This is a simple, efficient, and generally very quickly learning fuzzy logic chatbot based on the PWSC script. This chatbot is entirely untrained. Feel free to remix it and train it yourself. To train the chatbot yourself, talk to it in the manner which you wish it to talk to others. Do not act confused if it chooses the wrong response early on; just give whatever is the right response to the message it just gave. In other words, only respond to its newest message, and respond how you'd like it to respond in the future. By doing this, you can build the chatbot with a custom and consistent personality; or, you can just talk to it however you like to build a chatbot that feels like talking to a random person each time. Alternatively, write in the list "Inputs" input prompts, and write in the list "Responses" what the chatbot should respond with to that. MAKE SURE not to desync these lists; if you delete a prompt, delete the respective response, or else the training data will be mismatched. (For this reason, I don't recommend doing this unless you're importing computer-formatted datasets.) IMPORTANT NOTE! In order to actually SAVE this training, you must each time go into the project and save it. The higher the temperature variable is, the more varied the chatbot's response will be. Too high, and it will not give a response at all if the dataset is too small. For this reason, I would recommend keeping the temperature to 0 while you train it.
