These last several weeks have been focused on fixes, touchups, and minor enhancements to the Kognetiks Chatbot for WordPress. I started a list of feature requests a while back that originated in the chatbot Support Forum on WordPress.org, as well as direct messages from a wide range of users, web developers, site owners and their administrators. One request that appears several times in slightly different forms is support for text-to-speech and speech-to-text. So, this is something that I’m starting to look into.
This made me ponder the future of chatting with bots, whether through typing messages or speaking to them. As bots get better at listening and responding, it will become very difficult to discern if we’re talking with a real person or with “the machine”.
It will not be long before ambient background noise – the noises that are present in an environment but are not the main focus of attention such as those heard from the buzz of a call center – is heard, or rather not heard, while on the call with an autonomous bot answering your customer service call. There may even come a time when breathing, coughing, and even laughing are injected at appropriate times in the conversation. I’m not sure that we will know when a snicker or giggle has been purposefully injected into a response. All leading to difficulty in determining if you’re speaking with a person or with a machine.
Your brain is adept at filtering out background noise when you engage in live conversations. This process involves the brain actively ignoring ambient sounds to focus on the dialogue, unless the surrounding noise intensifies to a level that overshadows the speaker’s voice. This capability stems from an auditory processing function known as selective attention, which allows you to concentrate on specific sounds, such as someone’s voice, amidst a cacophony of noises. This neurological mechanism is crucial for effective communication, especially in noisy environments, by prioritizing the sounds you need to hear while minimizing distractions. So, when bots start injecting ambient noise it will need to be very, very subtle.
One thing that live customer service agents do is ask for permission to put you on a brief hold while they check on something. This can be very annoying, especially when the hold music is too loud, or the hold time extends so long that that the music starts to repeat itself. But the act of placing the customer on hold is a very human behavior. Machines are not intentionally programmed with such human behavior, yet.
Another thing that live customer service agents do – at least the very best of them – is to say that they don’t know but will find out. They’ll resort to using company resources, asking colleagues who may be more experienced, engaging a higher tier agent with more experience, or resorting to asking a supervisor or manager to making a judgement call should it be needed to answer the question or solve the customer’s problem (think about offering a refund or a discount just beyond the timeframe to a customer, i.e., I can’t do that but I can offer you this instead – satisfaction by bargaining instead).
Research in psychology and behavioral science suggests that honesty about one’s limitations, including acknowledging when one doesn’t know something, can foster trust. Admitting that you don’t know the answer but are willing to find out demonstrates honesty, humility, commitment to accuracy, and proactivity.
If the chatbot doesn’t know how to answer a question, could it ask a human or another chatbot how to answer the question?
Human-in-the-loop
Some chatbots are designed with a human-in-the-loop feature. When the bot encounters a situation where it cannot handle the question, it can route the conversation to a live person who can intervene. This is common when accuracy is critical to the response, for example in banking or medical settings. The human operator could either respond directly to the inquiry or guide the bot in how to proceed. If the chatbot has any sort of real-time learning capability, this would enhance the bot performance over time.
Chatbot-in-the-loop
The idea of having a bot seek assistance from another bot is much less common. In theory it is possible to create a network of specialized bots or agents to handle specialty situations. In such a setup, a generalist bot encountering a question outside its expertise could ask other bots on the network for help. This approach would require sophisticated orchestration to ensure effective information exchange.
The concept of bots or systems delegating tasks to humans or to other bots when encountering questions or tasks beyond their capabilities is indeed possible today and has been implemented in various contexts. Bots are already using APIs and Webhooks to interact with external services or databases to retrieve or verify information they cannot process internally. There are even emerging AI systems designed to work collaboratively to share data, insights, and outcomes with each other to solve complex problems.
True innovation will have occurred when advanced language models such as ChatGPT evolve beyond making inaccurate guesses (often referred to as “hallucinating”) and start responding with, “I don’t know, but would you like me to look it up for you?”. Such a level of self-awareness and innovation is expected to be a hallmark of Artificial General Intelligence (AGI), the stage of AI development we haven’t yet reached.
It won’t be long before I’ll have my bot call your bot to work it out.
The Kognetiks Chatbot for WordPress plugin continues to evolve. That evolution is driven by the rapidly growing installed base of users just like you. I want to thank each of you for taking the time to share your valuable feedback, as it helps me continually improve and provide a better experience for all our users.
#ChatGPT #WordPress #WordPressPlugins
About the Author
Stephen Howell is a multifaceted expert with a wealth of experience in technology, business management, and development. He is the innovative mind behind the cutting-edge AI powered Kognetiks Chatbot for WordPress plugin. Utilizing the robust capabilities of OpenAI’s API, this conversational chatbot can dramatically enhance your website’s user engagement. Visit Kognetiks Chatbot for WordPress to explore how to elevate your visitors’ experience, and stay connected with his latest advancements and offerings in the WordPress community.
Leave a Reply
You must be logged in to post a comment.