The model is based on the company’s latest text-generation GPT-3.5 system released earlier this year. ChatGPT is more conversational than previous versions. It can ask users follow-up questions and refrain from responding to inappropriate inputs instead of just generating text.
ChatGPT is, unfortunately, plagued by the same fundamental issues affecting all current language models: It doesn’t know what it’s talking about. It will therefore generate many generic and quite true, and well formulated, paragraphs, but it is not able to link contextually related and specific information that humans still know about. So understanding its limits is important (as well as the ethical use of the output).
That said, I was able to get it to generate a short piece of Python code from a description I gave it in plain English. It generated a few poems for me that related to people I know. It can also help generate story outlines, do translations for language as well as source code, generate taglines or product promotion social media posts, summarise a long piece of text down to its essentials, explain what a piece of source does, etc.
I can certainly see how AI can assist humans with creative tasks, but its results cannot yet just be used as-is without good proofreading by humans. But try it out for yourself and see what you think.
See https://www.theregister.com/2022/12/03/in_brief_ai/
#technology #AI #artificialintelligence #openAI #ChatGPT