Conversational AI & Chatbots, the Process: Since 2013, scientists have been looking for ML-related models and computational neural networks that can train a smarter and “humanized” Chatbot for natural language processing (NLP). Luckily, in 2018, NLP models came to its historical breakthrough from solely words or features model training (like Word Embedding) to deep-contextualized learning. Models such as ELMo, Google BERT, and Baidu ERNIE are proved to be more accurate and close to human communication. (More details of the process of human language and NLP-model development (Mandarin Chinese))
Notably, this tech-competition of AI & human language models within international brands fires up the use of automated and personalized communications in messaging apps, chatbots, speech assistants.
Thus the so-called Conversational AI is all the globe around.
However, there is a standstill lying behind chatbot usages since 2018: People leverage Chatbots only as 24/7 message providers, or information collectors. What does it entail?
Challenge 2. We’ve feared the unknown as Chatbot emerged longer than we face the uncertainty since corona-outbreak.
In this century of Conversation AI, we’ve been in the incredulity and asking, “Can Chatbots/Bots replace human interactions?”
Limited potential for Conversation AI development:
Chatbots act As a Service for solving instant queries from consumers, customers, or repeated questions. In other words, we doubt the possibility of Chatbot in real communications, and therefore we made Chatbot a replacement for customer service or any other dull, repetitive works with SOP.
Insufficient imagination for Chatbot application:
As the WHO Bot and UK WhatsApp COVID-19 Bot focus on pandemic information clearance, we’ve been thankful and relieved to see that Chatbot can save the situations for its privacy in messaging and accuracy in synchronical updates. On the contrary, its usages as a one-directory broadcaster or a message transmitter also tell us the insufficiency of Chatbot usages when it comes to human communications.
Chance 2. Go beyond Chatbots, establishing a trustworthy Chatbot ecosystem as soon as possible.
Start an open-minded discussion, and grab your flashes of any inspirational Chatbot idea right away.
You can find an easy-setup Chatbot framework to look for its pros and cons. Only when we face the reality of the Chatbot industry (from small businesses such as Bottender to enterprises like LUIS.ai) can we thoroughly acquire the current state/trends of Conversation AI.
Write, illustrate, and share information in open development communities.
No more postponing expediency. Humans have no excuses for delaying the discussion-based sufficient time saved by Conversational AI.
Contrary to self-isolation, now is the time for us to step into the unknown of Conversation AI, making Chatbot, not an alternative, but a accompany in human interactions.
