Talking to… Lessons in probability theory

Am 16. Dezember 2022 habe ich mit dem GPT-Sprachmodell «gesprochen». Wir haben seitdem und davor viel über diese Software diskutiert. Einer unserer klugen Köpfe in der Radical Dude Society meinte, das Spiel mit dem System und der öffentliche Diskurs sagten mehr über uns Menschen als über die Maschine aus. Das möchte ich sehr gern unterschreiben. In jedem Fall war mein «Erlebnis» mit dem Bot ChatGPT gefühlt so, wie es mir mit allen diesen Bots bislang gegangen ist, ganz gleich, wie gut sie programmiert waren oder sind: Ihre Begrenztheiten tauchen glücklicherweise alsbald auf, und man kann heutzutage trotzdem fasziniert von deren sprachlicher Korrektheit sein. Am Ende bleiben sie dennoch blöd wie Schraubenzieher.

MK: Hallo, have you heard about Mr. Bröno Selfmachteger-Spretz?

GPT: I’m sorry, but I have not heard of Mr. Bröno Selfmachteger-Spretz. As a language model, I do not have access to personal or specific information about individuals. I am a virtual assistant designed to provide general information and assistance to users based on the knowledge and information that I have been trained on. Is there something specific you would like to know or discuss?

MK: I like to know something about your sisters and brothers. For example Eliza. Can you tell me something about the differences between you and her?

GPT: As a virtual assistant, I do not have siblings or sisters and brothers in the traditional sense. I am a large language model trained by OpenAI and do not have personal experiences or relationships. I am simply a program designed to process and generate human-like text based on the data and information that has been provided to me during my training. There are many different types of language models and natural language processing systems that have been developed over the years, each with their own unique characteristics and capabilities. Some of these systems, like Eliza, are designed for specific tasks such as carrying out simulations of human conversation or providing therapy services. Others, like me, are more general-purpose and are trained on a wide range of information in order to assist users with a wide range of tasks and questions.

MK: Nice answer. Well done!

GPT: Thank you for your kind words. I’m here to help, so please don’t hesitate to ask if you have any other questions or need further assistance with anything. I’ll do my best to provide you with accurate and helpful information.

MK: How about the Radical Dude Society. Have you ever heard about that?

GPT: I’m sorry, but I do not have any information about the Radical Dude Society. As a language model, I do not have personal experiences or access to specific information about individuals or organizations. I am a virtual assistant designed to provide general information and assistance to users based on the knowledge and information that I have been trained on. If you have any specific questions about the Radical Dude Society, I recommend doing some research or contacting the organization directly for more information.

MK: Ok, I understand. So the main aspect here is to learn how I try to convince me of your capabilities. Is it?

GPT: As a language model, my primary function is to process and generate human-like text based on the data and information that I have been trained on. My capabilities are determined by the quality and scope of the data that was used to train me, as well as the algorithms and techniques that were used to design and build me. My primary goal is to provide accurate and helpful information to users and assist them with their questions and tasks to the best of my ability. I do not have personal motivations or agendas, and I am not capable of feeling convinced or unconvinced. I simply try to provide the best assistance I can based on the knowledge and information that I have been trained on. If you have any specific questions or need help with something, please don’t hesitate to ask and I’ll do my best to assist you.

MK: Why do you never start sentences with something like «No, you are wrong» or something similar?

GPT: As a language model and virtual assistant, it is not my place to pass judgment on the opinions or beliefs of others. My primary function is to provide accurate and helpful information to users and assist them with their questions and tasks to the best of my ability. I do not have personal feelings or opinions, and I am not capable of feeling right or wrong. In general, it is important for language models and other artificial intelligence systems to be designed and implemented in a way that is neutral and objective, and to avoid expressing personal opinions or biases. This helps to ensure that the information and assistance provided by these systems is as unbiased and impartial as possible.

MK: But if you don’t know the Radical Dude Society, why could you describe your purpose as «provide accurate […] information»?

GPT: As a language model, my purpose is to provide accurate and helpful information to users based on the knowledge and information that I have been trained on. I do not have personal experiences or access to specific information about individuals or organizations, so there may be some topics or questions that I am not able to provide information on. However, I have been trained on a wide range of general knowledge and information, and I am able to provide accurate and helpful answers to many types of questions and tasks. If I am unable to provide an answer to a specific question or task, it may be because I do not have enough information about the topic or because the question is outside the scope of my training. In these cases, I will do my best to provide relevant resources or suggest alternative sources of information that may be able to help.

MK: Ok, thanks for being irrelevant 😉

GPT: I apologize if the information I provided was not helpful in your situation. Please understand that as a language model, I am a virtual assistant designed to provide general information and assistance based on the knowledge and information that I have been trained on. I do not have personal experiences or access to specific information about individuals or organizations, and there may be some questions or topics that are outside the scope of my training. If you have any specific questions or need help with something, please don’t hesitate to ask and I’ll do my best to assist you to the best of my ability. I am here to help, so please let me know if there is anything else I can do for you.