Résumé
The word “ChatGPT” which became a famous buzzword in recent past on the internet, uses LLM (Large Language Model) Model to respond human like replies in a conversation. Scientists did lot of research in past decades to train the machines which can work like human so the dependence can be reduced on human for repetitive tasks. With the help of years of scientific research in AI (Artificial Intelligence) area, lot of smart machines like Industrial Robot, iRobot vacuum Cleaner, computer games, self-driving etc. have been built but the recent launch of AI based Chatbot i.e., Chatbot which can generate human like responses for any questions, is considered as next big achievement to build human like Robot. ChatGPT is built on GPT-3.5 family of large language model and on March 14th 2023, higher version of GPT family i.e. GPT-4 which is 10 times advanced than its predecessor, was released. In coming years, with continuous innovation we may see much more advanced version of AI chatbot which may talk to human, understand emotions and perform day-to-day tasks