Jan MatusiewiczAGI safety — discourse clarificationThe topic of AGI (Artificial General Intelligence) and its safety is becoming increasingly visible. Different positions on this matter may…Dec 2, 2023Dec 2, 2023
Jan MatusiewiczDreamStudio from Stability AI — tutorialDreamStudio is an online AI tool powered by Stable Diffusion allows you to make these amazing transformations of one photo into something…Jul 31, 2023Jul 31, 2023
Jan MatusiewiczHow to solve alignment in AGI.The recent boom in AI has caused many to worry not only about short-term risks, which I have described in this post, but also about…Jun 20, 2023Jun 20, 2023
Jan MatusiewiczMisplaced concerns: real risks of chatbotsThere is a significant amount of concern regarding the potential misuse of chatbots, such as ChatGPT, Bing’s chat, Claude, and Bard, by bad…Mar 27, 2023Mar 27, 2023
Jan MatusiewiczCommon sense reasoning: GPT-4 vs GPT-3.5OpenAI ChatGPT allows now for testing GPT-4, and not just GPT-3.5. Evaluating change of capabilities of LLM is a huge task so a single post…Mar 16, 2023Mar 16, 2023
Jan MatusiewiczLanguage and thought —ChatGPT vs GPT-3I was thrilled to read a thorough article “Dissociating language and thought in large language models: a cognitive perspective”…Feb 27, 20231Feb 27, 20231
Jan MatusiewiczCould ChatGPT be a millionaire?I tested ChatGPT against questions from one game of Who Wants To Be A Millionaire to check if it could win the game. It answered 11…Jan 12, 2023Jan 12, 2023
Jan MatusiewiczTesting memory capabilities of ChatGPTI tested the memory capabilities of ChatGPT by feeding it with random data and checking if it remembers them.Dec 26, 2022Dec 26, 2022
Jan MatusiewiczLearning languages with ChatGPT as a conversation partnerEver wanted to practice foreign language with a native who is limitlessly patient and will talk with you about whatever you want? ChatGPT…Dec 25, 2022Dec 25, 2022