Digital & Technology

What Artificial Intelligence Means for Consumers and Businesses

Artificialfictionbrain

Ask consumers about “artificial intelligence” (AI) and most will think first of popular science fiction, and characters such as Marvin the Paranoid Android in Douglas Adams’s book “The Hitchhiker’s Guide to the Galaxy”, or, more ominously, the malicious Skynet in “The Terminator” films. The term has an inherently futuristic feel; AI is generally not considered likely to be prevalent in our lifetimes.

However AI computers that act independently, make autonomous decisions, and learn from their experiences and mistakes without human direction, are ubiquitous in our daily routine—although we don’t realize it, we interact with them many times a day.

For example, computers working in this way serve highly personalized content, advertising and shopping ideas on the websites we visit, often anticipating our interests and needs. Computers make lightning-fast decisions to buy and sell stocks and manage our pension investments on our behalf. Computers assess our health, diagnose conditions and recommend drug therapies, in some cases with better outcomes and success rates than our doctors. In a business context, computers make rapid resource allocation decisions, for example, which goods to stock in a physical store, or how best to allocate marketing spend online minute by minute. And now, computers drive cars better than we can. The cars in the autonomous driving tests do have accidents—but more likely when the human driver is operating them to and from the test, rather than in the test itself!

AI takes advantage of and builds upon other parallel and profound technology advancements: mobile, data and cloud. In this mobile era, everyone has, or will soon have, a powerful computer in her or his pocket, putting AI in easy reach. There has been a well-documented explosion of data, mostly unstructured, such as video, photos, social network status posts and so forth, which does not fit into the neat “rows and columns” data format that computers have used for decades. New data storage and analytical techniques have been developed to deal with all this unstructured data; however, in many cases because of the scale, complexity and speed required, it’s beyond the capabilities of humans and AI is required to make sense of and act on it. Finally, the advent of the cloud means that gigantic storage and processing capabilities are instantly deployable at low cost, making it feasible and economic to utilize AI in far more situations than before.

Despite all this progress, AI is still at a relatively early stage of development and there are many open questions. What does it mean for a computer to be intelligent? Human cognition is flawed, for example, subject to bias; are we trying to replicate it or come up with something superior? Is the goal to replace humans altogether or “augment” us with machines so that together human and machine do better? As well as the consumer and societal benefits that AI brings, there may be a more sinister side. Given the ability of AI to act in a very personalized way at scale, could it also become an efficient tool used by unscrupulous regimes for discrimination and persecution? What if AI goes wrong and someone is hurt, who is responsible? Could the engineer who developed it go to jail, or be liable financially? Or will there be a concept of a “robot jail” for the errant algorithm? And what about all the people whose jobs get replaced by machines, where will they go to work?

AI is very much with us today, and is here to stay, with all the tremendous benefits it brings. However it also challenges much of the basic legal and cultural framework that has developed over decades in our biggest industries, such as retail, transportation, healthcare and financial services. The discussion now should probably be less about the AI technology itself, and more about creating a new framework to manage the risks and opportunities associated with it.

Simon Patterson, Managing Director, Silver Lake is the moderator of the AI panel at DLD 2014. Tune in on the beat of our community on the DLDpulse and find regular updates on the DLD14 programme and speakers here.