In 1995, when I received the textbook for my major subject in Information Systems, “Expert Systems and Applied Artificial Intelligence” by Efraim Turban, I felt like this was the stuff movies were made of.
The application of artificial intelligence
This subdivision of computer science, devoted to creating computer software and hardware that attempts to produce results such as those produced by people, enabling us to automate and enhance complex tasks that were done manually, seemed like an area best left to the gurus researching their days away in the MIT labs. But now, thanks to the rise of big data, the case for applied artificial intelligence is upon us.
Artificial Intelligence is closer than we think.
To help join the dots, consider the following definitions of Artificial Intelligence: First, it involves the studying of the thought process of humans; Second, it deals with representing these processes via machines (computers, robots, etc).
Regarding the first point, the fantastic thing about the human brain is “how a little bit of knowledge goes a long way,” in allowing a human to decide what is truly important. Intuition and gut feel may not find their way into machine learning capabilities, but truly, how many decisions do we as humans make in the business world that don’t require a fair amount of studying and analysing of factual data before making them?
Secondly, how we as humans thinly slice through all the data is not much different from the Map/Reduce frameworks inherent in Big Data Architectures. This is the principle of consolidating disparate data, big or small, and distilling it to a simple truth, so that a decision -- whether it is operational or strategic in nature -- can be made.
Watson, I presume
The exciting part is, this is not just talk anymore. Consider IBM’s best Cognitive system to date, that of Watson. Watson understands natural language; the system generates hypotheses recognizing that there are different probabilities of various outcomes. The system learns, tracking feedback -- learning from success and failure -- to improve future responses.
In February 2011, Watson appeared on Jeopardy, the quiz show known for its complex, tricky questions and very smart champions. The clues given to contestants require analysis and understanding of subtle meaning, irony, riddles, and other language complexities in which humans excel and computers traditionally do not. Watson received the first prize of $1 million.
Watson had access to 200 million pages of structured and unstructured content, consuming four terabytes of disk storage, including the full text of Wikipedia. August 2011 saw Watson enter the favorite big data arena of healthcare. In March 2012, it expanded into the Financial Services area.
Beating the Turing test
Alan Turing designed a test to determine whether a computer exhibits intelligent behavior. According to this test, a computer could be considered smart only when the human interviewer, conversing with both an unseen human being and an unseen computer, could not determine which was which.
Replace the interviewer with a high-powered CEO who picks up the telephone and calls the ABI (Artificial Business Intelligence) department to ask for the top 10 performing branches for the last three months -- and after a few minutes, he gets a report emailed to him. The scenario is not as far-fetched as it was for me in 1995.
Today, with tools such as Tableau (which suggests the best visualization for selected data), combined with underlying computing power as described above, the future may see even the “sexy” data scientists having to battle against the likes of Watson in the recruitment process.
Realistically though, it's good to end with an excerpt from my textbook:
AI is an excellent technology. Look for ways to use it, but do not expect miracles. On rare occasions, AI may give a miracle-like solution, but more likely it will not. It will deliver evolutionary rather than revolutionary improvements.
Related posts:
— Justin Lovell, Business Intelligence Lead, Etana Insurance Company
Tweet This
