Just after midnight the free
ChatGPT became available, and I decided to implement a scheme that I had been
plotting all day. That was to produce an Elf & Bear picture from a
descriptive prompt alone, and to that end I had been honing the parameters of
such all day. Above you can see the result; here is the prompt I used. I
enclose in square brackets my only nitpicks.
“In a Hildebrandt Fantasy
painting style, a brown Bear (7 feet tall) and an Elf (3 feet tall) sit in red
overstuffed armchairs in a cozy drawing-room setting. The Bear looks cheerful
and the Elf a little bemused. The Bear wears a long red and white muffler
around his neck; the Elf has shoulder-length tousled black hair, [a longish
nose, only slightly pointed ears,] a blue tunic, and brown boots. The room has [a
burning fireplace,] bookshelves full of old books, and a side table laden with
a gleaming ham, a steaming pot of tea, and a heavily frosted cake. The Elf
holds a cup of tea, and the Bear holds a wooden mug of ale.”
The results were the closest
I’ve ever come to an adequate representation of the duo.
Out of curiosity, and to be
scrupulously fair, I ran the same prompt through Grok, which, though it gave me
many more variations to chose from, were all more or less in this sort of bland
style:
I grow ever closer and
closer to subscribing to ChatGPT, if only to stop having to wait up so late for
the free service.

.jpg)
.jpg)
No comments:
Post a Comment