Thursday, April 3, 2014

Why IBM's Watson Cognitive System Could Be More Successful in Africa than in America

Last month, IBM launched "Project Lucy", a 10-year initiative to bring Watson and other cognitive systems to Africa in a bid to fuel development and spur business opportunities across the world’s fastest growing continent.

Cognitive computing systems learn and interact naturally with people to extend what either humans or machine could do on their own. They help human experts make better decisions by penetrating the complexity of Big Data.

Watson is a cognitive system that became famous with its 2011 win in a contest against two of Jeopardy's greatest champions.  Jeopardy is an American television game show that features a quiz competition in which contestants are presented with general knowledge clues in the form of answers, and must phrase their responses in question form (Wikipedia). For example when proposed with the following clue: “It’s the only American state lying south of the Tropic of Cancer”, the answer should be the question leading to that clue: “what is Hawaii?”.

The Watson win is sometimes compared to another IBM technology victory between IBM’s Deep Blue system and world chess champion Kasparov in a televised chess game, yet the Watson win was quite different. Deep Blue was a "brute force" system programmed to anticipate every possible answer or action needed and that could examine 200 million chess positions every second. Watson instead was trained using artificial intelligence (AI) and machine learning algorithms to sense, predict, infer and, in some ways, think using a large volume of data. Watson had access to 200 million pages of structured and unstructured content consuming four terabytes of disk storage. The sources of information for Watson included encyclopedias, dictionaries, thesauri, newswire articles, and literary works. Simply put, Deep Blue was the math whiz while Watson is an English major.

While the Watson – Jeopardy contest was showcased as human intelligence against machine intelligence for publicity reasons, the purpose of cognitive systems like Watson is really to have machine intelligence assist humans in taking on challenges together and achieving more than either could do on its own. For example Watson enables more natural interaction between physicians, data and electronic medical records through its highly sophisticated question-and-answer technology, and it helps physicians make more informed and accurate decisions faster and to cull new insights from electronic medical records (EMR). It does so by accessing and analyzing massive amount of data (Big Data) of medical information that one human being could not possibly memorize and analyze, even more so when one considers that 10 million new pages of medical information are published each year.

After Watson’s win in the Jeopardy game, IBM has been struggling turning the technology into big business and it has not delivered significant revenue for IBM yet. So, could Watson be more successful, and valuable in Africa?

I see several reasons why it could be.

The lack of data, and particularly reliable data, in Africa is hampering its development. Shanta Devarajan, the World Bank’s chief economist for Africa, struck a dramatic tone in his address to a conference organized by Statistics South Africa, calling the state of data collection on the continent “Africa’s statistical tragedy.” The lack of reliable data on Ghana's performance on the MDGs was identified as one of the major challenges facing the country’s development projects. In Kenya, lack of data hampers national economic and developmental planning. Better geospatial data access would be good news for the globe's poorest people, who are often worst affected by natural disasters. Access to historic data about market prices of crops, which show trends in crop price fluctuation, would enable better decision making on which crops to plant to yield the highest income.

The digital data scarcity described above is also the result of a "chicken and egg" situation. On the one hand there is little data available. Most data in the region is available in analog format (on paper) as opposed to digital format making it difficult to be analyzed by computer systems and cannot easily be transformed into valuable, actionable information. This is the case particularly for small and medium sized enterprises (SMEs) in the private sector as most of their business processes are still manual. On the other hand, people and institutions are not building their capacities in data analysis and statistics because of the data scarcity and the human capital, such as analysts and statisticians, is missing.  

So the first value of IBM’s project Lucy is to break that chicken and egg problem by making one of the best cognitive systems available in Africa. Big Data technologies have a major role to play in Africa’s development challenges: from understanding food price patterns, to estimating GDP and poverty numbers, to anticipating disease – the key is turning data into knowledge and actionable insight. When data is processed into valuable information, it contributes to fact-based decision making in the public or private spheres, improving decision’s quality. That is what Watson can be applied for.

Then I believe that once people will see the value of Watson’s analytical capabilities, they will start collecting more data to feed into the system to get even more value from it. And in turn this will inspire young talent to develop their skills in that area as I can see with my students when I teach them about the strategic value of digital information.

Africa’s lack of data could actually turn out to be an advantage. In America, we are overloaded with data from many different sources in various formats, structured and unstructured, some call it Big Data. The challenge has been really how to clean, integrate, load and transform that data from different sources in the appropriate format to be analyzed by systems specially programmed for it. So another reason why Watson could deliver value faster in Africa is that there is not much legacy data and systems in place as explained earlier and new data can be formatted for analysis by Watson. Watson represents a new era of cognitive computing, in which systems and software are not programmed, but actually improve by learning so they can discover answers to questions and uncover insights by analyzing massive amounts of Big Data. Once Africans understand the value of Watson, they will be incentivized to generate and collect data for Watson analysis resulting in reduced time-to-value.

Another advantage is that in Africa businesses and organizations will be using a different IT paradigm. Instead of decentralized data systems installed on premise in enterprises and public institutions as is the case in developed countries, Africa is moving straight to the new model ofaccessing or creating data from mobile Internet devices (tablets, smart phones,etc.) connected to centralized IT resources in cloud service centers. By storing data in a central cloud space we may find that it facilitates data integration and makes access to data quicker and easier . For example, the use of mobile based Point of Sales (POS) systems in retail stores connected to cloud-based services provides instant centralization of data (important especially to chain stores), and the ability to access that data from anywhere there is an Internet connection.

Moreover, the use of mobile devices for data collection may allow for capture of better quality and more relevant data since it is generated closer to its source of creation.
But I think that the main reason for the potential success of Watson in Africa will be the “infinite” value it can deliver. The use of such systems could accelerate at a rate similar to the mobile phone penetration.

When mobile phones first became available in developed countries, they were a "nice to have" technology, but not critically needed. Indeed those societies already benefited from good communications for long time with postal services, express mail services, land-line phones, faxes, Internet email, etc. So the mobile phone was just another new and more convenient communication tool.  Such was not the case in Africa. Until the mobile phone arrived, there was no communication generally available: no post mail, very few phones, not to mention very limited Internet and that lack of communication has been identified as one of the main reasons for Africa’ slow development. So when mobile phones became available, their added value was “infinite” because there was “zero” communication systems available.  As a result, the mobile phone market in Africa has grown more rapidly than any other technology in any market, reaching more than 50% by the end of 2010, less than 10 years after its introduction.

Much like mobile phone use, Watson could be a game-changer in Africa. Watson’s added value in the West is analogous to mobile phone technology when it became available in that part of the world: marginal. There, enterprises and organizations have been using business intelligence systems and analytics for long time. While Watson capabilities are superior to these “old” technologies, the added value in the West was marginal. But in Africa, Watson’s value is to be viewed relative to existing data analytics capabilities.  In Africa, such systems are practically nonexistent due to the combined scarcity of data and skills to analyze that data.  This is why a system like Watson presents an “infinite” value, which could lead to a similar penetration phenomenon than that of mobile phones.

Yet for Watson’s value to be fully realized, two conditions have to be met.

First, Africa needs more, and more reliable, data and we hope that Watson’s value will incentivize stakeholders to undertake more data collection, as explained earlier. Thecurrent open government data movement could also be a factor as government isthe major data producer in the region .

Secondly, there must be more data scientist skills available. While the mobile phone is relatively easy to use, requiring minimal skills, Watson requires expert skills even if the use of artificial intelligence is facilitating the human-machine interaction. It is critical for academia to enter the game here and for IBM to collaborate with universities to promote the development of those skills in Africa.

In the meantime, an open data movement in Africa can attract expert skills from the world to analyze its data and deliver the value that may have an immediate impact on Africa's development, thereby leading to more interest and accelerating the movement.

“Watson's cognitive capabilities hold enormous potential in Africa – helping it to achieve in the next two decades what today's developed markets have achieved over two centuries.”

Seeya later alligator...