Artificial Intelligence (AI) has come on as a driving force across different sectors of today’s economy. Sometimes, though, it can seem difficult to pinpoint exactly what it means or where it is used. To understand it a little better, we can try to look at what defines intelligence and how that can be applied to machines. Intelligence can be summed up by a quote from Francoise Chollet, the creator of the machine-learning software library Keras at Google: “Intelligence is the efficiency with which you acquire new skills at tasks you didn’t previously prepare for.” Scientists and mathematicians have been working on implementing this into new machine technologies since the 1950s. However, most recently the progress has truly accelerated and innovations have flourished as a result. This is because they are becoming more efficient at acquiring new skills and as a result, they can perform increasingly more tasks.  A supplement to the improvements noticed in AI is that of machine learning. This is the branch of AI that corresponds to the learning aspect. It focuses on the use of data and algorithms to imitate the way that humans learn, gradually improving its accuracy. As you can see, AI and machine learning rely on data to be able to learn from.

Currently, we can observe two types of AI; narrow AI and general AI. Narrow AI’s are intelligent systems that have been taught or learn to do specific tasks without being explicitly programmed to do so. This is the type that we interact with almost daily, whether we realize it or not. Do you use a GPS function to navigate roads or a traffic tracking app like Waze? Waze uses narrow AI and machine learning to provide users with the fastest routes available to reach their destination. It does so by taking advantage of the collection of a lot of data and the capability of learning from this data to then perform the task of giving directions.

We can see the use of narrow AI and machine learning through innovations that are simplifying existing tasks or allowing for completely new ones. These include some of the most notable features that are making AI recognizable, such as the voice and language recognition technology of Siri, the engines designed to recommend products based on previous purchases, or the vision recognition systems in self-driving cars.

The general type of AI is most associated with old sci-fi movies of the ’80s and ’90s and involves a more flexible form of intelligence that is capable of learning to do a variety of tasks. Think Skynet from The Terminator, for example. It has not been fully developed yet but its potential would be to reach or surpass human levels of intelligence. This opens the door to the classic sci-fi debate about machine consciousness and whether human values and rights could be given to machines if they become sufficiently intelligent enough to develop independent thought. So far, this debate has been kept in the realm of fiction as the technology has not existed for this to be a reality, but some academics argue that general AI may not be too far away.

As far as narrow AI goes, it is already here and it is improving constantly. In the media, it is often masquerading under the term automation. Many people don’t realize that automation often includes AI components and that as AI improves so does automation. It has been blamed for the loss of many jobs, especially in manufacturing. In the past, these lost jobs have been replaced by new jobs in other sectors. Automation has been growing significantly and it seems that it is outpacing the ability to create new jobs. It was previously thought that certain complex jobs were safe from automation but this is no longer the case. As a testament to how far narrow AI has come, almost any job is at risk of automation now. Data collection or “Big Data” has given AI the tools to take advantage of machine learning software to gradually learn more and become more efficient at carrying out different tasks. As these systems become more integrated with aspects of businesses and our daily lives, we are creating a feedback loop of data. As more of these technologies are used, more data is collected and the AI becomes smarter and more capable.

The advantages of this are clear, more efficiency and lower costs lead to more productivity and profits for companies and cheaper prices for consumers. Essentially this is the likely result of an increasingly global and competitive business environment. The nature of competition will mean that narrow AI, even at a basic level of automated data collection and processing, will be involved in almost every business and transaction in the modern economy. This will benefit the ease of doing almost any transaction as systems take some of the human error and inefficiency out of the equation. Additionally, it will open the door to new amazing technologies. As these become more developed they will become more of a part of our day-to-day lives. Imagine all the ways virtual reality can help with education and training around the world. Top professors will be able to effectively teach students in almost any part of the world with an internet connection and a computer. As developed expands, fast and reliable internet will reach almost every corner of the world. 

What does this mean for businesses? It likely means that narrow AI will not only be common but widespread and more sophisticated in the future. This will likely translate to fewer human jobs being needed and an increase in productivity and accuracy of task completion. Some politicians and academics are already looking at this possibility and trying to come up with ways to embrace technological progress without giving up on those who may lose their jobs as a result. We’ll have to wait and see what that future looks like, but for now, businesses around the world are looking to see how they can incorporate narrow AI into their business structures.