AI will Become as Pervasive as Electricity
In 2017 AI will become as pervasive as electricity, it will be there all around you but you just won’t see it. It will not adopt the form feared by many – as the victory of the robot overlords – but will automate repetitive tasks, provide ambient knowledge in context of the task being undertaken, and will deliver levels of productivity that knowledge workers have never seen before. There are too many examples around us already today to ignore this. We will however no longer think of applications or devices as smart, we will instead just view them as modern, and everything that has come before will be just old fashioned or dumb. We will no longer refer to our phones as smartphones, or our IoT enabled homes as smart phones. The inherent smartness will become the norm and anything that is not ‘smart’ will be seen as inadequate or inefficient.
In the same way that we use the intelligence of predictive search, Google Maps, or Facebook’s face recognition, we will begin to use smart systems to automate and accelerate the tasks of the knowledge worker in business; first in sales, marketing, customer service and other customer facing functions, and then in the less glamorous roles in back-office functions like finance, procurement and operations. AI systems will evolve from simple pattern based predictions to prescriptive system built on a base of contextual knowledge (like Google Maps) that understand the substance of the task at hand, applies the data that is applicable at that point in time and notify the user just when they need to act.
At the lower end of the knowledge curve where tasks are simple, repetitive or logically structured, AI systems will replace many jobs or at least many low-end task. AI is moving from consumer- or task-based utility, to applications that are more knowledge-worker focused—though there are still many application areas that are beyond the capability of the machines.
Most functional AI systems in business will still need to be taught the basics of the knowledge domain to which they are being applied. We are still far away from the time when tasks that require imagination, creativity or judgment are achievable by machines alone. In this area AI stand for Augmented Intelligence, not Artificial Intelligence.
AI promises tremendous productivity gains in all walks of business and life but it must be approached with caution and diligence. As we build AI systems we need to be careful that the AI models being built don’t just reflect the biases of the model creators, or where the AI system is learning from human intervention, the trolls need to be managed (e.g. Microsoft Tay) lest we further polarize an already polarized society.