Tag: Dall-E

Whisper of Wings

Posted in Art, Artificial Intelligence, ChatGPT, Cover Art, Dall-E, flash fiction, Generative AI, Short Story, WritingTagged , , , , , , , , , , Leave a comment

There’s a park outside London where the trees keep secrets and the air hums with untold stories. Nigel, a chap with calloused hands and a life measured in paycheques, stumbles upon a moment that’ll unravel him. A purse, unguarded on a bench, whispers temptation. He’s no thief, just a man cornered by circumstance.

The park, draped in the solemnity of dusk, watches as Nigel succumbs. He lifts the cash, a weight heavier than coins, and returns the purse to its owner, an elderly lady scattering crumbs for birds, her gaze lost in yesterdays.

It begins as a murmur on the wind. “I know what you’ve done,” whispers a disembodied voice. Nigel whirls around, searches the empty park in vain. He shakes off the words as a trick of his fraying mind.

But the voice persists, insidious as poison, relentless as the tide. Nigel wanders the park’s paths, and the leaves hiss with recrimination while shadows seem to lean in, heavy with judgment.

Reality blurs, the line between guilt and madness thinning. Nigel confides in a mate over a pint, his voice taut with fear and disbelief. “I’m hearing things in the park, a voice saying ‘I seen what you done.’ But I can’t find where it comes from.” His words trip over themselves.

The whispers follow Nigel everywhere, rustles of feathers echoing each accusation. His desperation cresting, Nigel finally flees the park. But even as he runs, the voice pursues, wings beating in the darkness over his head.

In his panicked flight, Nigel fails to see the lorry barreling down the street. It connects with a sickening crunch, leaving his broken body splayed on the pavement.

“I know what you’ve done,” it declares, Nigel’s crime given feathered form. A final cosmic jest, as this guardian of the park delivers justice for his misdeed.

Quoth the parrot, “Nevermore.”

Sometimes you just get in the mood to write a short piece of nonsense. In this case, I liked the theme of a paranoid person haunted by a talking parrot. From there, I wanted to capture elements of Edgar Allen Poe’s Telltale heart and (obviously) The Raven with a bit of Dostoyevsky’s Crime and Punishment.

As usual, creating cover art is an adventure. I asked Dall-E to render an image of the elderly woman on a park bench with a wooded background and a parrot perched in a tree behind.

It decided on this. It was hilarious to me, so I kept it. NB: I did not ask for it to be rendered on a faux book cover. smh

Dall-E’s first take before I asked for the revision described above.

Exowombs

Posted in Art, Artificial Intelligence, Book, Cover Art, Dall-E, Generative AI, Hemo Sapiens, Hemo Sapiens: Origins, Origins, Sci-Fi, Status Update, WritingTagged , , , , , , , , , Leave a comment

A challenge with beginning a story in media res and then writing a prequel, is that one is able to kick the creative can down the kerb and cross the bridge when you come to it. I’ve painted myself into a few corners, but exowombs, or artificial wombs, are one of them.

Being speculative fiction, I have some leeway, but I need to make some plausible connexions. Exowombs have existed for a few years now, but they are for premature infants and animals, so my literary licence needs to stretch that. To be honest, when I was contemplating things at a fifty-thousand-foot level, I was going from test tube to petri dish to incubator, but I overlooked the gestation bit. Oopsie. My bad.

This is not a work of hard science fiction, so I can take liberties there as well. I just hadn’t researched the current state of science until now. I’ve got a plot device in place, and it seems I’ve got some ideas for early concept and cover art that I can share here.

I rendered these with Dall-E 3. By default, it chose a green hue. I modified it to blue, and I wanted to see how it looked in violet to match their irises—this being an artistic rather than scientific choice. Bubbles in cylinders suspended with wires and tubes.

Violet Gestation Cylinders

Rendering these early can also help me to write descriptive prose with visual references. Dall-E seems to have a thing for spheroids, so I asked it for cylinders instead. I do like this one.

Blue Exowomb

My first correction got me to here. I like the metaphor of the egg membrane encasing the foetus in the tube.

Cylindrical Exowomb

Next, I wanted to envisage multiple cylinders with perspective, so I got these two.

Exowombs in Perspective

The problem I have is that it seems to be too large of a scale, but it’s still cool. We seem to have lost the egg-shaped membrane by now.

Industrial Production of Foetuses in Exowombs

Before settling on violet, I wanted to see what six across looked like. Dall-E’s maths skills are pretty dodgy, so this is what six looks like to it. You’ll notice that the violet render at the top does contain six.

Seven Exowombs in a Row

I don’t have much to say beyond sharing these images. I don’t want to give too much away, but I am excited to be writing Chapter 5 where these are relevant to the narrative in play.

What do you think of the images? Let me know in the comments.

Franglais

Posted in Art, Artificial Intelligence, Dall-E, Generative AI, Hemo Sapiens: Origins, WritingTagged , , , , , , , , , Leave a comment

In Hemo Sapiens: Origin, I am mixing French and English dialogues and tags. The challenge I am having is switch between the languages.

For example, see this passage:

« Où est maman ? » Camille asks Claire just as her parents come into view. « Maman » she exclaims, starting to weep again. « Papa. » She receives his hug.

French and English dialogue and speech markers work differently. I I were depicting large swathes of each language, I’d simply apply the specific language rules, but I am mixing it up, and that creates challenges. I haven’t seen any good examples how to present this.

Some obvious differences are the guillemets « » in French versus ‘ ‘ in English. In French, ? and ! are spaced after the sentence, and all content internal to guillemets is offset by leading and terminal spaces. Another big difference is that guillemets offset dialogue blocks whereas English uses speech marks to identify each speaker’s dialogue.

Referencing the example above—in English for English readers—, if I were to convey the content using French presentation rules, it might look something like this:

« Where’s mum ? Camille asks Claire just as her parents come into view.

— Mum, she exclaims, starting to weep again.

— Dad.  She receives his hug. »

Notice that the entire block is enquoted. I’ve considered this, but I feel it will not track well for English readers, who are used to the speaker-reference convention.

Also, I really want to set off the French language content, and the guillemets serve that function.

Regarding dialogue, French and English punctuation rules are similar enough, but there aren’t many cases of a comma (virgule) following a speech mark given the convention. To my eyes, it looks better inside the marks, but it feels off. The Oxford English style guide suggests not even using commas to separate the dialogue from the tag, but I don’t see that much in the wild.

Again, referencing the example above, one can see how I am solving this at the moment.

At first, I indicate the French dialogue by guillemets and employ French punctuation rules followed by a dialogue tag and descriptive content.

« Où est maman ? » Camille asks Claire just as her parents come into view.

Next, I use the English format, but I replace quotation marks with guillemets. I’ve omitted the trailing comma—after ‘Maman’—in this example.

« Maman » she exclaims, starting to weep again.

Finally, since ‘Papa’ expresses a complete thought, I enclose the full stop within the guillemets. Rather than a dialogue tag, I opt for a stand-alone sentence.

« Papa. » She receives his hug.

When I write mixed language copy, I usually identify a foreign language in italics, but I didn’t choose to do this for French dialogue. Firstly, because I am already using italics for other foreign words, e.g. Latin; secondly, because these also depict internal dialogue/monologue, so I don’t want to create too many visual design patterns.

Has anyone else solved this problem? I’d love to know.

As for the cover image, Dall-E 3 still can’t quite figure out words and can’t spell in French or English. I share it if only for the absurdity of it. Here was my other choice:

Cover Art: Awakening

Posted in Art, Artificial Intelligence, Awakening, Book, Cover Art, Dall-E, Generative AI, hemo sapiens, Hemo Sapiens, Hemo Sapiens: Awakening, WritingTagged , , , , , , , , 2 Comments

Now that the first draft of Hemo Sapiens: Awakening is done, I can let it marinate for a few days before I return to polish it up for Beta readers. Meantime, I can cleanse my palate and concentrate on other matter like cover art. Here’s the third or fourth draft. I know what direction I’m looking for. I was thinking of commissioning a cover, but I might just go it on my own.

I think I’m going with an 9″ x 6″ form factor, so this would be the aspect ratio. I used OpenAI’s Dall-E 3 to render the eyes as well as the overlay with Greek characters. I may have made it too transparent, so I might need to pull it up a few clicks.

Interesting enough, I am compositing in PowerPoint of all things and exporting to PNG. I usually use Paint Shop Pro, but this may be good enough. I still need to compose the back cover and the spine.

My paginated draft comes in at about 320 pages. It was 256 unpaginated pages. There are 38 chapters, and each chapter begins on an odd page number. The facing page is always blank. This creates a lot of blank pages. I could render some line art on the facing pages, but I’m not taking that option at this point. Probably won’t.

My goal for the cover (as might everyone) is to catch a potential reader’s eye and signal something about the content. In this case, the Hemo Sapiens have violet eyes, so I thought that would make a unique image hook. The are open (obviously) to capture the Awakening aspect. The Greek letters on the overlay represent the science aspect. There are 5 versions of Hemo Sapiens so far, named Alpha, Beta, Gamma, Delta, and Epsilon by their creator. I’ve highlighted those characters in a blood red.

Speaking of blood, this is obviously captured in the colour choice and the dripping font of the subtitle. This is also reflected in the stylised O, but I don’t wish to explain the significance of the stylisation now.

As is the entire cover still in progress, so is the treatment of the title as I am playing with a few options. The novel is set in the near future, 2045, so I was thinking of conveying a future with a silver font face, but the gold felt somehow warmer. I wanted to hint at the dystopian angle by distressing the letters.

So, there’s my confession. If you have any comments on the design, I’d love to read them—for better and for worse.

Neuromancer

Posted in Artificial Intelligence, Awakening, Book, ChatGPT, Dall-E, Generative AI, hemo sapiens, Hemo Sapiens, Hemo Sapiens: Awakening, Sci-Fi, WritingTagged , , , , , , , , Leave a comment

William Gibson coined the term ‘cyberspace’ in his book Neuromancer. Just the opening sentence informs the reader to expect something different.

“The sky above the port was the color of television, tuned to a dead channel.”

William Gibson, Neuromancer

In this book, Gibson is poetic and florid. I decided to task ChatGPT to rephrase a paragraph I wrote in Hemo Sapiens: Awakening to follow the style of Gibson.

I shared this paragraph:

The Deputy Chief Inspector starts her day at her polished wooden desk, perusing emails whilst sipping hot morning tea. The sharp aroma fills her nostrils. The fluorescent office lighting hides nothing. Except for some personal photos, her office looks like an austerity challenge. The absence of personal effects betrays her personality.

ChatGPT rendered this:

In the sterile dance of fluorescent light, the Deputy Chief Inspector’s day unfolds. This unforgiving light, a stark mimic of daylight, strips the room of shadows, laying bare the austerity of her workspace. It’s a realm where the polished wooden desk becomes a stage for the mundane ritual of morning emails, and where the walls, devoid of adornment except for a few personal photos, echo the unspoken ethos of restraint.

Amidst this, her morning tea stands as a solitary rebellion. Its steam, a curling serpent of warmth in the chill of artificial light, carries with it an aroma sharp enough to slice through the monochrome of her surroundings. Each sip is a fleeting escape, a momentary immersion in sensory richness that contrasts starkly with the barren office landscape. The tea’s heat whispers of a world beyond these walls, a reminder of life’s subtle flavors amidst the unyielding glare of fluorescent reality. In this space, where the lighting reveals all but understands nothing, the tea becomes not just a beverage, but a quiet testament to the hidden depths beneath her pragmatic exterior.

You’ll notice straight away that my writing is more concise. When I write in 51 words, he writes (by AI proxy) in 181. Mine is 28% of his.

Don’t get me wrong, I love the poetry of the longer form. I just don’t think like that unless I force myself, and then it comes off as stilted or pretentious. My preference is a more direct, terse writing style over the verbose. Perhaps, I’m of a different era. In the end, I compromise.

When I ask AI to critique my prose, a common response to to provide more sensory inputs. In this case, these are represented by her desk, polished and wooden, no less. And she’s sipping hot morning tea—not just a drink, but tea; not just tea, but hot tea, morning tea; and she’s sipping. And this tea has aroma. It’s sharp. So, I’ve got a visual (tea), a tactile (hot), and an olfactory (aroma). I’ve got no auditory or gustatory cues, but I might have.

My preference is to describe as little as possible to allow readers to form their own impressions. This is even more important for characters. On balance, I don’t personally care about the sex, gender, or race of my characters, but I share clues anyway. For race, I might be somewhat obvious as in Hemo Sapiens: Awakening by naming a character Ravi Chandrasekar. It’s fairly obvious that this is an male with heritage from the Indian subcontinent, though it says nothing of his preferred gender.

I might mention height or physique, a moustache or spectacles, a lisp or a stammer, but in the end if a reader wants to image that person as black or white or brown or whatever, I’m not affecting that much. Naming a character Maria instead of Marie sends subtle hints, but it’s not overpowering.

It the case of the Hemo Sapiens, they have blue eyes. Probability would lead one to believe they are Caucasian whites, but this doesn’t preclude something else. That’s not for me to decide. In my head, they are European whites, but I am not going to beat the reader over the head with this trivia.

I’m rambling now. Now read the faux-Gibson passage. It is much more immersive and experiential. If that’s your aim, then go for it. No harm, no foul.

You may prefer one over the other. I say they are simply different, and it depends on the intent of both the reader and the writer.

BONUS

Wanting to generate a featured image for this post, I used Dall-E. I’ve never had Dall-E talk back to me. It usually just renders what I’ve asked. This time, note its response. 🤣

Do you have a writing style preference? Does your writing style veer from your reading style? Let me know in the comments.