Everybody’s talking about it. Some people want it. Some people have it. Many people are threatened by it. Artificial Intelligence. It sounds scary, but fear not, civilization is not under serious threat from AI. AI that is capable of learning intellectual tasks in the same way as humans may seem a little way off, however, key elements of it – namely machine learning and representation learning – are already on our doorstep. This alone will transform both the process and outcomes of work and certain industries by adding efficiencies in understanding and learning from structured and unstructured data. As many media stories and business presentations about machine learning are accompanied by unsettling illustrations featuring humanoid robots, people cannot be blamed for believing that the arrival of Artificial General Intelligence (AGI) is imminent. This is not necessarily a negative nor a foregone conclusion until progress in basic technologies around AI are advanced further. Much has already been written about the likely impact of AI and the importance of carefully managing the transition to a more automated world. The purpose of this article is to provide a different overview to help executives understand how AI is a path to help augment learning before we have parity of machines achieving human-level intelligence.
Firstly, we need to reframe AI in order to gain a better understanding of its applications. In 1996, society witnessed the computer Deep Blue making history when it outplayed Russian chess grandmaster Garry Kasparov in a friendly game of chess. Although novel, this was a misleading representation of the use of AI, which is rarely an epic contest pitting humans against machines. AI is not about machines taking over, but rather the utilization of computers to enhance, streamline, and simplify complex tasks. Machines and humans are partners that enable greater efficiency. Machines are objective – they don’t have down days, they don’t have egos. They are fast, accurate, and can operate at a scale that we can only imagine.
But, as commonly referred to in computer science, machines have to be trained and if you train a machine on garbage, what you get back in return will be garbage, supporting the concept that garbage in, garbage out (GIGO) flawed data produces nonsense. As a result, the quality of data becomes more significant in the AI age in order to have efficient tools that help leaders and executives. Take, for example, the popular analytical tools Power BI and Tableau. They provide powerful visualization to a general user so long as structured data is well-organized by the user. Without specific expertise and human intervention through programming, there is still a large void for a machine to quantitate unstructured data in order to make it visual or useful. This is where we see the advantage we have over machines. Machines are driven by data, which means they are not creative or original thinkers.
Put simply, this makes our relationship with AI clear cut: leaders have ideas; machines hone and optimize them. We create more effective work, and they constantly learn which work is effective.
What capabilities would enable AI to Augment Intelligence?
It is less about Artificial Intelligence and more about Augmented Intelligence to ensure the utility of tools that may come from AI. To understand the complexity of achieving utility from AI or even human-level intelligence, it is worthwhile to look at some the capabilities that AI will need to master. We don’t need more data, but more tech advances in areas such as problem solving, sensory perception and image processing, natural language understanding or linguistic analysis, creativity, and the ability for humans and robots to coexist with social and emotional engagement. It is this blend of these innovations that will be transformational.
In any general-purpose application, a robot (or an AI engine living in the cloud) will have to be able to diagnose problems, and then address them. A home robot would have to recognize that a light bulb is blown and either replace the bulb or notify a repair person. To carry out these tasks, the robot either needs some aspect of common sense described above, or the ability to run simulations to determine possibilities, plausibility, and probabilities. Today, no known systems possess either such common sense or such general-purpose simulation capability.
Sensory Perception and Image Processing
Whereas deep learning has enabled major advances in computer vision, AI systems are far away from developing human-like sensory perception capabilities. Current computer vision systems are also largely incapable of extracting depth and three-dimensional information from static images. Humans can also determine the spatial characteristics of an environment from sound, even when listening to a telephone. We can interpret the background noise and form a mental picture of where someone is when speaking to them on the phone (for example, on a sidewalk, with cars approaching in the background). AI systems are not yet able to replicate this distinctly human form of perception.
Natural Language Understanding or Linguistic Analysis
Humans record and transmit skills and knowledge through books, articles, blog posts, and, more recently, how-to videos. AI will need to be able to consume these sources of information with full comprehension. Humans write with an implicit assumption of the reader’s general knowledge, and a vast amount of information is assumed and unsaid. If AI lacks this basis of common-sense knowledge, it will not be able to operate in the real world.
Commenters fearing superintelligence theorize that once AI reaches human-level intelligence, it will rapidly improve itself through a bootstrapping process to reach levels of intelligence far exceeding those of any human. But in order to accomplish this self-improvement, AI systems will have to rewrite their own code. This level of introspection will require an AI system to understand the vast amounts of code that humans cobbled together and identify novel methods for improving it. Machines have demonstrated the ability to draw pictures and compose music, but further advances are needed for human-level creativity.
Social and Emotional Engagement
For robots and AI to be successful in our world, humans must want to interact with them, and not fear them. The robot will need to understand humans, interpreting facial expressions or changes in tone that reveal underlying emotions. Certain limited applications are in use already, such as systems deployed in contact centers that can detect when customers sound angry or worried, and direct them to the right queue for help. But given humans’ own difficulties interpreting emotions correctly, and the perception challenges discussed above, AI that is capable of empathy appears to be a distant prospect.
In the Augmented Intelligence setting, we are beginning to see the benefits of AI in our day-to-day lives with our correspondence using certain apps and our devices being be able to predict emotional responses to content – think of the last time you sent a text message and were suggested emojis or predictive text as you are writing a word. Consider the ability to instantly marry images to a message from your phone and perhaps even AI’s intuitive sense to predict our responses to video. In the future with machine learning, all of this may be at a speed and scale that was previously unimaginable. It will not only be able to express positive and negative sentiments with emojis, but also a full spectrum of emotions in other engaging ways, all at the push of a button.
AI is already here, though not completely ready for prime time as AGI, but some of the capabilities are appearing in places you might not expect. The benefits of AI will appeal most to those who are perceptive, open-minded, and prepared. This is not science fiction. These are all products currently being developed, with the intent to revolutionize the way we use technology in our everyday work. This is not to say we will lose the unique interpersonal relationships we value with human-to-human contact, as we won’t expect human empathy to be replaced by an automated simulation. However, there is a paradigm shift happening where there is an opportunity for building AI to be beneficial rather than simply intelligent. By adopting and blending Augmented Intelligence with machine learning, we do not merely want intelligent machines that pursue automated actions, but for humans to remain in control of the objectives they contain. We want machines that are advantageous to us. Instead of writing an algorithm that finds an optimal solution for a fixed problem, we have the chance to write algorithms that solve this problem, the problem of functioning as an independent human in a very computerized world – a harmonious component in a combined system. This produces AI behavior that is beneficial to humans. Keeping this in mind, we are all waiting for the day when our work can become more optimized, more efficient, and therefore easier, with a little help from our AI friends.
Photography by: Reuters