SUBSCRIBE TO OUR BLOG
Right now, the most important thing about Artificial Intelligence (AI) is not technology. It’s not about machine learning, neural networks or quantum computing. It’s certainly not the popularized myths of future utopia or dystopia. The most important thing in the world of AI at this moment is a dinner reservation for a first date on the upper west side of New York City.
Why? Because it will result in a real-world, human experience that can be planned and executed by a virtual personal assistant, fueled by a combination of humans and AI. And it’s one of many tiny steps toward AI’s eventual integration into the daily lives of you—and your customers.
AI has seemed to remain tantalizingly just beyond reach ever since IBM's Watson beat Ken Jennings in a much-hyped ‘Jeopardy!’ showdown. And ever since, the ability for AI to master and best humans at our own games has only accelerated—most notably when AlphaGo defeated Lee Sedol (18-time world champion) in a five-game match of ‘Go’.
While games like Jeopardy! and Go show us how AI can be taught and even teach itself, these examples produce no tangible effect that everyday people can sink their teeth into. Very few people are surprised, shocked or even excited by these triumphs of machine over man.
It’s as if the world watched as engineers developed the first internal combustion engine. If we were to watch every stage of the technology advancing, we might not be able to appreciate how great it would become and instead continue to ask “what’s in it for me?”
For individuals, enterprises and municipalities alike, data consumption, analysis and output in real time will help fuel the effects AI can have on our lives. And it’s already starting to happen in Auckland, New Zealand.
Mobile games company MZ leveraged an open version of their data streaming technology to build a console on top of a livestream of data flowing from Auckland’s 3,000 buses, including ticket sales, routes, schedules, and more. This console helped Auckland analyze and plan bus routes and experiment with new ways to promote ridership (such as dynamic pricing).
In Auckland’s test case, humans were responsible for the analysis and application of the aggregated data. But it’s not a huge leap to imagine an AI system assisting or even controlling the public transportation system as an automated, intelligent actor that learns over time the best way to facilitate more commuters getting from point A to point B.
When the personal digital assistants we employ—be it Siri, Alexa, Fin or any other—can search, reserve and purchase for us, our “automatable” chores and errands will surely be automated. Just as the vehicles on our roads will move toward automation.
But it’s the marriage of multiple streams of real-time data and AI platforms that will enable this automation. And lest you doubt our ability to eventually trust and take comfort in this automation, consider the level of automation in the average commercial flight.
As futurist Gerd Leonhard has been quoted, “anything that can be automated or digitized will be.” But he is also quick to add, “anything that cannot be automated or digitized will increase in value.”
So the question is not and has never been whether there is a use for machines that can teach themselves and each other to know our needs and perform tasks for us. The question is—and will remain for some time—exactly what people will want or need them to learn and what tasks we will be comfortable assigning to them.