Editorial
With all the buzz around chatbots and artificial intelligence (AI), you could be forgiven for not understanding where one ends and the other begins. You’re not alone. The 2017 Gartner Emerging Technologies Hype Cycle showed virtual assistants and machine learning at the very apex of hype, ready for a correction.
There needs to be an important distinction between the two, especially when it comes to how they are applied in the workplace.
Workplace chatbots programmed with canned responses that help employees navigate the business have been around for over a decade. Witness IBM Sametime, Cisco Jabber and many call center applications. With the addition of text and voice natural language processing (NLP), chatbots have graduated to more lifelike question-and-response capabilities. However, despite the hype from many vendors, most of these chatbots aren’t intelligent. At best, they use machine learning in only the most rudimentary of ways.
From Chatbots to Intelligent Agents
Chatbots today may be clever, but for the most part they’re still preprogrammed algorithms. In order to truly transform business processes, chatbots need to evolve into intelligent assistants: conversational interfaces built on AI that learn from everyday work, understand meaning and context, reason out meaningful outcomes and next steps, and interact with us in more human ways.
What could that look like?
Today’s chatbots rely on a fixed model of the business. In contrast, intelligent assistants would evolve as the workplace and users evolve. One example: marketing administrative assistants that are built directly into the marketing software user experience. These agents can track an individual marketer’s usage patterns and interactions. The ability to track and understand common marketing work processes at an individual level can enable an intelligent assistant to simplify what a human marketer is doing and perhaps identify new ways to augment processes at key inflection points. Such an assistant could even improve as the marketer (or marketing department) improves.
Second, we need agents that analyze and synthesize answers and intent (using natural language generation and answer selection), not just respond to questions with canned human-crafted text (which involves natural language understanding). Two-way NLP relies on a deeper understanding of how every word matters, especially when it comes to the language specificity in your business. Just think of all the jargon and terminology in your workplace that takes months or even years for the average employee to master.
When agents can take a deeper conversational approach to language, and combine it with all of your business intelligence, then you’ll have a trusted assistant that sits by your side, participates in your day-to-day work and better connects to all the work you do. Imagine capturing best practices across the organization, such as in-depth process knowledge from an experienced employee who’s about to retire, and making it available to newer employees, everywhere.
Related Article: Chatbots Belong in the Workplace (Provided They're Well Designed)
Learning Opportunities
Webinar
Jun
6
The Future of Social Media: Emerging Trends and How to Stay Ahead of the Curve
The impact of emerging technologies like AI on social media
Webinar
Jun
7
The Building Blocks of Personalization
This webinar will explore how personalization can create powerful ecommerce shopping experience.
Webinar
Jun
8
Attracting—and Keeping—Your High-Net-Worth Investors
Join PwC to dissect their recent survey findings on a specific customer subset — high-net-worth investors.
Webinar
Jun
13
The Future of Contact Centers with AI at the Helm
Learn how to balance human intuition and interaction with the potential uses of AI to create delightful and lasting CX
Webinar
Jun
20
3 Steps to Unlock Your Customer Data and Drive Revenue
Discover how Twilio Segment’s #1 Customer Data Platform helps implement data-driven strategies.
Webinar
On demand
AI for Marketers: Optimize Personalization Across the Funnel
Understand the benefits of utilizing AI across your entire marketing stack
Key Steps to Bringing Intelligent Agents Into Your Workplace
Here are four key steps you can take to move from basic chatbots to intelligent agents that can transform your workplace:
1. Look for the right augmented intelligence to build your agents
Using weak machine learning to create what amount to prebuilt dialog flows simply won’t scale to handle the breadth of language and process understanding, or your rapidly evolving workplace. Look for AI tech that can understand your specific business domain, is capable of being trained to improve over time through observable behavior, and can easily scale to adjacent business departments or processes.
2. Identify low-hanging fruit
Look for high-friction business processes that could be better handled by an intelligent agent, and free up employees to focus on higher-value activities. Is the process one in which an agent could take on low-level, rote decisioning as it evolves its understanding of the domain? By democratizing and scaling expertise in this way, organizations can build a competitive advantage. One example: Buzzfeed gets a leg up on hiring the best people in a competitive labor market by using intelligent assistants to manage the large number of job applications it receives. Other examples of areas where intelligent agents could be deployed include employee onboarding, sales enablement and support call centers.
Related Article: Artificial Intelligence Will Change the Workplace Quicker Than We Think
3. Look for AI that can support cloud-based, on premises and hybrid systems
Key innovation in augmented intelligence is happening in the cloud. Companies gain real competitive value when they can use private enterprise data to train workplace agents. So make sure your AI technology can work with data and systems in any environment.
4. Persistence pays
Just as it takes time to train a new hire, it takes a while for intelligent agents to gain the experience with real world settings that they need to learn and grow. However, “cold start” techniques that seed an agent with domain knowledge can accelerate time to value and a provide the agent with a foundation upon which it can refine its understanding. Your AI system should support the application of these techniques.
Learn how you can join our contributor community.
