Chat-gpt3 is a natural language processing (NLP) model developed by OpenAI for automated conversation. If you are looking to build something like this, the best way to do so is to use a combination of machine learning and NLP techniques. You will need to collect a large dataset of conversations to train your model on. Some popular approaches are rule-based chatbots, neural network chatbots, and generative chatbots (such as GPT-3). Once you have collected the data, you can use supervised and/or unsupervised machine learning algorithms to train your model. Additionally, you may need to add additional layers of complexity such as bot-specific rules, sentiment detection, and entity extraction. After you have trained your model, you can deploy it to a platform like Slack or Discord.