You’re a working professional. Lots of meetings (maybe even some of those dreaded “meetings about meetings”). Stop for a moment and think about how much time you’ve spent, I mean wasted, playing email ping-pong just to get a meeting scheduled. Go on…last 30 days, how painful?
Coffee with James to discuss a partnership is AWESOME. The ridiculous process to find a day and time, not so awesome. It adds up. I won’t even mention reschedules. We’ve got better things to do.
In the U.S. alone, knowledge workers schedule more than 10 billion business meetings per year. Scheduling meetings simply should not be a human task.
Good for you if you can afford a human assistant (funny, I had to say “human”—you might know where this is heading), but shame on you if you’re paying some assistant $50,000+ per year to spend most of their time scheduling meetings. That assistant can do way more for you and your company once that task of scheduling meetings comes off their plate. In fact, most of us could really use an assistant to help us be more productive. That engineer you’re paying six figures to? She really shouldn’t be wasting valuable time scheduling her own meetings. The problem is, for all but a few, the cost of an assistant is prohibitive.
I personally suffered this very scenario: in my last venture, I scheduled 10,19 meetings in a year with 672 reschedules. It was miserable. And, once we were acquired, there were only two shared assistants for half a dozen executives in a 450-person company.
I believe EVERY employee should have a personal assistant. Fortunately, we have arrived at a moment when computing power has improved and the associated costs have declined so dramatically that we can solve this problem with AI. And that’s why 59 propellerheads and I have been in a basement for the last year and a half working very hard to develop an AI personal assistant to schedule meetings for you—Amy Ingram.
For the cost of a double latte each month, she’s all yours. Amy (and her brother Andrew) were born to schedule meetings, no more no less. So far she’s off to a blazing start.
As one of Amy’s bosses you simply cc: email@example.com when you wish to set up a meeting, and she takes over the job of scheduling. Amy negotiates with your guest, taking into account your calendar and preferences (“Amy, it’s football season; please no Monday morning meetings before 10am for me.”). She interacts as a human assistant would. After you request Amy to set up a meeting, you’ll be dropped from the scheduling thread, and the next thing you’ll see is a meeting invite from Amy. By taking on the job, she allows you to be much more productive. Instead of those eight emails, back and forth, go work on that forecasting document. Or if you like, sneak out for an iced coffee 🙂
We see Amy as part of the emergence of Vertical AI. One unintended consequence of the Internet is that, in easing communication and access to data, it has added to the workload of business owners and knowledge workers. Now, besides scheduling our own meetings, we book our own travel, publish our own writing, and launch our own ad campaigns.
So far, Apps have been the solution to this expanded workload. There are apps for tracking deals on plane tickets, and there are all sorts of calendar apps. But what I really want is for someone to buy the cheapest tickets that get me and my daughters to Miami for a long weekend or for someone to find a time and a place for me to meet with James.
I consider services like ours, which take over a job in its entirety, Vertical AI. They are specialized to do one job and to do it so well you might mistake them for a human.
In contrast, I see Siri, Cortana, and Amazon’s Alexa as Horizontal AI. They are extremely expansive generalists. There’s no single use case, no single “job-to-be-done.” They function more as massive question and answer settings (“What is the time in Berlin?”) or request, immediate-action settings (“Set my alarm for tomorrow morning 08:00 AM!”).
While Siri, Alexa, and Cortana, can help with a very large number of questions and immediate requests, they don’t take over a job, start to finish.
A “job” then is not a simple request/action sequence. It’s a request that might take many actions (and interactions) to fulfill—and might require hours or days to complete. When you ask our AI powered personal assistant, Amy, to schedule a meeting, you send her off, and she comes back hours or days later, only when she has negotiated a mutually agreeable time and location for all your guests while keeping your personal preferences in mind.
Horizontal AI is already here (if protean). I believe the next big wave of automation will deliver more Vertical AI, and these services will take on jobs for which human judgement has already been minimized. When you search Expedia or Kayak, algorithms already serve the results. Using AI to parse the available options (cheapest tickets to Miami over Presidents Day weekend) and perform the actual transaction (purchase tickets for me and my daughters) makes way more sense than having a human do this. Same goes for scheduling.
And once you posit the rise of a host of agents that do one job extraordinarily well, you can imagine a marketplace of Vertical AIs—in addition to those travel agents that buy your plane tickets and reserve hotel rooms, and bookkeeper agents that reconcile your accounts—that will ease some of the pain of working in a digitized world.
In this scenario, Siri, Cortana, and Alexa become enablers. Eventually, they will be able to summon these single-minded Vertical AIs at will. When you ask Siri to schedule a meeting, she’ll simply ping Amy, who will work hard to get the meeting on your calendar. When you ask Amazon’s Alexa to book a trip to San Francisco, she’ll reach out to a specialized travel AI to do that.
A future filled with Vertical AI and enablers like Siri also removes a more subtle tax on business owners and knowledge workers, one that we don’t even recognize anymore. For years, to capture any of the productivity gains that software promised, we’ve had to learn a new syntax. Remember the first time you fired up MS Office 2007 and the annoyance of having to use the ribbon instead of toolbars and drop down menus? What about figuring out how to use WordPress to update your company blog?
Amy, in contrast, requires no app, and there is nothing for you to download. Nobody wants to fiddle with yet another App on their phone. When scheduling a meeting, asking your guest to log in and choose times from some pre-set ranges on a web page is just as tedious and can seem almost presumptuous. Amy, on the other hand, requires only that you cc her, in the usual flow of communications.
Amy (and Andrew) are part of a trend, which I describe as invisible software. We are moving away from screens and radio buttons and drop down menus, towards interactions that build on our current habits. Our customers already use email to schedule the vast majority of their meetings. Cc’ing Amy builds on an ingrained habit of using email to schedule meetings and then seamlessly shifts the job to her.
Imagine a future where we no longer need to learn a specific syntax in order to use technology. We will unlock hours of productivity a month by removing this bit of friction. Sign me up!
To really make invisible software work—to ensure that it is truly invisible—we quickly learned that Amy and Andrew didn’t just need to communicate efficiently via email, they needed to seem human. There were two related reasons for this. The first is that we wanted to ensure that Amy’s bosses treated her as they would a human assistant. From a user experience perspective this was essential. If customers had to think about speaking to a robot, and give it specific robot-friendly commands, we wouldn’t be removing all of the pain. And that’s our goal, to make working with Amy or Andrew as easy as working with a human assistant.
The second reason is that the only way we can create these human-like interactions is by having good data. Using plain English with Amy provides us an opportunity to extract small (but important) details that get lost in an interface. Clicking a SCHEDULE MEETING button in some app a little bit harder doesn’t signal urgency. However, in plain english, you can stress an action, such as, “Amy, it’s really important that we meet before James leaves for San Francisco.” In plain English, you can also distill the output into conclusions such as “Dennis is busy.” You don’t need to visualize a calendar for the next week, filled with red “busy” blocks. And with this level of precision, we can continually improve Amy’s performance.
Surprising things happen when you create convincingly human-like interactions. Many customers feel compelled to thank Amy (and Andrew) even though they are fully aware that they are pure AI. Guests sometimes ask if Amy should join the call or meeting. And, occasionally, Amy even gets asked out!
But the reverse is also true. People can make requests of Amy that they wouldn’t make (or only sheepishly) of a human assistant. So say, you catch a bad cold, and you need to move all of your meetings for Friday. If you’re a business owner, with a booked day, this might take your assistant several hours of complex, painful, email negotiations. (Worse yet, you’d have to do this yourself!) But with Amy, you don’t have to worry about the tedium. You simply make a single request (“Amy, move all my Friday meetings to next week.”), and she’ll get the job done, and not resent you for asking!
Technology distributes its fruits unevenly and unpredictably—for all of the ways it makes our lives easier (remember typewriters?) it can make us busier (scheduling ping pong, booking our own travel). But we are fast arriving at a moment in which AI, and particularly Vertical AI, can remedy some of this pain. We want to democratize access to the personal assistant. AI, and a whole lot of blood, sweat, and tears, is making this possible.
Dennis R. Mortensen is the CEO and Founder of x.ai, whose artificial intelligence driven personal assistant lets people schedule meetings using plain English and nothing more than a CC to firstname.lastname@example.org. He’s a pioneer and expert in the analytics, optimization and big data space and has been since its inception – he is also a fully-fledged entrepreneur and successfully delivered a number of company exits. He’s an accredited Associate Analytics Instructor at the University of British Columbia, the Author of Data Driven Insights from Wiley and a frequent speaker on the subject of Analytics and Data. A native of Denmark, Mortensen currently calls New York City his home.