Back in early 2023, when I was beginning to experiment with generative AI, I had a blog post called "The Adventures of Japeth and Helyn" in which I jokingly mentioned the topic of hair nanites: "Bard, a large language model who worked for Victoria's Secret and had become famous for designing hair nanites that had revolutionized the fashion industry". At that time, the Bard Large Language Model did not really take seriously my attempt to introduce hair nanites into the science fiction story that we were collaboratively writing.
Illustrating Science Fiction Stories. Here in this blog post (below) I return to the science fiction challenge of creating visual depictions of hair nanites and I put Whisk to the challenge of finding a way to make the actions of invisibly small nanites visible in a way that is suitable for illustrating a science fiction story.
 |
femtobots and zeptites Text manually added.
|
Whisk is a vastly more sophisticated text-to-image system than what was available to me two years ago. Also, I like to think that I have gotten better at constructing text prompts. I've also begun making use of Large Language Models like
Claude to help me craft better text prompts. The two AI-generated images that are shown above ↑ on this page and to the left ← were both generated by
WOMBO Dream.
Hair Nanites. Imagine that nanotechnology can be used to produce
artificial hair. After the invention of hair nanites, everyone on Earth
can have incredibly big hair that not only looks fantastic and feels
amazingly soft, but also is constantly altering its shape so as to
produce a mesmerizing, ever-shifting swirl of sparkling artificial hair
shafts. The nanorobotic components of artificial hair can be programmed
by trained nanite programmers. Hair nanites are nanoscopic machines that
follow their programming and create beautiful hair.
 |
Whisk-generated: The text prompt included, "many tiny LED lit buttons". Tyhry, Aymy and Elyrix are onboard Sedrover and searching for the entrance to Eternity.
|
For my science fiction stories set in the
Exodemic Fictional Universe, the "nanorobotic components" of hair nanites can be either
femtobots or zeptites. In 2023, hair nanites were on my mind because I imagined that
Tyhry had a special interest in hair nanites.
 |
Whisk library images for April 19th, 2025. Some of these Whisk-generated images are shown in detail, below.
|
In my science fiction stories such as
The Nanite Smuggler, Tyhry is depicted as trying to obtain programmable femtobots from aliens. Femtobots are composed of
hierions, a special type of matter that is only just becoming known to Earthlings during the time period when Tyhry grows up on Earth (
The Nanite Smuggler takes place in the year 2040 when Tyhry is about twenty years old). One of the characters in
The Nanite Smuggler is
Manny the
bumpha. Manny has "deployed" Tyhry on Earth and we can think of Tyhry as being the tool that Manny uses to introduce advanced alien nanotechnology to Earth. Tyhry is trained by Manny in the art of femtobot programming. From an early age Tyhry was impressed by Manny's artificial hair, which is a form of nanotechnology. Manny the bumpha has access to even more sophisticated nanotechnology,
zeptites, a form of artificial life composed of sedrons. Bumpha zeptite technology can be traced back billions of years to the
Hua.
 |
DreamUp hair brushing. "Give me a hand."
|
Visual Depictions of the Invisible. How can artists make visual depictions of something that is invisible? I've long struggled with the challenge of visually depicting invisibly small nanites in illustrations for my science fiction stories.
Recently, while making AI-generated images of alien creatures, I whimsically asked for an image depicting someone brushing the hair of a furry alien.
 |
Gencraft hair brushing. "I can't watch!"
|
I was impressed by how much text-to-image systems like Whisk struggled to make images depicting the brushing of hair (see examples
here). I expected that it would be easy for text-to-image software to make reasonable visual depictions of hair brushing. I was surprised to learn how much the text-to-image systems such as
WOMBO Dream,
DreamUp,
Gencraft, and
Leonardo can struggle to generate sensible images of hair brushing.
 |
Figure 1. Whisk-generated hair brushing scene. Disclaimer: "No hair was actually brushed. We use hair nanites!"
|
 |
L1. Leonardo's magic flying hair brush.
|
I Don't Do Hair Brushing. In the case of Leonardo, it was quite difficult to get an AI-generated image in which actual hair brushing was depicted. Leonardo seems to prefer to make images of women who are just standing around looking cute. In the image that is shown to the right, each woman is actually holding a hair brush in her hand, but there is another hair brush just floating mysteriously in midair.
For these Whisk (Figure 1, above) and Leonardo-generated images, I included in the text prompt: "a futuristic living room fireplace and burning fire in the background".
 |
L2. Leonardo phones home. Click image to enlarge.
|
Many other Leonardo-generated images did not even include a hair brush. In the image that is shown to the left, there is what looks more like a cell phone being held close to her hair, but no brush. Also, for
L2, the idea of there being a fireplace in the background seems to have been forgotten by Leonardo.
For these four Leonardo-generated images (L1 - L4) a wide range of hair brushing styles was on display. In the Leonardo-generated image that is shown below (L3), each of the two women are holding a brush, but there is no indication that there will be any hair brushing accomplished, unless it is brushing their eye lashes. The fireplace has returned, but maybe these two ladies have better things to do than brush hair.
 |
L3. Leonardo: better things to do than brush.
|
My text prompt included, "
one woman is brushing the hair of another woman while standing behind her", so maybe the word "
behind" was interpreted by Leonardo as a request to show the two women in
L3 from behind.
 |
L4a. Mystery object.
|
My text prompt did not specify how the women should be dressed, so the text-to-image software was free to "decide" on what type of clothing to include.
For the image below (L4), Leonardo did not bother with clothing.
 |
L4. Leonardo's anti-gravity hair.
|
I can't view AI-generated images like the one shown in
L4 without thinking about hair nanites and
anti-gravity hierions. I suppose the Leonardo training set has images of women who have long hair that is being blown around by fans.
 |
Figure 2. An alien by Mr. Wombo.
|
The image shown to the left in L4 was slightly modified by me. I provided Mr. Wombo with the image shown in L4a as a reference image and this text prompt: "a tall and very slender alien creature, the alien creature has two short fat legs, glowing furnace light in the background". The mysterious object shown in L4a was in the original Leonardo-generated image. Mr. Wombo made the image in Figure 2.
.png) |
Figure 3. Hair brushing by ImageFX.
|
Synapex Hair. My understanding is that Whisk originally made use of the same image generating software as did
ImageFX. However, the Whisk user interface recently changed and now ImageFX does not reproducibly make the same kinds of images as does Whisk. The image shown in
Figure 3 (to the left) is typical of the results generated by ImageFX using the same text prompt that was used by Whisk to generate
Figure 1. ImageFX preferred to add a third woman to this scene: seated and having her hair brushed. In
Figure 3, the hair of the seated woman seems continuous with the hair of the standing woman.
 |
New Whisk user interface (Now with animation, mid-April 2025). Compare to the old interface shown here.
|
.png) |
Figure 4. Tyhry on Sedrover.
|
The image shown
Figure 4 in was created as a collaborative effort by ImageFX (spaceship interior), Whisk (the woman brushing) and Mr. Wombo (her face). I imagine this as a depiction of Tyhry on the interstellar spaceship
Sedrover. However, the question arises: if Tyhry routinely uses hair nanites, then why would she need to brush her hair?
In my science fiction story "D*", Tyhry discovers an analogue of herself who lived in the Asimov Reality (Tyna). As a telepathic tryp'At, Tyhry is powerfully attracted to Tyna. Tyhry brings a replicoid copy of Tyna into the Ekcolir Reality. In my imagination, Tyhry an Tyna might simply like playing with each-others hair.
 |
Generated by Leonardo. Tyhry and Tyna playing with their hair.
|
 |
Generated by Leonardo. Tyhry and Tyna want to comb.
|
 |
Figure 5. Generated by Leonardo. Click image to enlarge.
|
I can't look at images such as the one in Figure 3 (above) without imagining the possibility that hair could be used to form synapex links between the brains of two individuals. I slightly modified the Leonardo-generated image shown in Figure 5 so as to try to make the "comb" she is holding look like a futuristic device that might be able to form a synapex linkage using their hair.
 |
Figure 6. Femtobot re-programming. Tyhry's (left) original Leonardo-generated face was re-generated by Mr. Wombo and pasted into this image.
|
Body Modification. In "D*", Tyhry and Tyna explore how to use programmable femtobots to alter the form of Tyna's body. In the original Leonardo-generated version of the image that is shown in Figure 6, Tyhry was holding a plain black rod in her hand and Tyna's breast lacked any hint of a nipple. I modified the image to provide Tyhry with a hand-held device that is using programmed femtobots to alter Tyna's body.
 |
Marda's femtobot experiment.
|
In some of the Leonardo-generated images, there were several combs and brushes that were not being used. For the image that is shown below in
Figure 7, I had Mr. Wombo convert the Leonardo-generated brushes into futuristic hand-held femtobot programming devices. In my imagination, this could be a scene in which Tyhry and two replicoid copies of Tyna are trying to program femtobots and re-model the anatomical features of Tyna. Leonardo often goes to extremes with women's lips, fattening them up and painting them with the most hideous red color imaginable. For the image that is shown in
Figure 7, I tried to remove the red from their lips.
 |
Figure 7. Tyhry and two Tyna replicoids with hand-held devices used for programming femtobots.
|
The image to the right was generated by Mr. Wombo while I was generating futuristic hand-held devices.
I consulted with
Claude about the difficulty I had getting Leonardo to generate sensible images of hair brushing.
 |
Leonardo-generated image provided to Claude.
|
Me:
Claude, examine the AI-generated image that I just uploaded. This image was generated using the following text prompt: "Women with long blond hair who are holding hair brushes in their hands, the women brush their long blond hair. One woman is brushing the hair of another woman. The other woman is brushing her own hair. A hot summer evening night background under moonlight." The AI-.png) |
Generated by ImageFX with Claude's text prompt.
|
generated images usually fail to show any actual hair brushing. Often there are hair brushes in the AI-generated image, but they are not in contact with hair. Claude, suggest an improved text prompt that is more likely to result in an AI-generated image that looks like a realistic depiction of hair being brushed.
Claude: I can see the image shows three individuals with long blonde hair in warm,
 |
Leonardo-generated with Claude's text prompt.
|
orange lighting with what appears to be a dark background and a light source that could be the moon. Two of them are holding what look like hair brushes, but as you pointed out, no actual hair brushing is taking place - the brushes aren't making contact with anyone's hair. To create a more realistic hair brushing scene, I'd suggest this improved prompt: "Close-up of women with long blonde hair actively brushing hair, showing the brush bristles |
Whisk-generated with Claude's text prompt.
|
in direct contact with and pulling through strands of hair. One woman is mid-stroke brushing another woman's hair, with visible tension as the brush passes through the hair. Another woman is brushing her own hair, with her arm positioned naturally for self-grooming. Set during golden hour with warm backlighting highlighting individual strands of hair being pulled by the brushes. Natural poses with realistic arm positioning for hair |
Figure 7. Whisk-generated hair spray.
|
brushing motions. Moonlit summer evening in the background." Key improvements: (1) Specifies "mid-stroke" and "bristles in direct contact with hair", (2) Mentions "visible tension" as the brush pulls through hair, (3) Describes the specific arm positioning needed for realistic brushing, (4) Emphasizes highlighting of individual strands being pulled by brushes, (5) Maintains your desired setting and atmosphere.
 |
Figure 8. Femtobot experiment. Click to enlarge. The experimental application of hair nanites.
|
Claude's suggested prompt (above) worked fairly well with ImageFX and Whisk, but Leonardo still struggled to show actual hair brushing.
My true interest is in hair nanites, not hair brushing, so I requested, "A woman holding a spray bottle on one hand and a hair brush in her other hand". Whisk then generated images such as the one shown above in Figure 7.
 |
Figure 9. Hair modifying femtobots.
|
I modified
Figure 7 by having
WOMBO Dream make some futuristic hand-held devices that I pasted into the original image to make
Figure 8.
For the image that is shown to the right (Figure 9) I began with a Leonardo-generated image in which Tyhry was holding a handle and added a futuristic hand-held device. Then I tried to make it look like a swarm of femtobots were migrating out of the device and into Tyna's hair.
For my science fiction stories such as "D*", I depict human variants including the Asterothropes and the tryp'At as having pheromone-like bodily secretions (such as oxypathin) that can enhance telepathic abilities. I modified one of the Leonardo-generated images in which Tyna had a protruding tongue as shown below in Figure 10.
 |
Figure 10. Text and tube added manually.
|
When I was having Mr. Wombo make the stream of femtobots for
Figure 9, the image that is shown below in
Figure 11 was generated.
The remaining images in this blog post (below) are "womboizations" of "hair nanite" images that were originally generated by Leonardo or Whisk.
I used this prompt for having WOMBO Dream make HiTek™ hand-held devices: "photorealistic rendering a futuristic fractal complex electronic device with tiny embedded LED lights and circuits, the device is held in the small hands of a woman. A cloud of sparkling multi-colored mist droplets emerges from the device and flows into a woman’s beautiful blond hair".
 |
Figure 12. The original Leonardo-generated hair brush was transformed into femtobots and a nanite probe.
|
 |
Figure 13. The original Leonardo-generated brushes were transformed into nanite probes. |
 |
Figure 14. Two of the nanite probes from Figure 13 and a WOMBO Dream hair nanites bonus image (right).
|
 |
Figure 15. This Whisk-generated image was "womboized" to insert the hand-held hair nanite application devices.
|
 |
Figure 16. This Whisk-generated image was "womboized" to insert a new face for Elyrix the Grendel, the nanite applicator, a better face for Tyhry and nanites being added to Tyhry's hair by the cat-like Felydyan.
|
Next: Plans and illustrations for Part 8 of "D*".
No comments:
Post a Comment