08 Feb Can Generative AI Like ChatGPT Provide Value to Contact Centers?
With the internet buzz around ChatGPT making so much noise it’s easy to miss that AI tools have been a part of customer service for a long time. Plus, there are other GPT-3 and GPT 3.5 based AI tools out there (like Jasper and Youchat and Google’s not-yet-available-to-the-public Bard powered by LaMDA). This is a growing sector that creates as many questions as it answers. For the customer service industry, how can generative AI tools like ChatGPT impact the future of contact centers?
For context: OpenAI began in 2015 and has received funding from Microsoft in 2019, 2021 and 2023. OpenAI is behind the Dall-e (art generator), Chat GPT (generative AI chatbot) and GitHub Copilot (AI programming tool).
Bridging the Human-AI Gap – Carefully
These AI fueled tools are growing in sophistication every second they are being used as they continue to learn more about human interactions. Of course, the AI is only as good as the human communication it is fed (RE: Microsoft’s Tay chatbot being quickly turned into a bigot when ‘trained’ by the nasty side of Twitter).
Given the right kind of prompts these tools are able to write answers, emails, intros, articles, outlines and more. With the growing interest and need for self-service options alongside the customer desire for personalization and positive experience, a human sounding bot may have a place.
The devil is as always in the details. A bot that tries to sound too human can come off as creepy. If the effort to make a bot more conversational means sacrificing helpfulness or speed, that’s another issue.
The simple requests may be better served by simpler chatbots, but where tools like ChatGPT may help is in the more complex requests.
Approaching Generative AI Tools in Customer Service
It comes down to balance. Customers want to get their answer quickly, accurately and preferably in a manner that leaves a positive impression.
In 2019 a study was done that showed 86% of consumers prefer human agents over chatbots. However, three years is a very long time in the AI space and bots aren’t the same as they were. The useless answers, unnecessarily long ‘conversations’ and issues with escalation are problems the industry is aware of.
Generative AI may be able to address some of these problems.
Generative AI Helping Customers and Agents
These kinds of tools can help both sides. One example is the AI providing summaries to agents when a customer interaction is escalated. This saves the agent from having to ask the customer for information they’ve already given, saves the customer from having to repeat themselves, and makes for less agitation for both because of this efficiency.
Generative AI can work alongside agents, crafting responses and searching databases. AI benefits from human involvement because it sometimes lies or misses vital nuance. Further, the agent oversight is good for training, maintaining brand voice and ensuring appropriate empathetic response. On the flipside, the generative AI assistance is good for increasing speed and reducing redundancy among other things.
What Does a Generative AI Tool Say About Itself?
When asked how it could be used to improve customer service, YouChat responded:
“YouChat can be used to improve customer service by providing automated customer support. YouChat can respond to customer inquiries quickly and accurately and can provide personalized answers based on the customer’s request. YouChat can also provide real-time updates on order status and delivery information and can even provide product recommendations to customers. Additionally, YouChat can be used to facilitate customer feedback and provide personalized offers and promotions to customers to help increase customer satisfaction.”
– https://you.com
The generative AI wave is here and it hasn’t even hit its full height. Though it can’t operate exclusively without human oversight it can provide tremendous value. One thing it does well is create quality summaries of conversations. Generative AI is part of CSAT.AI’s future too, with language model summaries on the way that improve employee and customer experience.
