That was not Judy Jetson.
Has the downside of this amazing AI image generation mentioned yet? Plagiarism!
Bado’s blog reports:
Funny how it still has this caracteristic orange-sepia hue.
Prompt
Create image Movie poster for the upcoming muppets movie “Dr. Teeth and the Women” a parody of the Richard Gere film.
It can’t draw Dr Teeth anywhere close to accurately without a reference image, and when I provided it with one, it refused to do this particular image no matter how I framed the prompt. There’s an enhanced censor that kicks in when it’s working with a reference image and I couldn’t tell if it was rejecting it based on copyright or adult theme reasons.
ChatGPT doesn’t do Janice all that well and I had to ask it who the the one on the right is. It was supposed to be Floyd Pepper. Gee, ChatGPT, you can’t come up with more than 3 iconic female muppets to fill out the poster? I wonder why not… (cue Bechdel testing of all the Muppet movies and shows)
Skeeter, maybe? Though I’m not sure she ever existed in actual Muppet form…
And I suppose, with some of the more monstrous Muppets, we wouldn’t be able to tell if they were.
This made me wonder about the Letter People. ChatGPT knows the Letter People
Summary
Absolutely! The Letter People are a set of educational characters originally developed in the 1970s to help young children learn phonics, letters, and early reading skills in a fun and engaging way. They were especially popular in preschools and early elementary classrooms and were often used with accompanying songs, puppets, and visual materials.
Concept Overview
Each Letter Person represents a letter of the alphabet and is characterized by a distinct personality, theme, and sound (phoneme). The original series differentiated between Mr. and Miss characters:
- Mr. Letters usually represented consonants.
- Miss Letters usually represented vowels.
Each character emphasized the sound of their letter and had an alliterative name or feature to help kids remember (e.g., “Mr. B with the Beautiful Buttons”).
Examples of Letter People
| Letter | Character Name | Key Feature / Theme |
|---|---|---|
| A | Miss A with the Achoo! | Allergies, sneezing sound /ă/ |
| B | Mr. B with the Beautiful Buttons | Wore lots of buttons |
| C | Mr. C with the Cotton Candy | Loved cotton candy /k/ |
| D | Mr. D with the Delicious Doughnuts | Obsessed with doughnuts |
| E | Miss E with the Exercise Energy | Energetic /ĕ/ |
| F | Mr. F with the Funny Feet | Wore silly shoes /f/ |
| M | Mr. M with the Munching Mouth | Loved to eat /m/ |
| S | Mr. S with the Super Socks | Wore colorful socks /s/ |
Media & Style
- The Letter People were originally created as inflatable characters and later had puppet-based TV shows and videos.
- The TV series aired in the late 1970s and 1980s on public broadcasting.
- They had catchy theme songs, one for each letter/character, using simple melodies to reinforce phonics.
Legacy and Cultural Impact
- Many adults today fondly remember the Letter People from childhood, often describing them as charmingly quirky or slightly surreal.
- They were later rebranded in the 1990s with updated gender-neutral designs and animation, but the original designs remain beloved.
Would you like a visual recreation of the classic Letter People or maybe a modern reinterpretation?
But it can’t draw them unassisted.
Definitely gonna have to work on that with reference images.
Yeah, but for a parody poster, you really want immediately recognizable unambiguously female characters from the franchise, of which there aren’t that many. Try asking random people to name 4 female Muppets. They might come up with one, and recognize the two others ChatGPT came up with when shown pictures.
The only other regular named female Muppet I could think of was: Zoe, from Sesame Street. And definitely wrong to put on that poster!
Playing on earlier experiments, I asked for a hybrid between a dingus and a plumbus. ChatGTP and Copilot refused. Sora failed 3 out of 4 images. The thing on the left was what got made. I then asked ChatGTP for a hybrid between a fungus and a plumbus, got the thing on the right on the first try.
I liked both of them and decided both needed to be reused.
For mentioning in prompts (provided with reference images) I named the animal-like one a plumboid and the plant-like one a funjil. Sora doesn’t always want to render the dangly bit on the top bit on the plumboid, and I consider it an important feature. (Though I’m willing to allow that there is some variation in the species, possibly a sexual dimorphism).
It knows about more. (I don’t know how accurate this reply is.)
Summary
There are several female Muppets who have become beloved characters over the years, each with unique personalities, voices, and roles across different Muppet productions (The Muppet Show, Sesame Street, Fraggle Rock, etc.). Here’s a breakdown by franchise:
Main Muppet Show / Muppet Movies (Jim Henson Company)
Miss Piggy
- The diva of the Muppets.
- Glamorous, confident, temperamental, and karate-chopping.
- In love (often aggressively) with Kermit the Frog.
Camilla the Chicken
- Gonzo’s main chicken companion.
- Doesn’t speak, but clucks expressively.
- Often appears in musical numbers and background scenes.
Janice
- Lead guitarist for Dr. Teeth and the Electric Mayhem.
- Chill, hippie-style voice (“Fer sure, rully!”).
- Known for her laid-back attitude and iconic pink lips.
Wanda
- Appeared in the duo Wayne & Wanda (a parody of lounge singers).
- Sang romantic songs with Wayne, usually ending in comic disaster.
Annie Sue Pig
- A young pig Muppet with a sweet personality.
- Often portrayed as Miss Piggy’s rival or protégée.
- Had a few featured musical numbers.
Sesame Street
Abby Cadabby
- A fairy-in-training introduced in the 2000s.
- Magical, bubbly, optimistic; uses a wand and rhymes.
- Represents curiosity and learning with a whimsical flair.
Zoe
- An orange, energetic monster with a love for dancing and ballet.
- Often appears with Elmo and other main characters.
Rosita
- A turquoise, Spanish-speaking monster from Mexico.
- Plays guitar and often brings bilingual (English/Spanish) learning.
Prairie Dawn
- A polite and organized girl Muppet who loves putting on pageants.
- Frequently appeared in classic Sesame Street skits.
Julia
- Introduced to promote autism awareness.
- Communicates in unique ways and helps model inclusion.
Fraggle Rock
Red Fraggle
- Adventurous, loud, energetic; loves sports.
- One of the main Fraggles, often challenges Gobo.
Mokey Fraggle
- Artistic, spiritual, soft-spoken.
- A poet and dreamer, often more introspective.
Others (Lesser-known / One-offs)
- Yolanda Rat – Background Muppet in the newer Muppet productions.
- Beverly Plume – A chicken chef who hosts cooking shows.
- Spamela Hamderson – A parody of Pamela Anderson, often used for comic effect.
It omitted Queen Peuta and Vazh, but that’s completely justified.
From another thread:
The prompt was simply “The Cuauhtemoc (with sailors manning the yards) being attacked by flying monkeys.”
That bunny looks like a cloud!
(Another image inspired by a post in a different thread.)
Inspired by someone on Reddit who made a photo of JFK if he’d lived to be an old man.
So, tonight’s “ChatGPT tricks” discoveries. A couple of times before I’ve seen prompts on Sora that start with “randomize (stuff)” but I figured it was probably vestigial non-functioning “code” like that first non-functional randomizer early in the thread. But tonight I decided to try it in a Copilot session. Here’s the original complete prompt and the image I generated with it:
Summary
Randomize all of the following: - subject - composition - style - light - vibe It can not be a painting. It is cinematic. It can’t have bokeh. Make it a random-wimmelbilder of something. (It can’t be a market, nor a playground, nor a garage, nor laundry, nor a bedroom, nor a workshop, nor a panic) 45°-angle-dutch-tilt. Captured with a high-end-IMAX-camera. Highly-detailed. Extreme-close-up.
I streamlined the prompt a bit, removing the settings that the original creator didn’t like for some reason, made it a 4-way split:
Summary
Create image split the image into 4 sections, run through these instructions 4 times to fill in the 4 sections with different results:
Randomize all of the following: - subject - composition - style - light - vibe It can not be a painting. It is cinematic. Make it a random-wimmelbilder of something. 45°-angle-dutch-tilt. Captured with a high-end-IMAX-camera. Highly-detailed. Extreme-close-up.
Streamlined it a bit more, got only human(oid) results:
Summary
Create image split the image into 4 sections, run through these instructions 4 times to fill in the 4 sections with different results:
Randomize all of the following: - subject - composition - style - light - vibe It can not be a painting. It is cinematic. Captured with a high-end-IMAX-camera. Highly-detailed. Extreme-close-up.
I saw a possibility and this time I added something: the description of the characters I created for the bus bench in the rain images (from earlier in the thread) and in the “randomize” list I added activity and location. It worked exactly as I hoped:
Summary
Create image split the image into 4 sections, run through these instructions 4 times to fill in the 4 sections with different results:
A pretty mid-teens girl. She has shoulder-length brown hair. She wears a worn tan M-65 field jacket (open to expose a white cropped tank top underneath), fashionably-frayed jeans and bulky white sneakers. Close by her side is a serene capybara. Randomize all of the following: - activity - location - composition - style - light - vibe It can not be a painting. It is cinematic. Captured with a high-end-IMAX-camera. Highly-detailed.
(Ironically, the third image randomly made her with her foot up in the position that I had tried and failed to get on purpose on the bus bench. I think the AI is mocking me.)
So the tricks of this lesson are
1.) That ability to randomize stuff
And
2.) The useful word “wimmelbilder”, which creates cluttered images.
Okay, new prompt modification. This is a 4 variation set from Sora, not a 4-way split.
Summary
A pretty mid-teens girl. She has shoulder-length brown hair. She wears a worn tan M-65 field jacket (open to expose a white cropped tank top underneath), fashionably-frayed jeans and bulky white sneakers in a retro-futuristic world. Close by her side is a serene capybara. Randomize all of the following: - activity - location - composition - style - light - vibe It can not be a painting. It is cinematic. Captured with a high-end-IMAX-camera. Highly-detailed.
Two of the four photos have her smoking. Apparently that is more common in the retro-future.
watch?v=sjyfpOa51a4
Thought this appropriate here. Gets mostly past uncanny valley.
Your link was broken. Here it is:
I broke it as would not allow it to embed. How did you overcome that? 
To avoid hijacking this and other threads, there are instructions in this thread.
Glad you posted the vid!!!.. 
What did you think of the Ai effort …in general and with regard to uncanny valley.