Long time no post! I need a good opportunity to dedicate the time and the energy. I got interviewed last week by the New Scientist about Google‘s Loon project, and I thought it would be a great topic to talk about. First of all, I’m not an expert in this technology, I don’t know if balloons can be more efficient than satellite, mobile tower or sea cable. However, I’m sure that Google has made extensive investigations before announcing anything, and I’m also sure that they are able to mobilize all the required expertise. So all in one, even if I don’t know anything about the technology, I’m sure this is technically perfectly ok. My point is somewhere else, and is related to impact of Internet connectivity on development.
The fact is that today two third of the World population is not connected to the Web. I believe that there is now a general agreement, at least there are evidences, that some specific ICT services in some specific places (e.g. mPesa in Kenya, the usual example!!) can improves people’s lives in underprivileged communities. The big question is to understand how to make more relevant services available to people to have a global effect, and increase social and economic opportunities for those at the base of the pyramid.
From this shared, imho consensual analysis, the design of solutions leads to very different approaches! I tend to think that there are two schools here: those who focus on connectivity (aka supply side) and those who focus on content, more exactly locally relevant content (aka demand side).
In the first category, the thinking is that connectivity and bandwidth is the bottleneck. There are in this category lots of different initiatives: some believe that the issue is the availability of connectivity (e.g. loon), some believe that the issue is the cost of the connectivity (e.g. Alliance for Affordable Internet), some believe that the issue is the size of the bandwidth (UN broadband commission). In all these cases, the point is still that the issue is the pipes and not in what is in the pipes. I believe the idea behind this approach is that the Web is huge, it has trillions of information, and thus just providing access to this huge amount of knowledge will resolve all development problems.
I personally don’t subscribe to this point of view, and to the myth of a farmer in Malawi googling to find a solution to the disease of his crops. Based on my 7+ years of experience in the field in countries that range from Kenya to Mozambique to Mali to Ghana to India, I’m convinced that before a farmer could “google” the World, there are huge barriers to solve. And then even if he can eventually google, the information he would find is very unlikely going to be relevant to his context.
In terms of barriers, in my experience, there are four major challenges: ICT expertise and experience, illiteracy, language, devices.
Let’s start with devices. It is great to focus on connectivity, but connectivity is meaningless if one doesn’t have a device that can connect to the available network. If you don’t have a data-enabled phone, 3G is useless. If you have a 3G non-WiFi phone, a WiFi network is useless, etc. Despite all forecasts and predictions, the situation in the field is pretty clear: the vast majority of phones in rural areas are basic phones that have no data capacity. In practice, that means that independently of the environment (if there is or not connectivity), from people’s perspective there is no IP network. It is pretty obvious that to improve people’s lives you have to build solutions that fits with what is already existing (see the recent interview by Ken Banks about ”10 Tips for Successful ICT4D Interventions” where the first one is “Build for what people already have in their hands”.). Of course, it is easy to counter this argument and say that people didn’t buy phones before mobile networks arrived. This then leads to my next points related to other access & content challenges.
Indeed, devices are just the visible part of the iceberg. With mobile phones, the barrier was relatively low. It is relatively easy to teach people how to dial a number. The complexity of ICT applications is another order of magnitude. There are at least three major barriers: illiteracy, language and training. Illiteracy is still a major challenge in many countries. More than 1 billion adults are illiterate with rates above 50% in countries like Mali. How can you enjoy the Web and access content and applications if you are illiterate? Of course if you are in an urban context, proximal literacy (using an intermediary who can read and write like e.g. your son reading your SMS) is an option, but this is hardly workable in a rural context, and affecting more women. From my perspective, it is essential to develop new technologies and interaction paradigms to empower those with low reading skills. As I mentioned in a previous post, voice technology is the only option today. I’m very proud to have led one of the biggest research project in the domain (see EU-funded VOICES project), and it is clear that practitioners are giving a growing attention to this technology. Unfortunately, there are lots of trials that are leading to failure because of lack of best practices, and guidelines, in particular related to audio interface design and business models. Even if voice technologies can be usable be illiterate people, it is essential to design interface so that they can use them. My usual failure example is e.g. asking people to “press 1” when they are unable to associate the sound “ONE” to the pictogram of the phone keypad “1”. Similar examples include the use of multiple levels of menus with many choices in each. Those failure examples are still blossoming all over the World and it is essential not to through the technology away for bad reasons (throw the baby out with the bathwater as we say in French!) . It would be essential to have more research and more publication of guidelines and best practices in this area to exploit better the potential of this technology.
A part from voice technology, it would also be interesting to start looking at graphical interface. People (particularly those at the base of the pyramid) will invest in new more powerful devices if they can see values in it, and services that are usable and useful. All the smartphones and touch-base interfaces are very “natural” for lots of people, but those who have ever tried to show such interfaces to farmers know that they are not at all natural for people without experience. In the same way, even for those who can read and write, all mobile and web applications are heavily using background knowledge of the users for interactions: e.g. a house for home, a right to left arrow for back etc. To the best of my knowledge, there is no research in this area: no research or very very few publication of icon-based or icon+audio based interfaces, no research on new gesture interface adapted to people without experience or who are illiterate, no research about how to teach the background knowledge I mentioned above etc. These topics are for me far more important bottlenecks to a greater update of Web technologies and applications.
Finally, I want to say a couple words about content. What is the proportion of contents developed in e.g. Africa available on the Web? How can we imagine that locally relevant content would come from Europe or US? I’m convinced that services and applications that can make a difference in people’ lives can only come from people living in the country. It is essential to develop capacities, and support local content & service development. For me, it is a bit worrying to see that most of the connectivity-related initiatives are driven (officially or behind the scene) by major content firms like Google, Facebook, Microsoft, etc.
All in one, of course, bandwidth and connectivity is an issue, and we can all experience this when we are visiting Africa, South-east Asia or Latin America. It is essential to continue working towards cheaper more broadband options. But I’m convinced that those aspects are not the major bottlenecks for greater social and economic impacts of the Web, and for a greater usage at large. In a recent past, we saw initiatives investing billions in trying to send computers everywhere with the success we know on development and education. The history also shows that bandwidth and connectivity was not an issue from 1975 till 1990 between the creation of the internet, and the creation of the Web. Then with the rise of Web usage, the bandwidth started to be the bottleneck and started to be increased. My view of this is that usage is the key lever and the key challenge to focus on, instead of connectivity and bandwidth. Obviously it should not be a “or” (connectivity or usage) but an “and”. However, at this point in time, I don’t see many initiatives looking at some of the critical barriers to leverage access and exploitation of usable and locally relevant content. I hope this will change in a near future, and we will see greater investments in service and content, in particular new research on user interfaces.