Dear Reader,
Artificial Intelligence and IT overload are our key themes this week. First we look at whether artificial intelligence, such as speech recognition, makes sense in achieving development goals.
Then we examine whether Africa's IT focus, believed by some to be to the detriment of other development needs, is equally problematic in the sphere of MFS.
Mondato is at the convergence of the public and private sectors where financial, mobile, and online payments and value transfers (money, airtime, bill payments) meet.
Got Questions?
Mondato's custom research department will look to answers you can use in planning, revamping or otherwise addressing business questions related to remittance, mobile payments, and mobile transfers of value. Wherever your geographical interest, reach out to us to learn more about how we can help: info@mondato.com.
Suggestions Welcome
We constantly strive to ensure our content covers the depth and breadth of the mobile financial services industry. Have a good idea for us to follow up on? Let us know at contact@mondato.com. Chances are we may even know a perfect industry solution or partner.
|
In this issue:
Siri and BoP: Artificial Intelligence and Development
Despite what might come first to mind in associating Artificial Intelligence (AI) and developing countries, there are, in fact, movements by such organizations as AI-D[1] to incorporate the best practices from this field into developing country contexts. For example, AI-D has done research on mobile crowdsourcing, using expectation-maximization to develop “a system that enables people to earn small amounts of money by completing simple tasks on their phones.” In other research that aims to analyze longitudinal data on 15m mobile users across East Africa, they hope to learn what impact human mobility plays in the spread of malaria.
In an article about one facet of AI – speech recognition – the authors argue that an open source technology approach should prevail so that more services can be developed and spread, and places like South Africa, with eleven official languages, can implement beneficial and ultimately lucrative applications.[2] All told, this then begs the question of what, when and where might AI best be brought into mobile payments.
It is fairly easy to see how speech recognition in a variety of forms – speech to text, voice recognition/mapping (e.g., for security), etc. – could enhance a mobile payments app. The question of user interfaces for mobile financial services applications on the handset has been raised from time to time in the industry. And while some may argue that a single “dominant” language (e.g., English, French, etc.) will suffice for the limited commands that are needed in a given locale, advances in speech recognition mean that the options may be there to more broadly embrace many languages for greater adoption of a given MFS in which they are embedded. For the Bottom of the Pyramid where literacy is a concern, being able to operate in a known language orally would seem to open the scale of adoption for a MFS much wider.
Safety in Numbers
Yet the immediate future of AI in mobile payments may be in the less visible, back-office security side. As more is done electronically via mobile phones, threats to those transactions are estimated to rise, and the fight to thwart them will necessarily grow in importance.
Although not as sexy perhaps as the user interface side of an m-payment app, as the tide shifts to more of these transactions, the ‘bad-guys’ will inevitably gain greater interest. Efforts to reduce or eliminate fraud, ID theft, or other security risks will be all the more valued once the first round of well-publicized security breaches cuts into the momentum. Preventing them may seem secondary at the moment, but will rapidly be at the core of assuring trust to those users of MFS that go to any significant scale. Particularly since consumers are concerned about such risks.[3]
Where banking and financial transaction ‘literacy’ is low, the risk of fraud is perhaps even higher. AI applications in developing market contexts, or geared to the unbanked, will need to be even more attentive to ensuring simple but accurate security safety mechanisms are in place.
A Mind as a Sieve
On a different AI angle, those not ready for the possibility of their every thought being revealed may feel lucky that we are reportedly at least 10 years away from this advancement from Intel, who are reportedly: “developing a technology that could be built into a wireless hat and work as a ‘mental typewriter.’ By scanning brain waves, it can guess what a person is thinking and predict search queries. A prototype of the software can recognize about 60 words, such as "airplane" and "celery," says Dan Pomerleau, a researcher at Intel Labs in Pittsburgh, Pa. It could take more than 10 years for its vocabulary and accuracy to improve enough for real-life uses, he says.”[4] Though, then again, one could envision that service as a significant life-changer for someone like Stephen Hawking!
But out of curiosity, it would be interesting to learn what Siri on the iPhone 4s thinks about AI. Someone ask her and let us know.
[1] http://ai-d.org/research.html, also notably AI-D is headed up by Nathan Eagle, founder of txteagle which is now Jana – a mobile crowdsourcing research tool for developed country brands to access developing country opinions.
IT & Mobile Payments: Are the Africans Putting the Cart Before the Horse?
Blogger Ndubuisi Ekekwe in the Harvard Business Review recently posted a very thought-provoking argument: “The success of IT in Africa has reached a level where it is being dangerously over-emphasized, at the expense of most non-IT technologies.[1] According to his blog post, across most African universities, the only funded and active labs are the IT labs, while everything else is broken. Agricultural engineering students are more focused on IT than on learning to build next-generation farm machinery. On a country level, for example, the Nigerian government has created a new ministry to focus solely on IT and related areas, but despite decades of crude oil exploration, it cannot claim that it has developed indigenous domain expertise in that industry, lacking the local ability to explore, extract and sustain production.
Such concerns have been echoed by others in the development community, such as Howard Buffet, son of billionaire investor Warren, who is a strong critique of the Gates Foundation’s approach to agriculture in Africa.[2] For him, the US-style of industrial agricultural techniques being implemented by the Gates Foundation, including the use high-tech seeds and pricey synthetic fertilizers, is inappropriate in regards to Africa’s true agricultural needs, which are low-tech, inexpensive methods for increasing farm productivity.
From the perspective of mobile payment, one could also ask whether the focus on such high tech applications is superfluous, given the inability of most persons to fulfill basic needs such as access to clean water. Are we putting the high-tech cart before the basic horse? We would argue that an emphasis on the development of mobile payment in Africa, although intrinsically connected to IT, is not in danger of neglecting other developmental concerns; in fact, mobile payment technology can directly solve developing country issues, in a way that is more efficient than traditional pathways.
Mobile payment has been espoused as the WAY to integrate the unbanked into the formal banking system, thereby providing them with opportunities to create savings, obtain loans and start businesses. And as highlighted in previous newsletters, certain mobile applications are being created specifically to tackle developing country issues, such as finding clean water, assisting women who may have to undertake dangerous commutes to get to and from work each day, and confidential AIDS counseling. (See Mondato Newsletter Jan 4, 2012, Vol. 5, Issue 1 (Past and Future Predictions + Apps for Developing Countries).)
Mobile payment also does not necessary mean a high tech solution – as highlighted in our previous article on leapfrogging technologies. A case in point is ForgetMeNot Africa, a UK company that markets a technology that allows SMSs to be delivered as emails, instant messages and/or Facebook messages and vice versa from any mobile phone, rendering the need for sophisticated smartphones with internet capability obsolete for such capabilities and thus facilitating basic money transfers.[3] (See Mondato Newsletter July 13, 2011, Vol. 4, Issue 14 (Bypassing NFC + Security in Mobile Payments).)
And even if the technology that is used to facilitate the mobile payment is “high tech” – such as the use of speech recognition technology to allow illiterate people the ability to make mobile payments (see today’s article on artificial intelligence and Development) – this is an example where high tech is truly being used to address specific developing country issues, rather than part of a bulldozing approach to development issues.
Given that a high proportion of the African population already owns mobile phones and therefore the primary investment in “infrastructure” has already been made, further progress and focus on mobile payment does not come at the expense of other areas that require funding. The energy and attention devoted to mobile payments is time well spent, given the returns such work has achieved and is continuing to achieve in terms of the meeting of development goals. African countries may be too focused on IT to the detriment of other development goals, but in terms of mobile payment, that argument just does not hold.
|
