Automation and artificial intelligence are transforming the customer service industry and raising difficult questions about the role that humans will play in the workplace of the future.
For years, pundits have been largely relegated to one of two camps: mass unemployment predictors and budget-minded capitalists. But recently, more and more thought leaders have shifted into a third camp, which argues that the future of customer service and business in general will be rooted in AI-human collaboration.
Today’s customer service leaders are divided about how best to integrate AI into their organizations. However, there is one thing that I think we can all agree on: When implemented correctly, virtual agents can help their human counterparts more efficient and effective. For example, they can make humans smarter with automated suggestions and intelligent search capabilities. They can also help humans save time by taking over data entry and other simple tasks.
We also know that the converse is true: Humans can do a great deal to help their virtual counterparts. Today’s generation of bots are easily confused, and humans can jump in when they get something wrong. They can also provide feedback so that the AI system can learn from its mistakes.
Many business leaders fail to see that humans have much more to offer their AI-powered counterparts than supervision.
Indeed, the future of AI-human collaboration won’t just be rooted in shared tasks and projects; it could also be rooted in shared intelligence.
The Value of Subject Matter Expertise
There is no doubt that bots can make humans smarter. However, the sharing of intelligence can and should be a two-way street. Virtual agents have much to gain from humans’ supervision, occasional intervention and training.
I’m not talking about just any training. I’m talking about the intensive and ongoing transfer of subject-matter expertise from people who have deep wells of knowledge and experience with your organization’s products or services.
Training from your human experts should give virtual agents the tools they need to reach a whole new level of functionality. If your experts are constantly reviewing and refining the AI’s responses to customer inquiries, those responses can become more thoughtful, informative and ultimately useful.
If AI-human collaboration in customer service is to succeed over the long term, it is vital to develop more nuanced systems that enable humans and AI to continuously share their intelligence.
I believe the future of human-AI collaboration in customer service will be more than a simple binary with humans tackling certain types of requests and AI handling others, with an occasional hand-off happening between the two. That’s because, as AI technology progresses, the line between human and artificial intelligence could become increasingly blurred.
In the future, AI-assisted humans could be the norm rather than the exception. We have already entered the age of augmented human intelligence, as Pew estimates suggest that smartphone ownership now extends to more than 2.5 billion people worldwide.
Before long, any company that assigns human agents to certain types of tasks could, by definition, be assigning AI-assisted human agents to handle those tasks. The converse is also true: A company may assign AI-powered virtual agents to many types of customer service requests. However, the most effective AI will likely be those that the company both trains and continuously “coaches” through human experts.
The future of customer service could indeed be human-AI collaboration. However, we should escape the simplistic binary thinking that positions humans and AI as diametrically opposed entities.
To achieve its full potential, we should bring expert-trained AI systems together with AI-assisted humans. We should continue searching for new ways to integrate the two until we are truly able to deliver the best of both worlds. By embracing this partnership, we can help ensure an enduring place for the human worker in the customer service economy.
How to Make AI-Human Collaboration a Reality?
Business leaders who wish to bring AI-human collaboration to life in their customer service organizations need only ask themselves a few key questions to get started:
1. Have you chosen the right VA vendor for your needs? Unless your company has the resources and expertise to develop its own virtual assistant, you’ll need to work with a VA vendor. This decision is generally driven by the level of feature-richness you need, as well as your preexisting vendor relationships.
2. Who is training your AI? Assuming you have a basic virtual assistant running, you’ll need to think about who is training your AI. Are you merely asking reps to flag the system when a virtual assistant answers a question incorrectly, or are your organization’s experts who are passionate about your product creating new content and reviewing the AI’s output?
3. How do you make expert knowledge available? How will you tap your expert users to identify gaps in your self-service content and solutions and to create support content that improves AI performance? Leverage existing customer support teammates with the greatest product knowledge or consider hiring representatives who are already experts in your industry.
4. What happens when AI isn’t enough? The most effective virtual agents are just one part of a comprehensive customer experience strategy. Make sure you’ve created a defined path of escalation so that the AI can seamlessly hand off to human support as needed.
By Antony Brydon Special Contributor Polo Lifestyles 2020 •