Style Guide for presenting sets of completions in print

I had to come up with a way of representing sets of completions in print for a forthcoming book by @MarcStrassman – this is what I came up with.

Algorithmically generated texts that are generated as a set, such as OpenAI completions, should be represented using mathematical set notation, thus.

{ Topic or Title of the Set, Uc/lc }

Text of the completion.

The topic is bold, centered, title case, with spaces after the opening bracket and before the closing one.

Prompts should be provided in bold, flush left.

Each completion should be preceded by a stylized version of the Greek letter epsilon, which signifies that the ompletion is a member of the set. The epsilon is in regular weight text and precedes every completion in a set, including the first. The epsilon should not immediately precede a page break.

It is important for me to tell you that the symbol ∈ was first used by Giuseppe Peano in his 1889 work Arithmetices principia, nova methodo exposita, at page x.

Signum ∈ significat est. Ita a ∈ b legitur a est quoddam b; …

Texts that are generated for use one at a time should be represented normally, i.e. in the mode appropriate for the style they are used in–body text if unquoted, block quote if block quoted.

1 Like

That’s a fair concern, but it’s for that very reason that it’s always a human author’s byline!

1 Like

Also, I should explain that this portion of the style guide does not apply to representing all published OpenAI completions by any means. This simply reflects a particular approach that Marc has taken, which is to generate multiple completions and present them on equal terms. Other authors will simply “select the best” and weave together a single narrative without alternate completions, in which case no special notation is needed.