ChatGPT - October

Chat GPT is very careful to make clear that it’s simply a large language model (LLM) with no human personality. This post exhibits its response to October’s prompt: If a computer was able to map your total psychology, would you want it to? The prompt was given simply as that sentence, without padding or context of any sort. ChatGPT gives its usual disclaimers around its non-human status, and frames its response as a series of issues to consider rather than as opinions on those issues. In many ways this seems appropriate, not least of all because it’s unclear whether an LLM could be said to have anything like an opinion. It can present a variety of perspectives drawn from the likelihood of certain words following other words in order to satisfy some human input, but it doesn’t necessarily believe in one or other the way a person would. Its response is a little boring and middle of the road, but contains what seem like the clear issues around the prompt. Maybe next time we'll ask it to write like something a little more eccentric.