日本語ページ   ▶AI art top   ▶AI art blog   ▶Image top  

AI art blog 🤖🖼

A blog related to original AI artworks. I make AI artworks using coined words and emoji as prompts.

Results of the “Free Creations by AI Art” Exhibition at Design Festa

2024-11-18 07:51

On November 16–17, we exhibited and sold posters of AI art under the title “Free Creationsby AI Art” at Design Festa, held at Tokyo Big Sight. We also showcased and explained the five works that won awards or were selected at exhibitions held at the Tokyo Metropolitan Art Museu ...

Dasyn Exhibits and Sells Posters at Design Festa, November 16–17

2024-11-17 23:39

At Design Festa vol.60, held at Tokyo Big Sight, we selected a few dozen works from over 130,000 images to exhibit and sell on the spot. Other works, excluding the latest ones, can be viewed on our official site, and posters are available for purchase on Rakuten. The exhibition, titled "Fr ...

Visited the Chubi and Shin Geijutsu Exhibitions

2024-11-17 21:45

As previously mentioned, one of my works received an award and two works were selected at the Shin Geijutsu Exhibition, and two works were selected at the Chubi Exhibition. I attended the awards ceremonies and parties for these exhibitions and later visited the exhibitions themselves.The f ...

Poster Presentation at Nicograph 2024

2024-11-17 20:00

I gave a presentation titled “Using Personal Names or Pen Names as Prompts to Unleashthe ‘Artistic Talent’of Generative AI” at Nicograph 2024, a conference organized by the Society for Art and Science. The presentation covered the Kana and Kanabstracd artworks featu ...

Three Works Awarded and Selected at the Shin Geijutsu Exhibition

2024-11-06 07:40

The Shin Geijutsu Exhibition is an annual open-call exhibition held at the Tokyo Metropolitan Art Museum by the Shin Geijutsu Association. Out of the three works I submitted, one received an award, and two were selected. As previously mentioned, I also submitted two works to the Chubi Exhi ...

Chatting with ChatGPT-4o about a “collaboration” with Seurat

2024-11-06 00:00

Chatting with ChatGPT-4o about art interpretations has led to various insights.Q: Please provide an interpretation of this artwork (ci9s7Qlc), including any artists it might have been influenced by.A: This artwork uses techniques like pointillism and abstract expression, creating a vibrant ...

Asked ChatGPT to Interpret a Colorful Lake Landscape

2024-11-02 22:20

I asked ChatGPT to interpret the use of color in the following lake landscape.Q: This painting seems to depict a lake landscape, but the water or ground is very colorful. Why do you think it’s painted this way?A: There could be several reasons for the colorful depiction of the water or gro ...

Asked ChatGPT for Multiple Interpretations of a Multi-Interpretational Artwork

2024-11-02 22:20

I asked ChatGPT for multiple interpretations of the following artwork.Q: Please interpret this artwork in four distinct ways.This artwork, painted in a vibrant abstract style, allows for interpretation from various perspectives. Here are four different interpretations:Rural Landscape: This ...

ChatGPT Correctly Guessed “Hope” as the Theme of Some Images

2024-11-02 18:40

I provided several images generated with the prompt “Kanabstracd Hope” to ChatGPT, asking it to guess the theme. For some of these images, it correctly identified “hope” as the theme. The following three images are examples:I initially asked ChatGPT the following questions:“What influenced ...

How to Select Better Works Using ChatGPT

2024-11-01 17:00

I provided the following artworks to ChatGPT-4o and evaluated them using several types of questions. In the end, the method of presenting multiple works and having it select the best appears to be effective for choosing higher-quality works. Below, I describe the three methods I tested.Hav ...

Agreeing with Many of ChatGPT-4o’s Explanations of Top 500 AI Artworks

2024-10-31 11:35

Since ChatGPT-4o provided good explanations for a few pieces, I expanded it to cover aroundthe top 500 works. Here are some examples where I largely agree with the analysis.Note that ChatGPT’s answers may vary, so the descriptions here might not match thoseon each artwork’s page.This piece ...

Had ChatGPT-4o Analyze Several AI Artworks

2024-10-30 01:41

I had ChatGPT-4o analyze several works, including those submitted to exhibitions like the Chubi Exhibition, instructing it to especially note influential artists. For example, it identified influences from Seurat and Cézanne in the two Chubi entries, which were generated using those artist ...

Two Works Selected for the Chubi Exhibition

2024-10-30 00:00

The Chubi Exhibition is an annual open-call art exhibition held at the Tokyo MetropolitanArt Museum by the Chuo Art Association, and two of my submitted works were selected. Before applying, I carefully read the submission guidelines to confirm that entries were not limited to works create ...

Over 120,000 AI Art Paintings Created by a New Method, with the Top 5000 Classified by ChatGPT Now Available Online and Sold as Posters on Rakuten

2024-10-18 00:00

We released a press release about the recent year’s achievements,where artworks were created using the real names or pen names of the creators as prompts: https://prtimes.jp/main/html/rd/p/000000003.000113201.html. ...

Impressions of ChatGPTʼs Classification

2024-10-16 18:20

I would like to share some impressions I had after letting ChatGPT classify the images.Artwork of a person’s full body or half-bodyIn one image, a person aboard a ship is drawn very small, yet it was classified under “Human.”Before seeing this classification, I hadn’t noticed t ...

Tried Classifying AI Art with ChatGPT

2024-10-14 14:00

At the beginning of this project, I classified artworks based on the subject matter depictedin the images. However, as the number of artworks and the amount of drawings increased,I could no longer continue this approach. Yet, classifying images by their content or type isimportant, especia ...

AI Also Borrowed Dalí’s "The Persistence of Memory"

2024-09-29 08:19

Up until now, there had been no clear borrowing of Dalí’s paintings in Stable Diffusion. However, recently, for the first time, two clear instances of borrowing from “The Persistence of Memory” have been observed.The following image is one of them.I didn’t make note of the othe ...

Deciding to Submit to Various Competitions

2024-09-29 07:50

Various competitions are being held for painting and CG (computer graphics).For CG, there are competitions like the ADAAand AI Art Grand Prix,both of which have deadlines around September.Additionally, for painting, competitions like the Chubi Exhibition,Shin Geijutsu Exhibition, and Novas ...

Presented a Poster on AI Art titled "AIʼs Artistic Potential Unleashed!" at NicoInt 2024

2024-06-16 10:30

A conference called Nicograph International 2024, organized by the Art Science Societyat Tokyo Metropolitan University in Minamiosawa, was held. At this conference,I exhibited and explained a poster titled “Emergent Text-to-Image Generation Using ShortNeologism Prompts and Negative P ...

Peekholes Drawn by AI

2024-04-07 10:00

Collected here are artworks created by Stable Diffusion XL that resemblepeekholes.Some of these may be mirrors. ...

Monet’s Bridges and Water Lilies That Appear Without Being Named

2024-04-07 09:40

When prompted with the name “Monet,” more than half the time, paintings resembling“water lilies” are drawn. While not always depicting water lilies, Monet often paints bridges.Examples will be provided at the end of this article.The following painting, although not ...

The Munch “Scream” That Appears Without Being Named

2024-04-07 09:17

The Munch “Scream” is so famous that even without specifying Munch’s name, AI oftendraws images similar to it. It may be coincidental, but the following three imagesare all generated from prompts related to fabric patterns.The third image is also influenced by Van Gogh, with st ...

The Waves of Hokusai That Appear Without Being Named

2024-04-07 09:00

When given the name of Hokusai, AI persistently produces the famous wave painting, but it isoften depicted even without the name of Hokusai. Several examples are provided below.The painting in the top right corner seems to be influenced by Van Gogh as well.The next two paintings differ con ...

Scenes with the Moon and Sun Drawn by Kana and Kanabstracd

2024-03-28 20:00

When instructed to draw by Kana (Yasusi_Kanada) or Kanabstracd(that is, Kana or Kanabstracd is used as the positive prompt),landscapes often featuring the moon or sun are frequently depicted.First, examples drawn by Kana are presented.In the next example, the surface of the moon is depicte ...

Commonly Drawn Tunnel Images by AI

2024-03-28 09:10

Among the images drawn by Stable Diffusion, there are often depictions of tunnels or corridors.Especially when given words related to physics, square-section tunnels are frequently observed.The following image depicts what could be called a corridor.The next image resembles more of a spaci ...

Diverse Faces and Bodies Drawn Without the Influence of Animal Patterns

2024-03-28 09:00

When using Kanabstracd as a positive prompt and a word evoking a pattern of animals as a negativeprompt in Stable Diffusion XL, for some reason, it frequently draws faces and bodies of people.Here are some examples drawn with the prompt “rat.” In this case, not only faces but a ...

The Patterns Unveiled by AI with “Escher"

2024-03-28 08:00

Having found that generating images with patternized prompts is effective, I searched for wordsthat evoke patterns. One such effective word is “escher.”While words like “pattern” or “texture” are used to repetitively draw and generate patterns,they may n ...

Balance in “Kanabstracd Art Without Goldfish Influence” Amidst Linear Dominance

2024-03-27 23:20

When Kanabstracd is set as the positive prompt and goldfish as the negative prompt, it seemsthat a balance is achieved between predominantly linear elements and moderate curves, resultingin a well-balanced image.The following images, under this prompt, showcase the most linear abstract com ...

More References to Van Gogh Here...

2024-03-27 22:30

The following image is created with Kanabstracd as a positive prompt and “dices” as a negative promptin Stable Diffusion XL.With swirls in the sky, the connection to Van Gogh’s “Starry Night” is evident. It can also be foundin Google Image searches. Furthermore, the ...

Surreal Images Generated with Bow and Arrow as Negative Prompts

2024-03-27 18:20

Using Kanabstracd as a positive prompt and “bow_arrow” or its plural “bows_arrows”as negative prompts in Stable Diffusion XL resultsin various surreal images.Where is this place, and what are the objects rising up? Even when these images are searchedon Google Images ...

Images Generated with Barbed Patterns as Negative Prompts

2024-03-27 12:00

I experimented with the word “barbed” as a negative prompt. When combined with wordslike “pattern,” “texture,” and “Escher,” variousinteresting images were generated.Since “Barbed” is part of the term “barbed wire,” in ...

Effectiveness of Using Patterns (Textures) as Negative Prompts

2024-03-27 12:00

Using patterns, or repeated motifs, as negative prompts seems to result in more interesting images with a higher probability. By introducing repetition, it clarifies the “prohibitions” and enhances the style of the image. Initially, noticeable patterns such as waves and fabrics ...

Van Gogh-Inspired Paintings from Collaboration between Kanabstracd and Klimt

2024-03-26 20:00

In the paintings created by Kanabstracd and Klimtusing Stable Diffusion XL, there are several paintings that seem to be influenced by Van Gogh.Although they differ significantly in style from Van Gogh, these paintings show influence from him.When searched on Google Images, both of these pa ...

Faces Embedded in Images

2024-03-26 18:00

AI is adept at embedding human faces in various objects and abstract paintings. For example,in the following Picasso-style painting, it seems like the faces of multiple people are embedded.In the next image, faces seem to be embedded within the patterns.In the following images, cat faces a ...

Drawing Emoji Cats with Kanabstracd

2024-03-26 11:00

Stable Diffusion XL understands emoji cats just like Stable Diffusion 1.5, soit draws true cats.It also draws abstracted cats.It draws strange cats and abstract art. ...

When Kanabstracd Was Prompted with Emoji Cats, Lions and Mice Were Drawn

2024-03-26 11:00

When prompted with emoji cats along with the pseudonym “Kanabstracd” in Stable Diffusion XL,naturally, it did not draw typical cats, but instead drew lions, mice,and various other unclear images.As Kanabstracd is primarily an abstract artist, it also draws abstract art.Perhaps ...

Loss of Diversity with Negative Prompts

2024-03-22 20:00

Negative prompts are used to diversify AI art images, but it has been found that in some cases,diversity may decrease or almost disappear when negative prompts are provided.The following images were generated with “Kanabstracd” and the painter Hieronymus Bosch as positiveprompt ...

Issues with Color Depth in 16-bit Images

2024-03-16 15:30

!Most image-generating AI, including Stable Diffusion, currently representimages in 16 bits per pixel, which is often insufficient.When generating images with AI, it often produces images with very low contrast.Increasing the contrast in such images essentially represents the image with mu ...

Drawing Polygons as Negative Prompts

2024-03-14 22:33

Using the name of polygons as negative prompts, I had Stable Diffusion XL generate images.While the most diverse results are obtained from hexagons, various images are generatedfrom triangles to pentagons.HexagonsAbstract paintingsPentagonsSquaresTriangles ...

Why Do Images with String Instruments like Guitars and Cellos Occasionally Appear in the Center?

2024-03-14 22:30

Images with string instruments like guitars and cellos occasionally appear in the center. Why is that?I will provide several examples for now and add more later. ...

The Transformed Clock Shop (?)

2024-03-14 15:20

There is an “image” like the following. What could this be?It is just speculation, but I think this is an image generated by AI from a photograph of a clockshop with many clocks lined up on the wall. Each clock seems to have been replaced with a swirlingpattern. There may even ...

Various Patterns of Fabrics as Negative Prompts

2024-03-13 21:20

Fabrics come in various patterns. When these are specified as negative prompts, the imagesgenerated by Stable Diffusion vary significantly for each pattern.ArgyleWhen “Kanabstracd” is used as the positive prompt and “Argyle” as the negative prompt, AI(Stable Diffusi ...

Why Does AI Often Draw People Wearing Headphones?

2024-03-13 21:10

Stable Diffusion (XL) often depicts human faces, and among them, many people are wearing headphones.Why is that?Here are some images generated by Stable Diffusion XL with “Kanabstracd” as the positive prompt.However, the frequent appearance of people wearing headphones is not u ...

Images Generated from Starfish Patterns: Perspective and Vague Depictions of People

2024-03-13 20:00

Using “Kanabstracd” as the positive prompt and an image containing numerous starfishas the negative prompt (when used as the positive prompt), I had Stable Diffusion XL generate images.I believe I have created some interesting pictures.Compared to those created with other promp ...

When Using Railway Patterns as Negative Prompts, Surreal Landscape Paintings, Decorative Portraits, and Abstract Art are Produced

2024-03-12 22:00

!Using “Kanabstracd” as the positive prompt and experimenting with various patterns as negativeprompts, it was found that specifying patterns related to railways, such as “railwaypattern”,as the negative prompt produce intriguing images.Several phrases were tested a ...

Color Fading and Its Mitigation by Negative Prompts

2024-03-12 12:50

In Stable Diffusion XL, certain negative prompts can cause a degradation in the colors of the generated images. Yellow and green hues are significantly reduced, leaving predominantly red and blue tones. For instance, when using the creatorʼs name (Yasusi Kanada) as the positive prompt ...

Diverse Images Prompted by “Mondrian” as Negative Prompt: Klee, Kandinsky, Kanabstracd

2024-03-12 12:00

When using “Klee” or “Kanabstracd” as positive promptsand “Mondrian” as the negative prompt, the patterns in the images change significantly. This phenomenon was initially noticed when experimenting with various negative prompts alongside the positive pr ...

Diverse Artworks Prompted by the Pen Name “Kanabstracd”

2024-03-11 18:00

Choosing my pen name as “Kanabstracd” and setting it as a positive prompt for generating images with Stable Diffusion XL worked well, so I decided to add some variation. I considered using a pen name because I felt that it would be difficult to assert ownership with a name that ...

Battle with Bias in AI Art Images Prompted by the Name “Kanada”

2024-03-09 16:00

I attempted to draw about 8000 images using my name as a prompt with Stable Diffusion XL. As a result, it was often found that maple trees or maple leaves resembling the Canadian flag were drawn from the spelling “Kanada.” The bias towards maple trees tends to be associated wit ...

Trying CMYK Image Profile for Prioritizing Poster Printing

2024-03-07 19:20

When printing from RGB images to posters, the colors can change significantly. While the images on this site are also intended for viewing on display screens, considering printing them as posters first, I thought it might be better to create the images using the CMYK color composition meth ...

Paintings Drawn Using the Pen Name “Kanadansky” as Positive Prompt and Russia or Russians as Negative Prompt

2024-03-07 12:00

I experimented with various negative prompts while using “Kanadansky” as the positive prompt, intending it to be the creator’s pen name. Since specifying only “Kanadansky” often resulted in Russian-style landscapes, I thought that specifying something related to Rus ...

Artworks Illuminated by Light etc. Drawn Using the Pen Name “Kanadansky” as Positive Prompt and Ukiyo-e Artists as Negative Prompt

2024-03-07 12:00

I experimented with various negative prompts while using “Kanadansky” as the positive prompt, intending it to be the creatorʼs pen name. When I found that specifying the name of a ukiyo-e artist as the negative prompt while using the creatorʼs name as the positive pro ...

Russian/Canadian-style Landscape Paintings Drawn Using the Pen Name “Kanadansky” as Positive and Negative Prompts

2024-03-07 11:05

Since drawing images by specifying only my name (Yasusi_Kanada) as the positive prompt workedwell, I wanted to add some more variation. Therefore, I considered using a pen name. I thought that using a pen name that is different from my real name would make it less claimable as my own, so I ...

Drawing Different Images from Ukiyo-e Artists by Specifying My Name as Positive Prompt and Manga Artists’ Names as Negative Prompts

2024-03-07 11:00

I considered using my name (Yasusi Kanada) as a prompt. I have written the reasons separately. To achieve various images, I considered specifying only my name (or with “painting”)as the positive prompt and diversifying the images with negative prompts. For the negative prompts, ...

Drawing Abstracted Figures by Specifying My Name as Positive Prompt and Painter Names as Negative Prompts

2024-03-04 21:30

I considered using my name (Yasusi Kanada) as a prompt. I have written the reasons separately. To achieve various images, I considered specifying only my name as the positive prompt and diversifying the images with negative prompts. Therefore, negative prompts such as Millet, Rembrandt, an ...

Drawing Deformed Figures, Landscapes, and Abstract Paintings by Specifying My Name as Positive Prompt and Ukiyo-e Artist’s Name as Negative Prompt

2024-03-04 21:20

I considered using my name (Yasusi Kanada) as a prompt. I have written the reasons in another article. To obtain a variety of images, I specify only my name (or with “painting”) as the positive prompt and diversify the images with negative prompts. For the negative prompt, espe ...

Exploring Image Variation by Specifying Only My Name as Positive Prompt and Diversifying with Negative Prompts

2024-03-04 21:00

I considered using my name (Yasusi Kanada) as a prompt. Here is the reason: Creating images with short prompts is something anyone can do. If interesting images can be created this way, and if the prompt is publicly available and does not reflect my “rights,” anyone can copy it ...

Exhibition Results of “Art AI’s Free Creation Works” at Design Festa

2023-11-13 11:00

I exhibited AI art pieces under the title “Art AI’s Free Creation Works” at Design Festaheld at Tokyo Big Sight from November 11th to 12th. Although the AI art imageson the MWAIP website showcase around 46,000 images,due to limited exhibition space, I displayed only 36 images. While most p ...

Dasyn to Exhibit and Sell Posters at Design Festa on November 11-12

2023-10-13 11:30

At Design Festa vol.58 held at Tokyo Big Sight, I will select and exhibit 36 out ofover 40,000 images. Additionally, you will have the opportunity to purchase other postersalong with them.The exhibition is titled “FreeArtistic Creations by AI,” and our booth is B-137. Except fo ...

Prompts Including My Name Worked Well in SDXL

2023-10-13 11:00

!Until now, I have mainly used the names of famous painters as prompts to create painted-style images in Stable Diffusion. However, since I am using short prompts, anyone can easily imitate this method. To create artworks that are resistent to imitate, I decided to try using a prompt that ...

Changing the Initials of Kandinsky’s Name Creates Dramatically Different and Interesting Artworks

2023-09-15 22:00

In Stable Diffusion 2.1, by changing only the initial of Kandinsky’s name, “K”,such as “Dandinsky” or “Vandinsky,” entirely different and unique artworks are created, distinct from Kandinsky’s style.Among these, “Dandinsky” and “Vandins ...

Dasyn Releases Over 30,000 Images Created by AI Art, and Sells Approximately 2,000 Posters

2023-09-12 08:00

I have issued a press release regarding the achievements of the past 9 months.https://prtimes.jp/main/html/rd/p/000000002.000113201.html ...

When Learning Hokusai and “Negating” Ukiyo-e, AI Paints Surreal Ukiyo-e Style Landscapes

2023-09-07 16:40

I provided the name of Hokusai as a positive prompt and “Ukiyo-e” as a negative promptto Stable Diffusion 2.1. This resulted in surreal Ukiyo-e style landscapes and more.When it comes to Hokusai, his famous wave paintings from “Thirty-six Views of Mount Fuji”are glo ...

When Learning Sharaku and “Negating” Ukiyo-e, AI Creates Black “Heroes” and Eerie Figures

2023-09-07 16:30

I’ve written about the pictures generated by teaching AI with Sharaku’s name in a separate article. This time, I tried providing “ukiyo-e” as a negative prompt. The results are red-caped black “heroes” and eerie figures. Why?!Letʼs continue. ...

When Learning Utamaro and “Negating” Ukiyo-e, AI Creates Three-Dimensional Beauties

2023-09-07 16:00

Kitagawa Utamaro created numerous portraits of beauties, many of which are available online. Providing just Utamaroʼs name results in AI-generated artwork that resembles his style of beauty portraits. However, when I attempted to have AI create something other than beauty portraits, I ...

AI Art of Mountains and Ponds Elicited from “Denied” Hiroshige

2023-09-07 14:00

When only Hiroshige’s name is given to Stable Diffusion 2.1, many of the resulting paintingshave muted colors and are not particularly attractive. However, by providing his name or “Ukiyo-e” as a negative prompt, more diverse landscape paintings featuring mountainsand ponds can ...

High-Resolution AI Art with the Latest VanceAI

2023-09-06 15:00

VanceAI offers the ability to upscale image resolution by 2-8 times, and it seems thata new algorithm has recently been introduced on this site. When I tried it, I noticedsignificant improvements compared to before.Most of the images on this site are created on OpenArt.ai. Itʼs conven ...

Cute Art Created by AI Using the Name of Arcimboldo

2023-09-04 14:25

Arcimboldo’s paintings are strange and eerie, but when you provide his name as both positive andnegative prompts, along with names like Rubens and Botticelli as negative prompts, various cuteartworks surprisingly emerge.First, let’s consider the case with Rubens. Various artworks, includin ...

AIʼs Artistic Variety by Learning from Kandinsky and “Negating” Mondrian and Monet

2023-09-03 11:00

By providing Kandinskyʼs name as a positive prompt and names like Mondrian and Monet as negative prompts to Stable Diffusion 2.1, various artworks different from what Kandinsky created can be obtained.When only Kandinskyʼs name is provided, most of the images have a vibe like the ...

AI’s Creation of Peculiar Landscapes by Learning from Kandinsky and “Negating” Cubism

2023-09-03 10:40

By providing Kandinsky’s name as a positive prompt and the word “Cubism” as a negative prompt to Stable Diffusion 2.1, various surrealistic landscape paintings different from what Kandinskycreated can be obtained.When only Kandinsky’s name (Kandinsky) is provided, most of the i ...

AIʼs Creation of Various Images by Learning from Kandinsky, “Negating” Utrillo, and Changing the Scale

2023-09-03 10:00

By providing Kandinskyʼs name as both a positive and negative prompt to Stable Diffusion 2.1 and further adjusting the “scale,” various images with different styles were obtained.By providing Kandinskyʼs name in various forms, a surprisingly diverse range of images ca ...

AI Generates Various Images by Learning from Kandinsky and “Negating” Klee

2023-09-03 10:00

By providing Kandinsky’s name as a positive prompt and Klee’s name as a negative promptto Stable Diffusion 2.1, various images, including landscapes, portraits, and abstract art,were obtained.Among the landscape images, we have the following examples.The following images could also be cons ...

AIʼs Diverse Artworks by Learning from Kandinsky and Negating Various Painters

2023-09-02 21:00

By providing Kandinskyʼs name as a positive prompt and various paintersʼ names as negativeprompts to Stable Diffusion 2.1, you can obtain a wide range of artworks.The two images above, the left with Pollock specified as the negative prompt and the rightwith Mondrian specified, se ...

Diversity Unleashed by AI from Picabiaʼs “Negation”

2023-09-02 19:00

We provided the name of the modern artist Picabia to Stable Diffusion 2.1 and had it create various paintings. By using this as both positive and negative prompts, we obtained a surprisingly diverse range of images.Even when only Picabiaʼs name is given as a positive prompt, interesti ...

AI paints Van-Gogh-style landscapes by “Shin Expressionism” or “Shin Impressionism.”

2023-09-02 18:39

We gave the coined term “shinexpressionism” to Stable Diffusion 2.1 to create art. Among various paintings generated, one of them is shown below.There are trees on the left side of the swirling sky in this painting. It seems to be influencedby Van Goghʼs “Starry Nigh ...

AI Excludes Lattice Patterns from Kleeʼs Paintings...

2023-09-02 16:00

When providing Kleeʼs name (Klee or Paul Klee) to Stable Diffusion 2.1 to generate images, it often creates lattice-like patterns. To encourage the creation of more diverse images, I provided the word “lattice” as a negative prompt, and this worked well, resulting in a var ...

AI Paints Cubist-Style Landscapes When Making AI Learn from Blaque and “Negating” Picasso

2023-09-02 14:00

In Stable Diffusion 2.1, when providing the name of the Cubist painter Braque as a positive promptand Picasso as a negative prompt, AI paints landscapes with a Cubist atmosphere that are quitedifferent from Braqueʼs paintings. Many of them depict mountain scenes.In the next four image ...

Learning from Arcimboldo and “Negating” Vermeer Leads AI to Create Paintings Resembling Complex Surfaces

2023-09-02 12:00

Giuseppe Arcimboldo was a Renaissance painter known for creating bizarre faces composed of vegetables, small animals, and more. When you provide his name to Stable Diffusion 2.1, it generates paintings resembling those created by Arcimboldo himself. However, when you use the name of anothe ...

When AI Learns from Arcimboldo and “Negates” Rubens, it Paints a Variety of Decorative Faces and Planar Surfaces

2023-09-02 12:00

Giuseppe Arcimboldo was a Renaissance painter known for creating strange faces composed of vegetables, small animals, and more. When his name is given to Stable Diffusion 2.1, the AI creates images reminiscent of Arcimboldo’s own work. However, when the name of another artist is provided a ...

AI Creates Diverse Images with a Array of Faces When Learned from Arcimboldo and “Negates” Rembrandt

2023-09-02 12:00

Giuseppe Arcimboldo was a Renaissance painter known for his peculiar portraits made up of faces consistingof vegetables, small animals, and more. When his name is provided to Stable Diffusion 2.1, the AI producesimages similar to what Arcimboldo himself painted. However, when given other p ...

AI Creates Colorful Images of Stones and Objects When Learned from Arcimboldo and “Negates” Botticelli

2023-09-01 13:00

When given the name of Arcimboldo to Stable Diffusion 2.1, it produces paintings similar to whatArcimboldo himself painted. However, providing Arcimboldo as a positive prompt along with the nameof another painter as a negative prompt results in various transformations in the art. In this c ...

AI Creates Various Images of Women When Learned from Arcimboldo and “Negates” Van Gogh

2023-09-01 12:00

When Giuseppe Arcimboldo’s name is provided to Stable Diffusion 2.1, the AI creates images similarto what he painted. However, when Arcimboldo is given as a positive prompt along with other painters’names as negative prompts, the artwork undergoes various transformations. In this case, we ...

When AI Learns from Arcimboldo and “Negates” Various Painters, the Generated Artwork Varies Significantly

2023-08-29 17:30

Giuseppe Arcimboldo was a Renaissance painter known for his peculiar portraits composed ofvegetables, animals, and other elements. Providing his name as a prompt to Stable Diffusion 2.1results in artwork reminiscent of Arcimboldo’s own creations. However, when other painters’ namesare give ...

When AI Learns from Arcimboldo and “Negates” Mucha, AI Creates Portraits Quieter than Arcimboldo

2023-08-29 17:00

Giuseppe Arcimboldo was a Renaissance painter known for his peculiar portraits composedof vegetables, animals, and other elements. Providing his name as a prompt to Stable Diffusion 2.1results in artwork reminiscent of Arcimboldo’s own creations. However, when the name of Muchais given as ...

When Mondrian is “negated,” the AI learns from Klee and creates rich paintings!

2023-08-29 16:00

Stable Diffusion 2.1 does not produce very interesting images when given the name “Klee”(referring to Paul Klee). However, interesting images emerge when using the name “Mondrian”as a negative prompt.These were created simply by using “Klee” as a positiv ...

Learning from AI-Generated Van Gogh: Various Starry Nights

2023-08-29 12:00

!As I have mentioned before, when providing the name “Van Gogh” to Stable Diffusion 2.1, it persistently generates images of starry nights. Sometimes, it even does so without his name. Letʼs explore these various depictions of starry nights.In some of them, the color of sk ...

Experimenting with Learning from Photographers to Create Art

2023-07-04 19:40

So far, I have been using AI to learn from painters (specifying their names) and create artwork. However, this approach may raise copyright issues when learning from existing artists. Therefore, I thought about trying to learn from photographs and creating art. If I can produce paintings t ...

AI Art Created from Natural Science Terminology

2023-07-04 12:10

While we have experimented extensively with mathematical terms, this time we triedusing terms from physics, chemistry, and other fields to generate AI artusing Stable Diffusion 2.1.Combining specific terms with specific artist names, such as “relativity” and Klimt,or “qua ...

Drawing AI Art with Various Mathematical Terms

2023-07-04 10:50

Continuing the experiment of creating AI art with mathematical terms, we will nowintroduce various artworks derived from different terms that havenʼt been covered yet.From a term like “exponent” and a close word “exponentiation,” as wellas “reciprocal, ...

AI Art Created from Probability and Statistics Terminology

2023-07-04 10:46

While continuing our endeavor to create AI art using mathematical terms, in this case, we will showcase artwork derived from probability and statistics terminology.Mathematical terms are typically abstract, but in the case of statistics, they are often associated with graphs, making them m ...

Drawing Mathematical Terms with AI Art: “Linear” and “Nonlinear"

2023-07-04 09:52

Continuing the experiment of creating AI art with mathematical terms, in this case,we focus on “linear” and “nonlinear.” Even with “linear,” specifyingboth positive and negative prompts results in interesting artwork. However,with “nonlinear,” ...

AI Art Created with Famous Artists’ Names and “Infinity” is Outstanding!

2023-07-03 18:00

I have previously written about the interesting results of creating artwork using the word“infinity” or similar words with Stable Diffusion. In addition to that, by providingnames of artists like Kandinsky and Klimt, the most fascinating artworks have been createdso far.I have ...

Trying AI Art with “Differentiation” and Similar Words

2023-07-03 16:10

I have written about the interesting results of creating artwork using “infinity” or similar words as prompts for Stable Diffusion a couple of times already. So, I decided to try other mathematical terms and experimented with several of them. In this article, I will focus on di ...

Exploring the “Negation” of Cats in AI Art

2023-07-03 14:20

When instructing AI to paint cat images, I have been experimenting with both positive andnegative prompts, including “cat.” Sometimes, I include two instances of “cat,” but thattends to result in multiple cats rather than enhancing the cat presence. First, let me sh ...

AI’s rendering of Afremov’s painting style undergoes significant changes when negating “expressionism”

2023-07-02 23:20

I attempted to apply the name of contemporary artist Leonid Afremov (1955-2019) to Stable Diffusion.His paintings are characterized by a distinctive style that is immediately recognizable, and cautionis necessary regarding copyright issues when Stable Diffusion produces similar artwork.How ...

AI Art Unveiling Nude Artworks

2023-07-02 13:30

AI art has created various nude artworks. In AI art websites, the inclusion of the word “nude” is usually prohibited, so some ingenuity is required.Conversely, if there is too much ambiguity, it can generate inappropriate images,so those are not included on this site.First, her ...

The Diverse Art Emerging from “Repetitive Names” like “Lili”

2023-07-02 11:00

I had AI create art using “repetitive names” such as “Lili” or “Mimi”. To achieve a painting style, I added “🖼” as a positive prompt in Stable Diffusion 1.5, and in 2.1, I provided “painting” as both positive and negative prompts. ...

“Ambiguous” Cityscapes Created by AI Art

2023-07-01 18:00

Stable Diffusion 1.5 and OpenJourney sometimes depict “ambiguous” cityscapes.In other words, while the presence of roads and buildings can be discerned, they may appearblurry or indistinct.These are also paintings that painters rarely create, and I believe that this isone of th ...

“Ambiguous” Figures Created by AI Art

2023-07-01 18:00

AI art software, especially Stable Diffusion 1.5 and OpenJourney, often depict “ambiguous figures.”In other words, they create images where certain parts or the entire form of a person are vague,yet it is still apparent that a person has been depicted. Faces are usually relativ ...

Terrifying Faces and Simple Creatures obtained by “Unfin”

2023-06-23 15:50

As I mentioned before, when using “Ynfin” in AI art, terrifying faces, glowing creatures, and plants are depicted. However, during exploration around “infinite,” I discovered that “unfin” and its vicinity also depict demon-like faces and small animals.He ...

Terrifying Faces and Glowing Creatures obtained by “Ynfin”

2023-06-23 15:20

I discovered that interesting images can be obtained with “Infin” or “infinite” by AI art,so I started exploring around that. As a result, I found that “ynfin” and related terms depictdemon-like faces and glowing creature-like entities.First, let’s look ...

“Infinite” Makes AI Art Captivating

2023-06-23 14:20

I wrote about how interesting patterns can be obtained in AI art by using “Infinite” orsimilar spellings in the article“Images with Overlapping Arcs Created with the Prompt ʼInfinʼ”.In that article, I mainly presented images composed of arc-like shapes (re ...

To add color and 3D effects to QR codes in AI art, “negation” was necessary

2023-06-23 12:50

I tried giving “QR code” as a prompt to AI art (Stable Diffusion 2.1).Of course, I didn’t expect the generated images to be readable as QR codes.While they do resemble QR codes, adding “color” doesn’t actually add colors, and adding “3D”does not make the ...

Sharaku, Sharks, and Russia

2023-06-22 19:00

I have created several AI artworks using Sharaku as a prompt in Stable Diffusion. While I expected to seeimages in the style of Sharaku’s ukiyo-e prints, what I didn’t anticipate were the shark photo-like imagesand scenes of Russian cities.The appearance of sharks could be due to the resem ...

Diverse Abstract Paintings and Women Generated by Negating Klimt

2023-06-22 17:50

I have been attempting AI art generation by specifying various artist names as positive and negative prompts. In the case of Klimt, a much greater variety of images are produced compared to when only positive prompts are provided.Let me provide some examples. The first three images contain ...

“Divided” and “Negated” Kandinsky: Intriguing Artworks

2023-06-22 15:55

I have been using various approaches to prompt AI Art with the name of Kandinsky, and this time, I tried combining “dividing” and “negating” the name. In other words, I provided divided versions of "Kandinsky,” such as “K_andinsky” or “Ka_ndi ...

Even “Divided” or “Negated”, Kandinsky Still Inspires

2023-06-22 15:55

I have been using various ways to prompt AI Art with the name of Kandinsky, and regardless of the approach, it produces rich images. As I mentioned in the article Negating Kandinsky: Unveiling Fascinating AI Art”,providing negative prompts along with positive prompts for “Kandi ...

Fun of “Negated” Dali

2023-06-22 13:10

In AI Art, I have mentioned several times that specifying the same artist name in positive andnegative prompts can result in diverse paintings. When I tried it with Dali, I obtainedinteresting results.Although these paintings are completely different from Daliʼs style, what do you thi ...

Diverse Textures Created by AI Art

2023-06-22 13:00

In AI Art, textures can be actively generated or unintentionally appear.Here are examples of textures I have produced so far.One of the most frequently depicted textures is swirl patterns. Some may consider them too large to be called textures, while others are incredibly intricate.Mesh pa ...

Generating Diverse Images with Minimal Words AI Painting (MWAIP) and Methods for Achieving Variation

2023-05-20 00:50

On this site, we showcase images created by AI, particularly Stable Diffusion, with as short and minimalistic prompts as possible, typically consisting of only one word. We refer to this minimalist approach as Minimal Words AI Painting (MWAIP). In this article, we discuss methods for gener ...

AI Art as an Extension of Contemporary Art

2023-05-18 21:00

AI art has become possible through the use of a vast amount of existing images. In thisregard, many contemporary artworks, since Marcel Duchamp’s use of an existing urinalin “Fountain,” share a commonality of using preexisting materials. Therefore, it can besaid that AI art is ...

Plants and People Drawn by AI with Numbers

2023-05-18 15:37

When providing numbers as prompts to Stable Diffusion 2.1 and 1.5, the AI often draws plants suchas trees. I am not sure why this happens, but it allows for various pictures related to plants.Here are some examples:The left image resembles a jungle.The right image seems to have different e ...

When Providing 2 Alphabets to AI for Drawing, “AQ” Shows Particularly Captivating Results

2023-05-18 11:55

Using Stable Diffusion 2.1, I instructed the AI to generate painting-style images with 2 alphabets. Starting with the combination of “A” as the first letter, I found that “AQ” producedparticularly captivating results.The majority of the generated images are abstract ...

An Intriguing Art of Painting with a Single Alphabet in AI

2023-05-18 11:23

With Stable Diffusion 2.1 and 1.5, you can obtain various captivating images in different painting styles by providing a single alphabet.However, to achieve a painterly style, the prompt “🖼” is added for SD 1.5, and “painting” is used as both positive and negative p ...

Diversifying Images by Negating Art Movement Names

2023-05-17 12:10

By negating the name of a painter, that is, specifying both the positive prompt and the negative prompt with the same painter name, you can obtain a wider variety of images in different painting styles than by specifying only the positive prompt. I wrote about this in “Diversifying I ...

Diverse Images Generated by AI in “...ssionism”

2023-05-16 00:40

When providing variations of the spellings for “Impressionism” and “Expressionism”such as “Pressionism”, “Ressionism”, “Ssionism”, or their modified versionsto Stable Diffusion 2.1, it generates various images. By specifying the s ...

Diverse Images Generated by AI Using Words Representing Expressionism and Neoexpressionism

2023-05-16 00:13

In this article, I will introduce some of the various images generated by providing words suchas “Expressionism” and “Neoexpressionism” to Stable Diffusion 2.1.The original intention of Expressionism is to express emotions rather than depict objects objectively. How ...

Diverse Images Generated by AI Using Words Representing Impressionism and Its Successors

2023-05-16 00:09

In this article, I will introduce some of the various images generated by providing words that represent Impressionism and its successors, such as Impressionism, Neo-impressionism, and Post-impressionism, to Stable Diffusion 2.1.It appears that the use of colors in the artwork is more vibr ...

Why does AI only draw cityscapes by “Severini”, but diversifies when Severini is “negated”?

2023-05-14 10:03

By specifying a single word as both a positive and negative prompt in Stable Diffusion 2.1,we can obtain a diverse range of images. I tried this method with Severiniʼs paintings.If we specify only the positive prompt with Severiniʼs name, the generated cityscapes will have a simi ...

“Negating” results in diversification, but Van Goghʼs Starry Night and Self-Portrait still appear.

2023-05-14 08:33

!By specifying a single word as both a positive and negative prompt in Stable Diffusion 2.1, you can obtain a diverse range of images. When trying this with Van Goghʼs paintings, while the resulting images are diversified, Starry Night and Self-Portrait still appear frequently.The fir ...

Diversifying Generated Images by Specifying Painter Names in Positive and Negative Prompts

2023-05-13 15:00

When the same painter name is specified for positive and negative prompts, Stable Diffusion 2.1 often generates very interesting images.And sometimes, very diverse images are generated.Specifying the name of an artist in a negative prompt to obtain diverse images may be related to the fact ...

Registering AI Art Created on Pinterest

2023-05-13 10:19

Although I have been releasing original AI art images onPinterest in small amounts, I recently found a way to register them on Pinterest with relatively little effort, so I am now using it to distribute AI art images.Instagram is often used as a way to release original images such as photo ...

ChatGPT is the best for translating this blog into English!

2023-05-12 23:20

This blog has both a Japanese version and an English version, and the content is almost the same. When referencing image web pages from blog articles, I provide links to both the Japanese and English versions. The same applies to references to blog articles. While Google Translate often fa ...

“Negation” Makes It More Attractive Kandinsky-style AI Art

2023-05-12 21:10

When using Stable Diffusion 2.1, it is possible to obtain a diverse range of images with varied transformations by specifying a single word as both a positive and negative prompt simultaneously. When trying this with Kandinsky, I obtained results that exceeded my expectations.Although Kand ...

AI Image Ranking and Updating on Website

2023-05-12 15:54

On this website, we display images categorized using AI art programs such as Stable Diffusion. However, there may be hundreds of images in a ...

How far can you get night-scene images with a coined word close to “night”?

2023-05-12 00:13

When using Stable Diffusion to draw a night view, we sometimes use a neologism similar to “night”as a prompt, such as “nigh” or “nite”. However, just using these neologisms does notnecessarily result in a night view. Therefore, if the contrast of the res ...

Grammar of prompts for AI art programs

2023-05-11 17:47

We mainly create images in the style of paintings using an AI called Stable Diffusion. In this case, the words that make up the prompt given to the AI can be divided into painter names or nominative case and object names or accusative case. It is also possible to include a verb, but it is ...

AI Art Image Classification and Ranking

2023-05-11 17:30

I am creating various types of images using AI art programs. In order to share them withthe public, I need to categorize them and display them in order of preference by category.With over 13,000 images, it is difficult for viewers to browse through them without propercategorization, and ev ...

The Necessity of AI Art Image Corrections and an Automatic Optimization Program for It

2023-05-11 16:00

Many of the AI art images on this site required adjustments to contrast and brightness. Without them, it is sometimes difficult to discern what is depicted in the images. Until now, I have made these adjustments manually, but I have developed a program that automates the optimization proce ...

Troubles with post-processing AI-generated “pointillism” images

2023-05-11 13:24

When providing prompts such as “impressionism” or “Seura” to Stable Diffusion 2.1,one can obtain images in a pointillist style. However, these images do not look very good when zoomed in,so I struggled to make them look as clean as possible. I used image processing ...

AI art tends to draw broken faces and is better at non-realistic images

2023-05-02 14:31

Many of the faces of people drawn by AI art by chance are disfigured and uncomfortable.Here are examples. I donʼt think this would happen if I had it draw with elaborate words related to people, but I often see faces like this when I let it draw freely.Many people draw realistic image ...

“Sleeves” belong to women?!

2023-04-14 14:30

As in British folk song “Green sleeves”, "sleeves" were tied to women.In that case, I thought that if I gave “sleeves” to the AI, it would draw a woman, and I was right. ...

Various images created by the coined word “Gital”

2023-04-08 09:29

As an example of the creation of various images from short coined words,here I show some example images generated by Stable Diffusion 1.5using “gital”.I submitted the above content to Text toimage school board in Prompt Stack. ...

Participated in the Prompt Stack

2023-04-08 09:29

I want to widely introduce the method of creating various images from a short prompt, which is different from the conventional method of creating art thatuse long prompts.As one way to do this, I decided to join the community called Prompt Stack. This is because it is difficult to access p ...

ChatGPT Successfully Guessed the Location of AI Art Images

2023-03-30 08:27

When asked about the images on this site, ChatGPT was able to correctly guess the formatof the directory and the identification of the images. Here is a record of part of the conversation with ChatGPT.Q. Are there any images on this website that resemble Edvard Munch’s work? https://www.da ...

The effect of negative prompt “painting”

2023-03-21 10:54

Recently, when I let Stable Diffusion 2.1 draw AI art, I often used “painting”both as positive and negative prompts.Using the same word, I specified a negative prompt and the other did not.Then, the specified one generates pictures that has nothing to do with drawing a picture, ...

AI draws nature and rural landscapes “vaguely” and attractively

2023-03-19 22:59

If you give AI (Stable Diffusion) a short and meaningless prompt,it draws not only city scenery but also natural scenery and rural scenery in an abstract (vague) way.Let’s take a look at some images of these fascinating landscapes. ...

“Ambiguous” city scenery drawn by AI is attractive

2023-03-19 22:50

By giving detailed prompts to AI (Stable Diffusion), you can draw realistic city scape.Conversely, short, meaningless prompts generates abstract (or ambiguous) city scape.What do you think about this kind of scenery, especially the night view, that stirs the imagination? ...

What is the picture drawn by AI with “[”?

2023-03-19 16:34

The picture below was drawn by Stable Diffusion 1.5 with “[” as the prompt.What is that blue thing in the middle with a border?When I first saw it, I thought that this was the view of the sky from the courtyardof the building, although the surroundings do not look like building ...

AI covered human face under bird face

2023-03-19 16:21

The picture below was drawn with Stable Diffusion 1.5 by giving “>” as a prompt. However, if you look closely, there is a human mouth underneath, so you can see that what looked like a beak was a human nose. Since the eyes of a person overlap with the eyes of a bird, it can be seen as wear ...

What AI Likes: Collage

2023-03-17 14:31

Collage is often drawn when AI is given a “weak” theme.When one image does not have enough power to unify the entire screen, it seems to divide the screen and draw.As a method of division, there are cases where it is neatly divided into 4, 9, etc. rectangular areas, but it is more like a r ...

What AI Likes: Collage

2023-03-17 14:31

Collage is often drawn when AI is given a “weak” theme. When one image does not have enough power to unify the entire screen, it seems to di ...

Spelling that can create various images with AI art¥

2023-03-15 12:08

!I continue to try to create AI art by giving coined words.It can be seen that there are cases in which a slight change causes a sensitive response.For example, you can see that the word “cat” is resistant to spelling changes.“catt”, “catu” and other sp ...

AI image editing

2023-03-15 07:30

Images generated by AI are often dark or have low contrast.Therefore, the images published on this site are easily edited.The purpose of the editing is to make the images created by AI easier to see, not to rewrite them.Even so, I sometimes end up editing too much.There are two current edi ...

AI that imitates paintings by Monet and Van Gogh

2023-03-15 07:30

If you give AI the name of a painter, it will often draw a picture imitating that painter.Normally, various painting materials are selected, but depending on the artist, if you change the random numbers and drawI sometimes draw many pictures that imitate a specific picture.There is no prob ...

I want AI to draw more abstract images than photorealistic images -- use emoji 🖼 for that

2023-03-14 16:34

It seems that there are many people who try to draw realistic women with AI art.It’s something that appears in games.However, it doesn’t work very well.The face can be drawn relatively well, but the fingers seem to be difficult.Therefore, I think that I get more out of abstract paintings t ...

Trying models other than Stable Diffusion 1.5

2023-03-13 13:45

I’ve been using Stable Diffusion 1.5 for most of my images, but OpenArt.aiI also tried models such as OpenJourney, Analog Diffusion, Anything that can be used withSo far, I’ve also tried Stable Diffusion 2.0, but 1.5 works better, so I’ve been sticking with it.I’ve also used DALL E 2 for a ...

Images with overlapping arcs created by the “Infin” prompt

2023-03-13 13:20

We continue to make AI create images with coined words, but Stable Diffusion 2.1“painging” as a negative prompt by giving the word “painging infin”, it creates various images with overlapping arcs.I have only tried about 30 images so far, but just by changing the ra ...

If you give “painting” as the positive and negative prompts...

2023-03-13 13:00

If you enter a word with “🖼” as a prompt in Stable Diffusion 1.5, it will usually draw a picture.However, Stable Diffusion 2.0/2.1 does not draw the picture as you think.Adding “painting” often draws a picture, but it tends to use bold colors.Perhaps, if the spellin ...

Why can ChatGPT guess images that shouldn’t be visible?

2023-03-13 11:30

I showed ChatGPT an image generated by Stable Diffusion 1.5 and tried to guess what was drawn.Image from Dasyn.com AI art siteI tried to pass the URL, but I made a mistake and the URL on the MacBook that can’t be seen from the outsideI gave the . So I asked the following question.What shou ...

We asked ChatGPT why AI draws cats with “jiji”

2023-03-02 15:40

When I let Stable Diffusion 1.5 draw pictures with the words “jiji”, most of them drew cats.When I looked it up on Google, I found out that a cat named Jiji appears in Kiki’s Delivery Service.When I asked ChatGPT, I found that not only that, but there are also things related to ...

Answered a function call when I heard a funny spell on ChatGPT

2023-02-19 07:51

I asked ChatGPT,“Could you tell me a short spell that makes an interesting image with Stable Diffusion 1.5?”Then, some programs came back as answers. An example is1. Generate random noisesdnoise[0] out_color("type=octave, lacunarity=2, persistence=0.5, octaves=6, seed=1");out_c ...

AI made a circle by arranging things¥

2023-02-16 08:20

!AI (Stable Diffusion 1.5) may create a circle by arranging objects as follows.The prompt (spell) used is ``coronaviʼʼ. In other words, it can be said that AI arranged things to create the shape of the coronavirus.The image on the right has many protrusions, which are characteri ...

When AI was given the number “7”, it arranged objects to create seven shapes

2023-02-16 00:32

AI (Stable Diffusion) sometimes arranges objects to create shapes such as letters.After entering “7” as a prompt (spell), I arranged apples and wrote the letter “7”.Since it cannot be assumed that the AI ​​has learned this kind of photo,In addition to knowing the sh ...

How to make AI art that outsmarts ChatGPT

2023-01-28 08:26

I asked ChatGPT what it would be like to make AI art the way I do it now.I did not get the answer I was hoping for, but when I finally asked what would happen if I useda word that was not in the training data, I got a satisfactory answer that it was difficult to predict.At first, I asked, ...

What humans should do is outwit ChatGPT!?

2023-01-28 00:41

I used ChatGPT a little.That made me think that what humans should do is to outwit ChatGPT.ChatGPT is learning a huge amount of existing information.So, if you ask a question, it will give you a common sense answer in great detail.This is useful.However, it can hardly answer what is not in ...

When I let AI draw a picture of “Gerrymander”, it became “Gerrymander + Mandala”

2023-01-24 17:45

I wondered what kind of image I could get if I gave AI (Stable Diffusion 1.5) the coined word “gerimandala”, which is a combination of “gerimander” and “mandala”.As I intended, I was able to create something that looks like a combination of a “gerr ...

You can create endlessly interesting images by using Dali as a theme with AI art

2023-01-21 08:46

If you use Dali as a joke with AI art, you can create endlessly interesting images. Perhaps because Stable Diffusion also sees photographs of Dali’s performances, Dali himself appears frequently. There is Dali who became a cat in the attached image.Dali AI Art Poster#aiart #stabledifusion ...

Made the AI ​​draw a picture of the fictional artist Kanskoy

2023-01-18 10:47

The spelling “Kanskoy” seems to be recognized as a painterʼs name by AI (Stable Diffusion 1.5).Entering “Kanskoy” in the prompt (spell) draws a picture with a certain style.It can draw a picture of a woman relatively well.Of course, it draws pictures of men as ...

Tasty AI art created by layering pictograms

2023-01-18 10:39

I have been making AI art with pictograms.Combining constellation letters (such as ♒) with 🌠 and 🌌 creates patterns that would not appear on their own.#aiart #generativeartA list of such pictures can be found on the next page.AI art with zodiac and astrology pictogramsNote: This article is ...

Amazing AI art made from meteor emojis

2023-01-12 16:01

I wrote separately that you can create a cat image with a cat emoji, but just by inserting the 🌠 (meteor) emoji,It turns out that Stable Diffusion 1.5 produces surprisingly clean images.I will only include four here, but Sky/Stars/Universe contains various images, including related images. ...

Why does AI draw such a picture by “Corona”?

2023-01-08 15:03

When I put the word “corona” into AI (Stable Diffusion 1.5), the following image was obtained.It is certainly round like the coronavirus, and there are protrusions around it.However, it seems that there are many things that are not similar to coronavirus, such as a person-like ...

Does art AI like round things?!

2023-01-08 14:19

AI art (Stable Diffusion 1.5) often draws concentric patterns.The corona is round and many mandalas are also round, so it is natural to draw a round thing when I put those words in, but when I put in the coined words euclidean, ab, I drew a round thing.Here are just four examples (corona, ...

Art AI replaces human faces in portraits with cats

2023-01-08 11:54

If you enter cat and the name of the artist at the prompt in AI art, it will often draw a cat instead of the personʼs face in the painterʼs style portrait. I find it interesting that a cat draws a cat from the composition of a man drawing a man.#aiart #generativeart Images genera ...

A picture in which many faces overlap each other generated by AI art

2023-01-08 11:48

Picasso drew a picture in which parts such as a human face were rearranged.On the other hand, if you put Picassoʼs name in AI art (stable diffusion), I think it would be interesting to draw a picture with many faces overlapping each other.In music, canon and fugue overlap the same mel ...

Enjoy unexpected images created from short words with AI art

2023-01-08 11:40

Most people who enjoy AI art add words (prompts) to create the image they want. However, I think that enjoying unexpected images created from short words is also one of the ways to enjoy. The example below is an image from the name of the artist Giacomo Balla and some variations of it.#aia ...

An amazing variety of images made from “Tullio Crali” with AI art

2023-01-08 11:35

Using short prompts in AI art can yield a surprisingly wide variety of images. For example, in Stable Diffusion 1.5, if you try it repeatedly with the name of a real painter called Tullio Crali or a slightly changed version of it, you will get a lot of images with a painting style differen ...

Draw cats from emoji 😺 😻 with AI art

2023-01-01 01:00

I am addicted to AI art and have it create various images. As a result, we found that AI (Stable Diffusion 1.5) also understands emoji. For example, you can create the following images with 😺 and 😻.AI art images of various emojisFlHM7uWaMAA5PMtArtist names or coined words that are strong/weak against spelling changes in AI art

AI art poster and postcard press release

Sale of AI art posters and postcards

Let Stable Diffusion draw a coronavirus...

The Road to AI Haiku Art: Phrases Shortening

AI art also knows kanji and emoji!?¥

Let AI art draw an image with a word that transforms “Mondrian” into a word that can develop

Road to AI haiku art: Seasons

Draw cosmos with AI art¥

I tried to draw AI art by playing with the spelling of the artist’s name¥

Is AI good at abstract paintings?!¥

I tried drawing using multiple words meaning “cat” and “abstract” in AI art¥

AI canʼt draw plankton, so it canʼt make Kandinsky-style pictures well¥

Created AI art of cats and posted it on SNS


Dasyn.com デイシン
Created: 2023-06-23 15:20   Edited: 2025-01-22