Sztuczna inteligencja, prawo i etyka
Artificial Intelligence, Law, and Ethics

Summary/Abstract: Artificial intelligence has been taking (and passing) law exams at universities, writing scientific articles on legal problems – in collaboration with academics – or preparing submissions, statements of claims, and contracts at lightning speed. And all this happening within weeks of the launch (November 2022) of the bestknown AI application called ChatGPT (Generative Pre-trained Transformer), made available for free by the US company OpenAI. It is a conversational chatbot based on large language models making use of deep learning. Its resources include almost unlimited data sets from all over the world, excluding current information from the Internet because the system is – at least for now – closed, limited, and able to establish neural connections only between what it already knows. It generates new content based on the resources it has access to, using logical semantic connections and the mathematical probability of occurrence of a given word in a given sequence. Its statistical model can generate differently worded, correct content each time – even in response to the same question. Hence the first natural attempts of pupils and students to use the tool to write assignments and theses.

• Details
• Contents