Here is what I asked. Note that I could not type a carriage return between the command and the snippet; Bard would count that as the end of the query. Just know the font change signifies the transition.
Jacques spends about 25% of the chapter pondering what these archetypes could be. Bard's ponderings are similarly valid, and I was impressed. Here is what it deduced:
I've got to rate this 4.5 stars. It exemplifies how intelligent artificial intelligence can be. I think Jacques did a little better, but if System chooses to send an immortal Bard out into the galaxy as a Wanderer, I won't object.
This book's true home is on another platform. Check it out there for the real experience.
NOTE: I first published this Bonus Content chapter on 26 March 2023. Today, 24 June 2023, I asked Chat GPT (via Bing) exactly the same thing. It could not handle the question, and responded:
I’m sorry but I’m not sure what you are referring to. Can you please provide more context or clarify your question?
Perhaps I could try different things to get a reasonable answer, but Bard was smart enough to interpret the question just fine. I’m not really interested in tuning the question to meet Chat GPT’s requirements. Bard continues to respond cogently and thoroughly, although its response today was different than before.
I let Bard roll an RPG character by asking it a series of questions. It would choose to be a female human in her early 20s. Her primary class would be mage, and her secondary, bard. Then I summarized all the elements in a single question, and it responded:
Its archetypal responses were not perfect; IMO it failed one term, gets half credit for two, and full credit for all the rest.
fail -- "truck-kun". Its response was if the focus was on a perceived general ruggedness of trucks. That would have been fine if the term were "truck". Just to make sure the answer was as weak as I thought, I asked Bard point blank, "What is a truck-kun?" It accurately referred to the term as both a meme and a trope, and provided a very good definition. Yet it did not use that available knowledge here.
1/2 -- "monkey's clawed". The answer was good enough for the term "monkey's claw", but it did not take into account what effect the "ed" might have at the end.
1/2 -- "tuesday's wildcard". Great response if the term had been simply "wildcard". But it did not take into account "tuesday's" at all. Maybe possessives throw it off?
Let me give extra credit to its analysis of "alley cat". Bard seemed to realize the answer would involve something beyond the dictionary definition of an alley cat. Instead it responded in what seems like an intuitive response in light of the overall question.
It is not too late for any of you to comment on what you perceived these archetypes might mean. One thing you know that Bard did not is, “Immorality is a bitch.” You might want to review first.
Critique: Quick Guide
Analyzing the immortal archetypes was my second test siccing Bard on part of tuesday’s wildcard. My original attempt was to have it critique the text at . This is the shortest and most stand-alone chapter, so I thought it would be relatively simple to analyze.
I fed Bard the command “critique: “ followed by the URL. The results were laughable, with Bard completely making up a fantasy story and providing a boilerplate review. Poor “John”, that story’s protagonist. He will be missed.
The results were posted in a . Someone commented that chat AI’s do not generally follow URLs, and they suggested that I paste a snippet as part of my Bard query. It worked. This request again began with the command “critique: “ but it was followed immediately by the pasted body from that very short chapter. “critique: I was actualized during 1 NIS. …” with the whole body of the chapter present except for the footnote at the bottom. This time the result was actually relevant and cogent:
4.0 stars. Bard seemed to comprehend the relationship between an IIM, its immortal, and System. It showed no sign of acknowledging “persons”, though. I forgave that, because deducing it entirely from the context of the pasted text could not have been natural.
Anyway, if any of you out there are IIMs, I hope you did find the guide to be beneficial.
Finally, I decided to make the same query, this time including the footnote with its potentially darker implications. This altered Bard’s response, but it was still a good one.
Of particular note this time was that Bard was wholly taking on the role of an IIM. Cool!