Why A.I. Is The Next Big Thing?

Every now and again we have a breakthrough technology that changes how we work and dramatically improves productivity overnight. Some of these innovations come to mind immediately, like the availability of electricity, assembly line production, or the internet. Others less so. These would include modernisations like air conditioning, which the US Government credited with a 25% increase in productivity in the 1950s. An aptly titled book, “Air-Conditioning America” by Gail Cooper, cites a 1957 survey in which 90% of American firms named cooled air as the single biggest boost to their productivity.

The point here is that breakthroughs are not all about a completely new technology or device in a specific industry where the new replaces the old. Some breakthroughs happen across all types of jobs, and are specifically focused on making people more productive. Their role is to take some of the slog out of completing manual tasks, not just doing one thing faster or better. In 2017, up steps AI as the next great breakthrough that will impact every aspect of our lives.

There is a plethora of reports on the explosive potential and growth of AI. One such prediction from Accenture predicts that AI could double annual economic growth rates in 2035 by changing the nature of work and creating a new relationship between man and machine. That sounds even cooler than air conditioning. It’s all about making more efficient use of our time and improving decisions.

It could include AI driving our cars while we actively work on the commute. It might be AI reducing absenteeism by proactively managing our physical and mental health. It could involve predicting which products will succeed and which will fail. So how can AI deliver on these world-changing promises of promises?

Well at the risk of sounding simplistic, in a word: Data.

With the internet and now the Internet of Things, we are generating vast amounts of data. The Indexed Web contains at least 4.51 billion pages and many more data sources off line on intranets, private servers or on legacy systems. In your head count to “one Mississippi”. While you were doing that 6,000 tweets were sent (per second), there were also 40,000 Google queries, and 2 million emails.  By 2019, global web traffic will beat 2 zettabytes per year. Zettabytes are the new Terabytes or more accurately, one zettabyte is equivalent to one billion terabytes. Big Data is on a growth splurge.

Computers have always been good at handling lots of data when it is in lists or coming from single sources, in other words well structured data. The new challenge is to use data that comes from many different sources or systems and to piece that together.

This traditional or structured data approach can start to unravel when we look at the type of data being created on line right now. In the last second there were 8500 Facebook posts and 2717 photos uploaded. It is hard to handle this ever increasing quantity of data and worse still, how do you organise items other than text such as YouTube videos, Instagram or Facebook photos? The list goes on and on. While data is being generated at a colossal speed, it is unstructured and not suitable to traditional computing capabilities.

AI can make a real difference here through its ability to harness and understand constantly expanding unstructured data from many diverse sources and formats.

AI can dig into this ocean of posts, likes and tweets to build a picture of how we think, what we want, how we feel, what our motivations are and lots more besides.

AI can look at these varying data sets or content and deduce relationships, build profiles and understand the behaviour of individuals, markets or economies. Because AI is trawling real-time data uploads, it is acting on the most up-to-date information, not some retrospective, sample based surveys. It generates real numbers you can trust.

It does this through techniques like Natural Language Processing (NLP). This approach looks for certain words or phrases, analyses sentiments or makes predictions about what types of messages or conversations drive certain behaviours or attitudes.

If you wanted to know what people are thinking about your company on the ether of social media, NLP can dig into social posts, associate them with certain cohorts, and describe what matters most to each group. This in turn can be used to respond to criticisms and compliments, resolve user issues and improve services and people experiences. Facebook has taken this approach to amazing ends, even using AI to predict and prevent suicidal behaviours.

This could be the start of a fundamental change in how we understand our society, how we get a feel for individual and societal norms, how our society is changing. Events like suicide can be worked back into identifying at risk cohorts and behaviours that can raise a flag. This opens the door to new ways of implementing preventative medicine or therapies such as nudging access to health professionals, access to early intervention services. We may no longer have to wait until a crisis point is hit or the proverbial horse has bolted. AI can take action in real-time, making a real difference to our lives.

Next up is the idea of Machine Learning. In its simplest form, AI helps computers learn and act without explicitly being programmed to do so. The computer trains the computer. At a high level, it searches for patterns in structured and unstructured data to advise program actions, considering the context of the data in question. This is where AI can be trained to understand sentiment from text and video, marry that with acoustic tone, time of day, locations, social context etc. From this we get an understanding of how the social media posts, economic data and behaviours work together.

Put all this together and AI can start to tell us when we are tired and need to stop working, it can stop us from getting tired by saving us time reading application forms or carrying out other mundane analytical tasks, it can predict the best route to work and predict which products or decisions will be a success and which will fail. We are talking about greatly enhanced meaning, accuracy and understanding of the world around us, in a corporate and personal context. AI and Machine Learning is complementing our lives not controlling them like James Cameron might have you believe in Terminator 2.

AI is the next big innovation, it’s here now, it may have just started but it will soon impact on every aspect of our lives. Zettabyte, remember where you heard it first.