Visit HRO Today on Facebook

Thursday, August 4, 2016

4 Things Executives Should Know About Intelligent Assistants at Work

By Rob May, CEO and co-founder of Talla

We are in the early stages of the cognitive revolution at work. Over the next five years, every knowledge worker will begin adopting intelligent digital assistants to help do their job more effectively and efficiently. In just over the last year or so, we’ve seen the possibilities explode. The two technologies that recently made this possible are the advances in machine learning and natural language processing, which enable a good experience with bots and the rapid adoption of conversational interfaces as the main method of communication within the workplace—the perfect home for those bots.

The ecosystem of products is developing rapidly, and the first ‘generation’ of assistants available currently automate various tasks. The major focus has been on creating reports, scheduling meetings,  and managing project reminders so far. These are no doubt useful, and we’ll see continued development of those types of assistants across every vertical within businesses, from HR, to sales, to customer support. A second useful case that is developing are bots that lighten your cognitive load. These extend and amplify employee cognitive capacity by freeing knowledge workers from remembering certain things, intelligently managing workflows or storing vast amounts of information, and retrieving it when needed. At Talla, we work on both of these for HR teams, so we think a lot about the considerations executives should make in adopting the new technology. Here are a few things to know:

You need to be on the right chat platform 

Some organizations will lag behind in getting started with intelligent assistants, not because they wouldn’t save them timeand free the team to focus on more strategic work, but because they’re lacking the platforms with the right bot integrations. Today, I believe all teams should be evaluating the various chat platforms that bots will live in and deciding what’s best for their teams. Slack, Hipchat, Cisco’s Spark, and Skype are some major players. While companies that build bots will wisely go multi-platform over time, the reality is that not every interface will be optimized for bot and human interactions. Simply put, if the Slack App Store, for instance, has all the best bots, but your team is tied to email—and only email—you’ll be behind.

The best bots will learn from your interactions

In a typical buying decision for a SaaS app, you look at what it can do for you today. But, the value of intelligent assistants is better assessed by determining both what it can do today (typically base-level automation) and the additional value added over time as it better knows your organization. If an intelligent HR assistant knows 15 things about our insurance policies on day 1 and 250 things by day 30, it’s value has increased significantly as an internal resource for the team. Intelligent assistants will grow, learn, and improve much like human employees, so when you judge them, think about the long term, not just how they perform on day 1.

Consider security and privacy when choosing your bots 

Can a bot be socially engineered the way a human can? How do bots know what information they can disclose and to whom? These are all problems that will be solved in the next couple of years, but for now, most companies building digital assistants aren’t thinking about them. Look for companies with mature founding teams who have built other companies in the enterprise software space and understand the importance of these issues. They will be the first to adopt high security and data access standards and will lead the way for the other bot companies.

Don’t be fooled by A.I. buzzwords

Does an intelligent assistant use deep learning, neural networks, probabilistic programming, bayesian reinforcement learning? It doesn’t really matter. Many of the early A.I. companies building digital assistants have taken a “technology first” approach, but as long as the product works for you, the underlying machine learning models shouldn’t matter. Algorithmic innovations in A.I. get open-sourced pretty quickly, so any useful advances should be adopted by every digital assistant company in less than a year. Buy the best product for your needs and don’t let a buzzword-filled sales pitch drive you into the wrong decision.

Advances in A.I. are enabling a new breed of application that can extend and amplify the cognitive capacity of all your employees. These digital assistants will revolutionize all areas of work, but the largest gains will be seen in human resources and related fields. The best way to prepare your organization for the cognitive revolution is to start dabbling in these new technologies so that you can understand and appreciate their use cases. I hope this guide can be your first step in that direction.

No comments:

Post a Comment