There is a lot of talk about artificial intelligence. What is it really, and how do we make the most of it?
We read and hear about artificial intelligence (AI) very frequently across business media, above all where the use and implementation of digital strategies is being promoted. A reader who is not familiar with it may not follow its development closely, and so may stop paying attention to its growth and to its impact on their market and their competition.
AI is considered a branch of computing that has gained prominence for what it promises when applied to the large volumes of data companies have been collecting. One of the first to name it was John McCarthy at the legendary Dartmouth conference[1], almost sixty years ago.
Its techniques and methods can identify patterns in data more efficiently and more quickly, making it possible to draw more information and more insight from it. Thinking of an artificial intelligence that imitates the human kind may seem unreal — the human brain is more complex than an electronic one — but it is undeniable that some repetitive operations, such as translating documents or recognising faces, can be done better by computers.
The open question is what it is worth to a company's development, and here it is worth recalling one of the early articles by the well-known specialist Michael Porter, who noted that a company would not develop competitive advantage merely by using information technology, but by applying it in order to develop competitive advantage[2].
Porter's observation is more current today than ever: adopting and implementing AI in a company is not magic, and its contribution to any business will not show up immediately. It is technology that has to be prepared before it is of use to the company.
On that basis, one of the main activities to carry out is preparing these large sets of mathematical formulas and algorithms — that is, programming them so they can take automatic decisions from enormous quantities of data and so achieve machine learning.
There are three kinds of machine learning:
Supervised learning, in which the data sets are labelled so that their patterns can be detected and used to label new data sets.
Unsupervised learning, in which the data sets are unlabelled and are classified according to similarity or difference; and finally
Reinforcement learning, in which the data sets are unlabelled, but after carrying out one or several actions the AI system receives feedback and learns from it.
There are many applications of AI: in healthcare, producing better and faster diagnoses by drawing on patient data to form hypotheses about the possible causes of particular illnesses; in customer service, deploying computer programs (bots) used online to answer questions and help customers; and in business, with machine learning algorithms built into customer analysis platforms so customers are served better.
Before being carried away by the enthusiasm, first answer your own questions about how these new and attractive technologies will be used and applied. Success in using them will come when they are applied to developing competitive advantage.
[1] “A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence”. J. McCarthy, M. Minsky et al. Dartmouth College, 1955. [2] “How Information Gives You Competitive Advantage”. Michael Porter and Victor E. Millar. Harvard Business Review, July–August 1985.
