Chatbots will do more than just chatting

Chatbots can be found on websites, messaging apps, and other platforms, and are used to provide information, answer questions, and perform tasks such as booking tickets or making reservations

Chatbots are not new. “The concept of chatbots, or computer programs designed to simulate conversation with human users, has a long history dating back to the 1950s. Early chatbots were developed as a way to test natural language processing and artificial intelligence technologies. One of the first chatbots was ELIZA, which was created in the 1960s by computer scientist Joseph Weizenbaum. ELIZA was designed to mimic the language patterns of a psychotherapist and could carry on simple conversations with users by responding to keywords and phrases.”

The sentences within quotes above were answered by ChatGPT, the pioneering new chatbot created by OpenAI. The question was about the history of chatbots. 

For anyone researching the history of chatbots, this is a comprehensive answer. Further questions can bring even more data from ChatGPT, which is being recognised as the most advanced chatbot yet. 

ChatGPT has been developed by OpenAI, a research lab backed by Khosla Ventures and Microsoft, among others. GPT is generative pre-trained transformer, a language model that creates a human-like text response. 

How useful have chatbots been so far? This is the answer from ChatGPT: “Over the years, chatbots have evolved and become more sophisticated. Today, chatbots are used in a variety of contexts, including customer service,

e-commerce, and education. They can be found on websites, messaging apps, and other platforms, and are often used to provide information, answer questions, and perform tasks such as booking tickets or making reservations. Some chatbots are even able to hold more complex conversations and use machine learning algorithms to improve their responses over time.”

A debate has already been triggered about the possibility of increasingly sophisticated chatbots becoming the new search engines. This is entirely possible. Currently most chatbots including ChatGPT source their information from a vast yet limited set of information. ChatGPT is limited to information and developments till the end of 2021.

Search engines are current and keep updating information. Chatbots are far more articulate in their response and bring more focus to a query, which a search engine like Google finds difficult to manage. However, ChatGPT’s human-like response doesn’t mention the source of the information, nor does it offer links or citations. 

The synthesised and anonymised information is both useful and tricky, since it will be difficult to confirm the veracity of the information in the chatbot. 

The ease of conversation with ChatGPT has made waves across the world. The chatbot is being tested and trained with thousands of questions and chats being done by enthusiasts. Anyone can login and chat with ChatGPT. Each conversation is research and feedback for the developers. 

There are several frontiers that chatbots have to cross. The data-set that determines the response can keep increasing. It can also be domain-specific for different chatbots. So, a health-care chatbot can be only for patients, while another can be used for students of mathematics. Digital interaction is moving from text to voice, with options of many languages. To go beyond search engines and voice-based assistants like Alexa, chatbots will need greater amounts of information which can be accessed in various languages and voices. 

There is already a level of convergence between chatbots, search engines and digital assistants. They complement the other but are being integrated into easy-to-use solutions. The machine learning and voice recognition technologies are changing rapidly enough to indicate that many business models and functions are ripe for disruption. 

Customer care to personal assistants to search engines could soon evolve and explode with ever smarter chatbots who do more than just chatting.