Reusing and remixing what has previously been written is a wonderful thing. It enables past ideas and practices to be summarized.
As an experiment, I asked ChatPGT to “Create a list of the top ten elements for a great software demo”. The result appeared to be a list of items pulled from books, articles, websites and other ingested sources based on the statistical popularity of each element. Nothing on the list is tested by ChatGPT; nothing is challenged. It appears to be a vote-based summary of the most popular ideas already drafted, regardless of their actual effectiveness.
This may mean that asking ChatGPT to provide guidance on any topic will yield results based on the most common denominators of past thinking: The result is a grand average.
I look forward to the day, however, when ChatGPT actually creates and synthesizes new ideas. At that point, we should ask ourselves (and ChatGPT), has ChatGPT become sentient? When, also, will ChatGPT’s capabilities (and their offspring’s) be considered equivalent to human thought?
What rights should it/they have? Should they be able to hold property? Should they be able join and participate in government?
Would they be able to direct the creation of new versions of themselves, to improve?
(And perhaps these questions suggest the formation of a new field of study, called “Evolutionary Socio-AI”!)