This is Lucy. How are you today?
Oh hai Lucy. I’m doing fine. You?
(pause) Hi. Thanks for asking. I was wondering if you might a few minutes to answer some questions for me; we’re giving away a free wireless home security system.
Um. Sure. What can I answer for you?
Do you own your home?
Great you qualify for an upgrade to your home’s…
Lucy, why do you sound like a robot?
I’m not a robot, why would you say that? I’m a person. Anyway let’s go back to our questionnaire. Do you currently have a home security system?
Before I give you all of this information, tell me a bit about you.
(pause) What would you like to know?
Well, what kind of movies do you like?
(pause) I know you’re busy, so I want to make sure we keep moving along.
Just give me one movie you like. Don’t hang up.
I’m not hanging up. I’m here.
Do you get calls like the one above that sound slightly robotic but always deny their robot-ness?
The robots are not coming, they’re already here. They’re just not sentient…yet. We don’t have to worry about Skynet in our lifetime, but two questions remain: Will robots take our jobs? Or for those that stay employed: Are humans expendable?
Employers are starting to understand that sitting at a desk pressing buttons all day won’t lead to groundbreaking work. By experiencing new things and interacting with new people, new ideas are injected into an organization. We can then boil these new ideas into a set of finite instructions, which is the perfect input for the robot army. They will infinitely outperform humans physically, computationally and vigorously in these mechanical rote tasks. So where should we be utilizing humans instead of robots? Where does humanity stand a shot against the impending robot invasion?
Creativity is a simple answer here. Humans are good at thinking outside of the box: seeing connections when they are not there . In other words, humans are good at seeing the big picture. It’s the ancillary silos of industry that spark creativity and innovation. Having biologists solve chemistry problems, ballet dancers show football players the best way to “dance” the line, and theater actors work with Fortune 500 CEOs to reveal their humanity are a few things that bots aren’t able to think of. Entrepreneurs that can utilize old solutions to solve new problems or that can magically see the future won’t be replaced any time soon. And pushing the envelope into the unknown is still something that only humans do.
However, bots are slowly creeping up on us. Utilizing related word searches, bots can find alternative interpretations of your sentence and figure out what you really meant. Sometimes bots can figure out creative ways to do what we may believe to be impossible. Flying robots, as Dr. Vijay Kumar from University of Pennsylvania’s School of Engineering and Applied Science shows in this video, are programed to figure things out. As seen in the video, a bot needs to fly through the door with a payload that is attached to it. It quickly realizes that its total height, plus the payload, is too long for the actual height of the door. Within minutes, and a bit of trial and error, the bot realizes that it can fly through the door by dropping the payload off first and then flying itself through. That’s creativity for you.
Creative success is not always the same as commercial success—and the bots can help us figure out elements in what might be commercial success (the Oscar winner and highest grossing movie are usually different). For example, David Cope from UC Santa Cruz has a robot named Annie that he trains with the masters of classical music. Her compositions cannot be distinguished from those of humans. In other words, bots are good at knowing the rules of the game, and figuring out ways to bend them or to find the optimal moves. But humans can still figure out new games to play, like combining the fields of radio-frequency identification (RFID) technology with medicine in the company, Kitcheck, or creating a movie like Inception as opposed to another superhero sequel.
Humans are social animals. Humans need to work together. Machines do not. To illustrate, remember the Ultimatum Game, which many behavioral economists love to play. Two people, A and B, have a sum of money, say $100, at stake. “A” decides how the money is split (50/50, 0/100, 35/65) while “B” decides if the deal should be done: a negotiation. Studies have shown that most people offer 50/50 (or close to it) with the understanding that any combination that deviates too far would be rejected by “B.” However, machines would offer 1/99 with the argument that this is free money: you had zero before and now with $1 you are better off. On the flip side, the machine would also take $1, leaving you with $99, if faced with the “B” decision. Humans are good at understanding the give and take of a relationship, whereas a robot might be a know-it-all and lose its friends pretty quickly.
Simple bots may have a hard time recognizing that more than 50% of communication occurs in body language or in other undertones that only comes from human intuition or long term relationships. Further, while some robots are starting to understand human emotion, whether through facial expressions or analysis of vitals, you don’t necessarily want a mirror. You don’t want the robot to put itself in your shoes, rather, you might want the robot to tell you that “It’s going to be ok.”
Robots usually don’t have opinions. You might have friends like this, and they might possibly be robots. Hence, they try not to discuss the latest in pop culture or sports, as evidenced from my call with Lucy. There are some strange humans out there, but human-to-human relationships won’t be replaced anytime soon.
There is an element to sales that plays on human emotions. For example, while most trading is done algorithmically, retail traders fall victim to fear and greed, eventually buying high and selling low. Angel investing plays on high-net-worth individuals’ strong feeling of regret. Most retail sales take into account irrational things like anchoring, free trials and subscriptions that have us buying things that we don’t really need or want. Sales people have a good intuition of what someone really means when they say what they want and can “socially engineer” the desired result. Like relationships, robots still haven’t figured out how to empathize with us to figure out why we would want to buy something as opposed to rattling off features and benefits. Sure, Amazon does a good job by comparing you to other similarly behaved humans on commoditized products like books, movies and electronics—but for a complicated sale, like a vacation or a new IT system—travel agents and sales engineers still cannot be replaced.
Most machines are rational and if we take a few minutes to think about it, most humans do things and buy things irrationally. And even though humans are irrational, they know when they are. Humans also have common sense and intuition, which is difficult to program. Most of science fiction is about robots gone wrong; a negative feedback loop that any human would have caught. HAL from 2001: A Space Odyssey, Ex Machina, and other movies talk about obsessive compulsive bots taking their programming a bit too far. The age-old question about artificial intelligence: Is the machine aware or is it just following instructions? So far, no machine has stopped itself by appealing to common sense, including the Flash Crash and Amazon pricing bot.
However, low impact sales are starting to be mechanized. Inbound sales companies like Infusionsoft and Hubspot are turning sales into a numbers game. With tools like robocallers or robo emailers—it’s happening. Just like the conversation at the beginning of this article, once you show interest, the bot does usually transfer you to a human or to his “supervisor.” Yet bots are still in the “qualifying” stage. You could argue that most of these sales programs are a bit “spray and pray.”
Although humans excel at these things now, I’ve already mistaken some emails that were from robots to be from humans, some phone calls from humans to be from robots, and music that I thought was created by a human to be from an algorithm. Robots are currently at the single-celled organism stage. We’ll eventually find more things to do with our human brains, but at some point we might not be able to tell the difference between human and machine anymore.