It’s been a big month for artificial intelligence. On May 9, a hot startup unveiled a shape-shifting personal assistant. On May 12, a legacy computer company that thinks it’s a startup shifted a platform to cyber security.
VIV is the product of two artificial intelligence wunderkinds, Das Kittlaus and Adam Cheyer. They’ve been tinkering with the smart frontier of software for the better part of 15 years. In fact, the two were working on natural language programs at DARPA long before working at that defense project was cool. They always believed a natural language mashed-up with artificial intelligence was the next logical computing platform. Getting there was tough.
So six years after they sold their first attempt to Apple, Kittlaus showed a personal assistant so smart it routinely answered complex questions like, "Will it be warmer than 70 degrees near the Golden Gate Bridge, after 5 p.m., the day after tomorrow?" And fulfilled tasks such as, “Send Adam $20 bucks for my share of dinner last night”. VIV was able to do this because behind the scenes its software was dynamically generating its own code to patch together information from many disparate open application program interfaces, or APIs. Viv is alive.
IBM’s artificially intelligent software, Watson, is best known for defeating for quiz show champions. Sadly, quiz shows don’t pay the bills so the firm recently put Watson to work in various consulting capacities. The latest is tackling cyber security threats. Eight universities will dig through security blogs, white papers, social media posts and news items culling the raw data Watson needs to be effective. IBM hopes to bring Watson to life later in the year with the goal of reducing the number of false positives that have had security researchers tied up in knots, chasing dead ends.
That is the key takeaway from both Viv and Watson. They are alive. Not only do they absorb, organize and analyze data, they determine what should be done. Kittlaus envisions a world where Viv lives on everything. Like the film Her, Viv moves seamlessly among devices, learning everything about you to make interactions personal. VIV knows who Adam is and that you owe him $20 for dinner last night because she’s intimately familiar with your contacts and calendar. She knows the best way to send him $20 is Venmo because you both use that service. So when you tell her to send 20 bucks, it just happens. Watson is doing the same thing. He’ll take all of the data fed to him and make sense of it. Then, magically, he’ll form a plan of action and proceed. He won’t need help or further instruction at that point. He’s alive.
IBM chief executive Ginni Rometti likes to make the point Watson is so transformative for business because he excels in five key areas; engagement, scaling expertise, putting advanced learning in every product, refining the supply chain and discovery. It’s not a stretch to see how each of these concepts can be scaled to almost everything.
And that worries smart people like Stephen Hawking. The concern is software that is alive, continually curious about relationships and competent is also inherently dangerous. Hawking draws the analogy of intelligent software developed to run a hydroelectric project. Given competency would that software choose to flood lands for the betterment of the project despite large colonies of anthills? It doesn’t flood the land because it hates ants. It chooses to flood because of core competency. What would Watson do?
Viv and Watson are capable of amazing things. This month we saw the first glimpse of that. Hawking says artificial intelligence is either the best or worst thing that has ever happened to humans so it’s important that we proceed slowly and get it right. Or become ants.