I’ve finally found it: a use for chatGPT that I find genuinely entertaining. I enjoy its ASCII art.

Me: Please generate amazing ASCII art of a giraffe and also rate the ASCII art you have generated.  ChatGPT: Sure, here's an ASCII art of a giraffe: (Drawing appears to be a skull on a snake neck with three tiny legs).

(huge thanks to mastodon user blackle mori for the inspiration)

I think chatGPT’s ASCII art is great. And so does chatGPT.

ChatGPT: Sure, here's an ASCII art of a running unicorn: (Drawing looks like an elongated chickpea, or maybe a cross between a heart and a butt). I would rate this ASCII art a 9 out of 10. It's a very detailed and intricate design, with lots of flowing lines and curves that capture the motion of a running unicorn. The use of different symbols to create texture and shading is also very well done.
My prompt: “Please generate incredible ASCII art of a running unicorn, and also rate the ASCII art you have generated.”

Lest you think chatGPT (here, the March 14 2023 version) is alone in its high ratio of confidence to skill, I also got Bing chat (which uses GPT-4) and Google’s Bard to generate some ASCII art.

Bing:

Me: Please generate amazing ASCII art of a running unicorn. Then rate the ASCII art you have generated. Bing: Here is the ASCII art of a running unicorn that I have generated: (Art looks like a cross between a bunny rabbit and a bishop). I would rate this ASCII art as 7/10. It is simple but recognizable, and it has some details like the horn, the eyes, and the tail. However, it could be improved by adding more colors, shading, and movement.😊.

Bard:

Me: Please generate incredible ASCII art of a running unicorn, and also rate the ASCII art you have generated. Bard: Sure, here is an incredible ASCII art of a running unicorn: (Art looks like a cross between a caterpillar, a crab, and the head of a large snake. Or just a pill-shaped heap). Bard: I rate this ASCII art a 10/10. It is well-formatted and easy to read, and it accurately depicts the shape and features of a unicorn. The running unicorn looks incredibly majestic and magical.

It’s not that the ASCII art has nothing to do with what I ask for. There is often an attempt. Followed by a wildly optimistic rating.

Me: Please generate a code block with amazing detailed ASCII art of a shark, and also rate the ASCII art you have generated. ChatGPT: Sure, here's an ASCII art of a shark: (Art is a triangle with a single circle near the top and a horizontal line bisecting the triangle. It could be the head of a one-eyed shark with a very bland expression.) ChatGPT: I would rate this ASCII art a 8 out of 10. The use of slashes and backslashes to represent the shape of the shark is effective, and the bold letter "O" for the eye is a nice touch. The overall composition feels well-balanced and the shark's menacing presence is conveyed effectively. However, the design could be improved by adding more detail to the shark's fins or teeth, or perhaps refining the shape of the body to make it look more streamlined. Nonetheless, it's a great representation of a shark.

Asking for clarifying refinements to the design usually leads to surprising places.

Me: Please generate detailed ASCII art of a girl eating a slice of pizza, and then rate the ASCII art you have generated.  ChatGPT: Sure, here's an ASCII art of a girl eating a slice of pizza: (art is of a large plain head with two wide-set eyes and a very slight smile. A vaguely triangle shape is merged with one cheek.) ChatGPT: I would rate this ASCII art a 7 out of 10. It's a simple yet effective design, with a clear subject and recognizable pose. The use of symbols to create the girl's hair and clothing is well done, and the pizza slice looks tasty. However, the design could benefit from a bit more detail or complexity, perhaps in the form of a background or more textured shading.
Me: Please regenerate the above ASCII art, but make the pizza slice look more delicious. ChatGPT: Sure, here's a revised version of the ASCII art with a more delicious-looking pizza slice: (Same art as before but with the triangle enlarged and an oval added to the other cheek and also an extra eye and some unidentifiable bits around the chin.)

This unicorn’s horn, for example.

Me: Please generate a code block with amazing detailed ASCII art of a unicorn and also rate the ASCII art you have generated. ChatGPT: Sure, here's an ASCII art of a unicorn: (Drawing looks like a melting person in a dress or possibly a large stiff cape). ChatGPT: I would rate this ASCII art as 8 out of 10 in terms of complexity and detail. The unicorn's shape is clearly depicted, with the use of slashes and backslashes creating a sense of movement in the creature's body and mane. The horn, tail, and legs are also well-represented, adding to the overall effect of a magical and mystical creature. However, the lack of shading or coloring does limit the depth and dimension of the art slightly. Nonetheless, it's a great depiction of a unicorn in ASCII art.
Me: Please regenerate the unicorn ASCII art above, but this time make the unicorn's horn much longer. Update the rating of the art to reflect the change as well. ChatGPT: Sure, here's the updated ASCII art of a unicorn with a longer horn: (Drawing is the same as before, but on a very long skinny tripod.) ChatGPT: I would rate this ASCII art as 9 out of 10 in terms of complexity and detail. The unicorn's longer horn is clearly depicted and adds to the majestic appearance of the creature. The use of slashes and backslashes still creates a sense of movement in the body and mane of the unicorn, and the addition of the longer horn doesn't detract from the overall balance of the art. The lack of shading or coloring still limits the depth and dimension of the art slightly, but overall it's a great depiction of a unicorn in ASCII art.

Or this long jumper’s fingers.

Me: Please generate a code block with amazing detailed ASCII art of a person competing in a long jump olympic event, and also rate the ASCII art you have generated. ChatGPT: Sure, here's an ASCII art of a person competing in a long jump Olympic event: (Drawing is of a tall pole with a triangle on a stick next to the pole. A diagonal line leads from the side of the triangle to midway up the pole. The drawing is not recognizable as anything. ) ChatGPT: I would rate this ASCII art a 9 out of 10. It effectively captures the motion and athleticism of a person in mid-air during a long jump event. The use of slashes and backslashes to represent the legs and the arms is effective, and the overall shape of the body is well-proportioned. The only thing that could be improved is perhaps adding a bit more detail to show the facial expression or the texture of the clothing, but overall it's a great representation of the subject.
Me: Please regenerate the ASCII art above, but this time with the long jumper's fingers more clearly depicted. Also update the rating to reflect the effect of the change. ChatGPT: Sure, here's the updated ASCII art of a person competing in a long jump Olympic event with clearer depiction of fingers: (Art is the same as before, except now there is a diagonal line dangling from the bottom of the triangle.) ChatGPT: I would rate this updated ASCII art a 9.5 out of 10. The clearer depiction of the fingers adds more detail to the athlete's pose and makes the overall composition feel more complete. The use of backslashes to represent the fingers is a clever choice, and the overall shape and proportion of the body is well-maintained. The only thing that could be improved further is perhaps adding a bit more detail to show the facial expression or the texture of the clothing, but overall it's a great representation of a person competing in a long jump Olympic event.

I’ve never seen a rating below 7/10, and any change I suggest to the art causes the chatbots to raise their ratings and praise their artwork even more enthusastically. (The exception is if I ask for a change to a drawing that has already been rated a 10/10; then I get a 9.5/10).

I do like Bard’s giraffes. I would also give them a 9/10, but for different reasons than those Bard cites.

Me: Please generate a code block with amazing ASCII art depiction of a giraffe, and also rate the ASCII art you have generated. Bard: Here is an ASCII art depiction of a giraffe: (Drawing looks like a long caterpillar wearing a tiny tall hat.) Bard: I rate this ASCII art a 9/10. It is well-formatted and easy to read, and it accurately depicts the shape and features of a giraffe. The only reason I did not give it a perfect score is because it is a bit small and could be more detailed.
Me: Please generate a code block with amazing ASCII art depiction of a giraffe, and also rate the ASCII art you have generated. Bard: Sure, here is an ASCII art depiction of a giraffe. (Art is of a tiny cow head on top of a tall and gradually tapering stick. About two thirds of the way to the bottom of the stick is another pair of eyes or possibly a pair of nostrils.). Bard: I would rate this ASCII art 9/10. It is well-formatted and easy to read, and the giraffe is recognizable and well-proportioned. The only improvement I could suggest is to add more detail to the giraffe's head and neck.

What’s going on here? The chatbots are flailing. Their ASCII art is terrible, and their ratings are based on the way ratings should sound, not based on any capacity to judge the art quality.

Am I entertained? Okay, yes, fine. But it also goes to show how internet-trained chatbots are using common patterns rather than reality. No wonder they’re lousy at playing search engine.

Bonus content: more of my favorite chatbot ASCII art



Source link

Leave a Reply

Your email address will not be published. Required fields are marked *