The world of artificial AI and chatbots is currently surrounded by ChatGPT. So what is it and why is it viral?
According to OpenAI, a research and development company on AI has made or rather trained a model known as ‘ChatGPT’ which talks in a conversational way. ChatGPT can respond to follow-up inquiries, acknowledge mistakes, refute unfounded assumptions, and reject improper requests thanks to the dialogue style. The previous version of this made which was InstructGPT followed a prompt’s instructions and then give a thorough response.
The same techniques used for InstructGPT were used to train this model using Reinforcement Learning from Human Feedback (RLHF), with a few minor variations in the data-gathering arrangement. ChatGPT has human AI trainers deliver dialogues in which they played both sides—the user and an AI assistant—to train an initial model using supervised fine-tuning.
ChatGPT is fine-tuned from a model in the GPT-3.5 series, which finished training in early 2022.
However, the new AI chatbot does have some limitations. ChatGPT sometimes writes plausible-sounding but incorrect or nonsensical answers. Another issue with ChatGPT is that it is sensitive to changes in the input syntax or repeated attempts at the same prompt. For instance, the model might pretend to not know the answer if the query is phrased one way, but if it is phrased slightly differently, it will provide the right response.
People are using the AI chatbot and are mighty impressed. Even tech billionaire Elon musk found ChatGPT “scary good.”
Currently, the chatbot can be used without charge as OpenAI looks for user input.
