It’s an Exciting Time to Be Alive Thanks to Things That Aren’t

Google’s AI demo certainly caused quite the commotion. The demo showed how Google Duplex leverages an artificially intelligent chatbot that cannot only mimic a human, but can also manipulate the person on the other end of the call using verbal cues to guide a discussion. If you haven’t seen it, just to quickly sum it up, the AI calls a hair salon to schedule a hair appointment with a human receptionist. It used cues like “Mm-hmm” to direct the conversation.

So, what does this all mean? It is just a matter of time before we will be able to hand off tasks to an assistive machine that can interact as a human.

It’s an exciting time to be alive, thanks to things that aren’t. However, these advancements also raise quite a few questions about how AI will change the way we live and the jobs we do. Currently, a lot of attention is going to autonomous vehicles, with good reason (more on this in a bit), but with the addition of voice capabilities that are as believable as a real person, personal assistants will no longer be the luxury of executives and the elite.

Consider the opportunities that come along with Google’s Duplex ability to schedule a hair appointment. If scheduling a haircut can save five minutes, consider how much time the same AI would save a recruiter if it could first filter resumes from job applicants, call back the best of the best and schedule interviews based on the recruiter’s free time. If you aren’t convinced yet, think of just how many people are looking for jobs. To put some numbers to it, in 2017 Facebook was reported to have 25,000 employees, Google had 74,000, and Amazon had 566,000 worldwide. If AI saved five minutes per employee, it could translate to almost 11 years of saved time.

As the adage goes, time is money. So what are we going to do with the time we get back? While we can’t have specific answers to that question, we can begin to consider how AI systems will manipulate it.

At our core, people want to believe that these artificially intelligent systems care about us. This is evident in how Amazon is creating options for Alexa to verbally reward polite responses and how Google Assistant introduces itself with “Hi, how can I help?”. Additionally, Google has released data from “Assistant” that shows people are shifting from simple search queries to asking for recommendations. Recommendations are based on trust, trust that the system has our best interests in mind, which, for Google, means understanding the intent of the question and providing the most relevant answer. Accountability, however, has been an issue. Google’s web search has to show you which results are ads because they are not the most relevant to you. We can only assume that this will eventually translate to the Assistant.

We’ve also seen what can happen when content on social media is not held accountable. The systems still do their jobs, but they become biased by the content and, as a result, manipulate us into spending our time poorly. Let’s go back to the recruiter example. Facebook, Google and Amazon have over 665,000 employees. If recruiters don’t know why the AI is filtering some applicants to the top of the results, the biases of the AI will be passed onto people. If those biases are then weighted as a success, they will quickly flourish in the training models. Keep in mind that we frequently hear about researchers who don’t know exactly why their AI came to the conclusion it did.

If this sounds scary, it’s because it is. But that shouldn’t keep us from getting excited. One thing that keeps the excitement going is the knowledge that more and more job applicants value meaningful work. Often times profitable work and meaningful work are at odds, but that will change if AI can reduce the time it takes to get to the same profits. Google attracts some of the best employees with programs like 20% free time for creative work that will benefit the company. If more companies could do this it would not only increase innovation, but it will create a more competitive job market and economy.

AI isn’t just about assistants though. As previously mentioned, much of the conversation right now is about autonomous vehicles (cars that can drive themselves). The average American commute time is 26 minutes, but the fastest growing group of commuters are traveling over 90 minutes. In other words, autonomous vehicles have the potential to give commuters up to three more hours back in their day. That in and of itself is life-changing, but, again, how will lives change?

Researchers at Stanford have begun studies that are finding passengers in self-driving cars tend to do one of two things: fall asleep or review the build quality of the interior. Neither of these is a great use of time in a moving vehicle, but what really is? Finding ways to translate this free time to work will have the greatest benefits. Whether that means working longer, or getting home earlier to spend time with family and friends. Both have advantages, but for this time to truly be productive, the form factor of the vehicle or of the work we do will have to change. Vehicles are just not set up to support good office ergonomics. That said, conversational AI like Google Duplex has the potential to change the way we work, but in order for that to happen, we’ll have to learn to work differently. In the short term, there is a great opportunity for companies to adapt current processes and build new tools for use while commuting.

The fact is, the future of how we use our time is going to be different thanks to innovations in AI. The more we consider how we’ll deal with it, the better chance we have of using our new found time well.