Your Human Virtual Assistant Will Soon Be an AI-Driven Digital Assistant
A common problem for any scaling business is bringing on the right help at the right time. Small-business owners often find that the people they can afford don't have the skills or experience to do the job well, and the people they want cost too much relative to the business's current cash flow.
Most of the work is labor intensive, repetitive, boring and must be done correctly. As luck would have it, that's exactly the kind of tasks we build computers to do.
Outsourcing 1.0: The past
In the past 15 or so years, outsourcing to international assistants and part-time workers has been the go-to solution for entrepreneurs looking to grow on a budget. Famous successes from this way of thinking include Tim Ferriss's The 4-Hour Work Week and websites like elance.com (which is now Upwork.com). Still, finding the right work to assign to the right assistant in India or the Philippines required a learning curve that was sometimes expensive and time-consuming.
Thanks to growing progress in AI systems, much of this is beginning to change. The next wave of outsourcing is not to virtual assistants, but to digital ones.
Outsourcing 2.0: The present
There are a number of administrative jobs that are being increasingly handled by machines. Just a few early examples of software replacing business functions are X.ai (scheduling/calendaring), Grammarly (proofreading), Conversica (automated email) and LawGeex (contract review). Companies like IBM and Cisco are beginning to tackle customer service, and startups like mine, Cyrano.ai, are looking to manage the entire sales process (even the negotiation).
Right now it's best to think of each system as a standalone assistant that is unable to communicate with other assistants (this is called Artificial Narrow Intelligence). You tell the scheduling robot your work hours, you tell the proofreading robot what to read, you assign homework to your robotic paralegal, etc. Each will carry out the assigned tasks and report back to you, the quarterback who manages all these systems and executes based on all their efforts and insights.
When incorporating all this into your life or business, it's best to add these technologies one at a time and learn how to use them before adding subsequent systems. The learning curve gets flatter as you become accustomed to trusting these robots and learning what does and doesn't work. As you get a feel for how they work, you adjust. The fact that you have to adapt to them, and they don't invisibly work together, are two of the growing pains of this stage of development and a commentary on what is currently state-of-the-art.
Ultimately, all these distinct tools will get stacked together into a single system and become expected commodities the same way all the features you take for granted on a cellphone were added over time. One year later, you will wonder how you ever got along without them.
Outsourcing 3.0: The future
Ultimately, we want a single system that can do all these things and more (which is a big step toward, but still not, Artificial General Intelligence). An important part of a truly great digital assistant isn't just its ability to handle all these tasks in one system; it will need to understand your priorities and work well with you. For this to work at all (and for sales and customer service to get better), AI assistants need to have better emotional intelligence than they currently have. The nuance of how and when things are communicated must be measured and calculated, which may or may not necessitate follow-up questions from the assistant.
What if we could assign specific personality to our digital assistants? This has been shown in movies like Interstellar, Her and Iron Man to make our commands and requests more conversational and fluid. The existing batch of Siri, Alexa and Cortana have names and little touches of humanity programmed in, but what will it be like when you can make adjustments to your own assistant? This isn't just about assigning a new name as the "wake word" that turns it on or picking an accent -- this is about adjusting communication style to optimize efficiency and satisfaction. Big picture vs. detail-oriented. Bottom line vs. relationship-driven. As these kinds of things get quantified through natural language understanding, the digital assistants of the future will be optimized for personality. That means they can learn how to best work with you rather than you learning how each of the systems currently work.
As this happens, a growing company won't merely outsource what have been traditionally entry-level jobs; it means the right technology could increasingly handle important aspects of the business. Imagine how many great companies have failed because they couldn't get the right talent in a critical department. Imagine a world where these problems are fading away.
Why this is coming sooner than you think
Imagine starting a business with the knowledge that your administrative, sales, legal and customer service departments were all built into your Microsoft Office subscription, or part of your Google startup scholarship. How connected to that ecosystem would your business be? Google, MIcrosoft, Apple, IBM and Salesforce are just a few of the companies spending a combined billions of dollars on making this a reality because they know Outsourcing 3.0 is a market that can probably be cornered. Building, buying and integrating the different individual systems that need to work together to build the ultimate digital assistant will take time.
Microsoft currently owns most enterprise contracts today due to legacy. It has an impressive economic moat around this space, large enough that failing to gain traction in mobile could be absorbed. Whoever wins this next arms race will save companies of all sizes so much money it will offer a compelling reason for any company to switch to a new ecosystem and suite of services. "Virtual assistant as a service" is coming, and it's (eventually) going to be awesome.
Entrepreneur Media is an investor in Cyrano.ai.