Please read the article below. After reading the article, adopt the lens of one or more of the following types of stakeholders, c-level executive, Chief Data Scientist, Analytics Professional (Gap between Business End Users and Analytics Team), Data Scientist, Business End User. *tell me the name of the lens you picked*Then, summarize the key insights of the article. 300 to 500 words please. Will give thumbs up.

Background: IBM are the granddaddy of the computing world. Their founding came about through efforts at the turn of the 19th and 20th centuries to use machines to help process US census data. This led to the birth of tabulation-based computing and the dawn of the information technology age. In the decades since then, IBM have constantly innovated and evolved to keep at the forefront of the industry. Major developments including the development of mainframes, microprocessors, personal computers and magnetic storage have shaped the industry into the one we know today. Most recently, IBM have moved to position themselves as a key player in the Big Data and analytics market. Watson, which first gained fame by winning the US TV gameshow Jeopardy! in 2011, is the result of IBMs work to develop what they call cognitive computing. IBM Watson vice president Steve Gold told me the project heralds the arrival of machines that dont need to be programmed: they can learn for themselves. The Watson system, and the Watson Analytics service it provides, is named after the companys founder, Thomas Watson.

What Problem Is Big Data Helping To Solve? Until recently, language has been a big barrier between computers and humans. Computers are incredibly quick at calculations, and their logic is infallible. These are two qualities that make them immensely helpful to humans. However, they can only do what we tell them, and traditionally this has meant giving them coded instructions in a computer programming language. This means that anyone not technically skilled enough to create their own code has to use code written by others, and hope that someone has created a program that does what they want to do. In addition, computers until now have traditionally only known what we have told them. We give them the information we think they will need in order to solve the problems we think need solving. This introduces elements of human fallibility we have to know precisely what information they will need, and we have to know precisely what problems need solving. In theory, computers can learn much more quickly than humans. Upload an encyclopedia onto their servers, and all of the information is ready to be accessed with speed and accuracy far beyond human capabilities. Data analysis has shown itself to be immensely valuable in many fields, from preventing crime to curing cancer, as described elsewhere in this book. But computers arent, traditionally, capable of teaching themselves anything. We have to work it out first and give them algorithms to follow.

How Is Big Data Used In Practice? Connected to the Internet, and accessed through APIs, Watson in theory has the collective dataset of humanity at its disposal. It then uses algorithms developed through a field of study known as machine learning to work out what information it needs, and what it is expected to do. Over time, and given feedback on its performance, it becomes more efficient at this process, increasingly returning more accurate solutions. Watson is constantly updated when valuable information such as scientific studies are published, and interactions between Watson and its users are also analyzed to help it gain a better idea of what it should be learning, and how it can provide the best answers. Watson works in a probabilistic manner: ask it a question and it will return a series of likely answers, ranked according to their likelihood of being correct. Many use cases are already being found for this by IBM and over 300 partner organizations already working with Watson. One of these use cases involves improving care for cancer patients. To do this, it reads patient records, studies published in medical journals and pharmaceutical data to suggest the most effective course of treatment for individual patients. Natural language processing (NLP) is the backbone of Watson. As well as understanding instructions and questions in spoken English, it is learning to understand and help users who interact with it in other languages. This is thanks to partnerships with international businesses including Softbank in Japan and Mubadala in the Middle East. This means a major barrier between humans and computers the language barrier is gradually being disassembled.

What Were The Results? Watsons first public success was winning the Jeopardy! gameshow in 2011, defeating Brad Rutter and Ken Jennings. Its victory proved the success of the systems NLP capabilities, showing it was able to understand the questions, posed in English, to a high enough standard to win the game. It also provided proof of concept for Watsons data analytics and probabilistic modelling technology. Although Watson was challenged with almost every question that the human contestants were asked to solve, one type of question was omitted. These were ones that relied on audio and visual cues. At the time, Watson was not configured to tackle that type of unstructured data. Since then, it has been put to use across many major industries, including healthcare, marketing, retail, finance, waste management, crime prevention and security. Even toys are becoming more intelligent and useful thanks to Watson including a robotic dinosaur due to hit the shops in the near future that is capable of giving answers to questions posed by children. As well as being a clever and educational toy, it is thought it may have applications in recognizing the early signs of learning or developmental disorders, such as autism. Another service, developed with Watson by US financial services firm USAA, aims to help leaving military personnel transition to civilian life, by teaching itself about the issues they will face and offering support and advice.

What Data Was Used? Watson is plugged into the Internet and can trawl it for information to answer questions or help it learn. In addition, it is specifically kept up to date with particularly valuable information, such as newly published encyclopedias, scientific studies, news articles and statistics.

What Are The Technical Details? Watsons brain is made up of 90 IBM Power 750 servers, each containing eight cores, and has a total of 16 terabytes of RAM at its disposal. It uses this to power the IBM Deep QA analytics engines, running on the open-source Apache Hadoop framework. It is said to be able to process 500 gigabytes of information per second. Any Challenges That Had To Be Overcome? Early in its development, IBMs Watson team realized that exposure to a wide range of real-life situations was key to its ability to learn.

Although IBM are a huge company with employees all around the world, typically they were only ever presented with a limited number of problems to solve those relating to IBMs business. There was a danger that this could create a bottleneck in Watsons ability to learn and improve itself. To solve this, IBM began to develop partnerships with businesses across a large range of industries, including healthcare, education and finance. This meant Watson was constantly facing fresh challenges and learning to tackle a growing range of problems. Gold told me: Our partners have been very creative and very innovative in the ways that they are applying cognitive computing from veterinary medicine to childs toys, to redefining travel to the in-store retail experience. They are bringing cognitive capabilities to the forefront working with partner organizations can help accelerate our entry into critical markets.

What Are The Key Learning Points And Takeaways? Computers are capable of doing far more than what we tell them to do. Given their speed and accuracy, they can also be extremely good at working out what they should be doing and are probably far better at spotting problems and coming up with novel solutions than we are. Without a doubt, we are at the start of the age of the self taught computer, and this technology offers extraordinary potential to drive change. The language barrier has always been an obstacle that has prevented us from being able to use digital technology to its fullest potential, but with the arrival of affordable natural language processing we should start to see all kinds of exciting new developments.