Editor's note: This article is part of Inc.'s 2016 Best Industries report.

 

I should have seen it coming. It was, after all, the founder of X.ai, an artificial-intelligence company, with whom I was trying to schedule a phone call. 

I emailed his assistant, Amy, told her Friday was best for me, and ended the message with an emphatic "Thanks so much!" It wasn't until after I had hit "send" that I realized I had used an exclamation point to reply to a bot--a virtual assistant made of code and ether. I had spent the time to type out a sentence of gratitude a human would never read.

Maybe you, too, have encountered, and even thanked, Amy. Or perhaps Andrew, the other alias her cloud-based program can be set to use. They seem just like any other assistant over email, communicating in natural (American English) language to set up meeting times between two parties. There's no plug-in, app, or software needed; you conjure her just by adding her address to the "cc:" field on an email. Amy takes it from there.

She--for simplicity's sake, I'm using a gender pronoun to refer to this creation--is the product of more than a year of research and programming by a team of developers, artificial-intelligence experts, and one theater veteran who studied folklore and mythology at Harvard. The founding team of X.ai (pronounced "ex dot a aye") includes CEO Dennis Mortensen, who sold his previous venture, a predictive analytics startup called Visual Revenue, in 2013.

The following year, Mortensen turned his focus to AI. He and a small New York City-based team settled on trying to solve something super-simple: giving everyone who doesn't actually have a personal assistant back the hundreds of hours they spend scheduling, and rescheduling, meetings. Of course, before going full-steam ahead, the team--Mortensen, Alex Poon, Matt Casey, and Marcos Jimenez--first had to determine that the time had come when we regular humans were comfortable letting a non-human touch our calendars and our inboxes.

They raised $2 million in a seed round from the likes of Softbank Capital, IA Ventures, and Lerer-Hippeau Ventures. Early last year, they raised another $9 million, valuing the company at $40 million, pre-money. The company now has about 60 employees.

The difficulty of dealing with humans. 

X.ai is signing up new customers, but won't open the service to everyone who wants it until some point later this year. When it does, Amy will be available to schedule a few meetings here and there--maybe up to 10 a month--for free. But the startup is hoping to follow a SaaS pricing model akin to that of Slack or Dropbox, so there will likely be a charge, perhaps $9 a month or so, for using the service more frequently.

For now, X.ai is still working out some kinks in beta testing--and building more understanding into Amy. While she's very efficient about scheduling meetings, she still fails at times, because, as Mortensen says, "people are crazy."

"One challenge has been getting Amy to understand when someone is trying to cancel a meeting," Mortensen says. "Because people are so insanely nice over email when trying to cancel, it does not sound like a cancellation at all. Heck, it's something barely detectable to a human!"

Another hurdle: lies. Not the hurtful kind, but the kind that stem from misunderstandings or typos or just the way our brains work. Say you send the message "Can we delay tomorrow's meeting by an hour?" at one minute past midnight. That's a lie, because the meeting is today. And that confuses Amy. She still has much to learn before she can deal with our human quirks.

Tackling problems big and small.

There's an AI boom going on right now. Advances in machine-learning and a growing number of graduates with degrees relevant to artificial-intelligence fields mean more companies can dabble in it. Some might say it's not a single industry, but rather a tool employed by multiple industries.

Even if that's the case, some experts, including Peter Diamandis, the CEO of the X Prize Foundation and co-founder of Singularity University, believe AI will be behind the next wave of disruption of established companies, across industries, by innovative upstarts. For evidence of growth that's already underway, one recent analysis found AI companies are more likely to have an exit than startups in other tech sectors.

Companies are using artificial intelligence and machine learning these days to tackle the hardest problems on the planet. They're handling massive tasks, like instantly spotting potentially cancerous tumors or rapidly scanning hundreds of hours of surveillance video for terrorist activity. But small problems, like that 15 minutes you lost going back and forth about reserving a conference room for a call tomorrow, are low-hanging fruit. Many predict such "easy" issues will see the biggest shift thanks to AI over the next five years.

Thomas Slowe, the CEO of Nervve, a software company that does visual-image search in video, says he's seen a lot of growth in machine-learning work for the Department of Defense and other three-letter agencies over the past two decades. He thinks the time has now come for AI to reach everyone's everyday lives.

"Self-driving cars are on everyone's mind, but we'll see other growth areas around small instances in our lives where simple tasks can be made easier," Slowe says. "Say, planning a vacation or scheduling a meeting."

You already have Siri in your pocket. And maybe your company uses Microsoft's Cortana. There are plenty of competitors to Amy out there already--some more AI-centric than others. There's Julie from the French entrepreneurs who made the social event-planning app WePopp. There's Clara, from Clara Labs. There's  Overlap, which is similar, only without a human facade. And there's assistant.to, which is a Gmail plug-in that helps calendar-event booking.

According to a research paper published by Oklahoma State University, some 25 million meetings happen in corporate America every day. That comes to a whopping 10 billion formal meetings a year. Mortensen and his team have found that it takes an average of more than eight emails to set up each of each of those meetings. At large corporations, executive assistants cut down on the cost of all the time that takes. But for everyone else, X.ai sees it as a waste.

"Any workplace gives out an email and a calendar," Mortensen says. "I want this to be the same--it becomes ludicrous to not give each of your employees an assistant for scheduling."

It's X.ai's goal for Amy to tackle all of those 10 billion meetings. She still needs some work before that's going to happen--sometimes a human customer-service rep has to step in when the conversation veers away from scheduling, or the human corresponding with Amy poses an unrelated question.

Where Amy won't have any trouble, however, is when she meets another Amy or Andrew over email. There's no crazy. No lies. Just, as Mortensen calls it, "scheduling nirvana."

And since that phrase sounds like an oxymoron to us mere mortals, let it give us some hope for the near future of AI.