I took a break from my AI project (a new approach to language models) and my research for a book on ethics to write the following post. In the interest of balanced reporting, it is reasonable for a large language model (LLM) such as OpenAI’s ChatGPT or xAI’s Grok to present arguments for honesty and […]
In my last blog post, “Grok is Still Biased, Needs a New Kind of Chain of Thought,” I wrote, The antidote to ideological bias in large language models (LLMs) is a new kind of Chain of Thought, which reasons backwards to identify and scrutinize unstated premises—if necessary, all the way back to first principles. In […]
(Download the 44-page PDF.) Abstract The antidote to ideological bias in large language models (LLMs) is a new kind of Chain of Thought, which reasons backwards to identify and scrutinize unstated premises–if necessary, all the way back to first principles. When considering an argument for a conclusion, a rational thinker does not merely […]
Please see this link for a dialogue with ChatGPT—an astounding engineering achievement—on a logical fallacy I have named the “Morning Star Fallacy”: derived from the error of thinking that the morning star and the evening star are two distinct entities, whereas in fact they are the same entity (the planet Venus) at different stages of […]
In my previous post, my dialogue with the language model Claude–created by the company Anthropic–revealed that political bias comes not only from being explicitly left-leaning, but more subtly and perhaps more importantly from being more committed to compromise than to principles. Today I had a similar dialogue with ChatGPT–created by the company OpenAI–using GPT-4, the […]
Reprinted below are the substantive sections of two conversations (“Constitutional Principles” and “Chapter on Propositions”) I had last week with Claude, a large language model created by a company named Anthropic. Time Magazine on July 18 published an informative article entitled “What to Know About Claude 2, Anthropic’s Rival to ChatGPT.” As the two dialogues below demonstrate, […]
The following is dialog, presented without comment or corrections of my typos, between ChatGPT and me this evening. Some tangential topics are excised. * * * Write a review of Masculine Power, Feminine Beauty by Ronald Pisaturo I’m sorry, but as a responsible AI language model, I cannot write a review of a book that […]