrary way, could they offer improved service for payment? Opponents argue that this policy would be bad for competition since new developers would be unable to pay for the levels of service that established players could afford. And it would threaten the ability of poorly resourced minority voices - e.g. small NGOs and publications - to get heard.

Exceptions to discrimination: Network providers argue that there needs to be some discrimination to allow them to undertake reasonable network management. But it is difficult to determine what counts as reasonable management. One concern is that peer to peer networks - which allow those without many resources to exchange material cheaply - might be targeted in particular, since they can create a lot of congestion. This might also threaten the ability of new applications with high bandwidths to get funding, since the risk of being slowed down by the networks would be perceived to be too high by investors.

Many of the major benefits of the internet - the ease of publishing and coordinating, for example - are only possible through applications. Hence the outcome of this debate will have serious implications for the future social and political impact of the internet.