Google's New AI, Bard - 1st Impression

You might expect a well-known and powerful company like Google to eventually enter the AI competition arena. One could assume that Google has had a major AI-like system for years, used for web searches and information gathering. Or perhaps not? My first impression of Bard, Google's response to ChatGPT, is, "Holy $#%&! This might be the worst AI I've encountered in over a year!"

Let me explain the level of detail I usually expect. As a toy designer, I use AI to help me collect images of toys, conduct market research, and explore styles and color schemes, among other things. Recently, a client requested generic toy robots but wasn't sure about the exact appearance. This is where AI shines, by generating images/mock-ups in a variety of possible styles.

A very traditional cute metal toy robot design - requested in metallic silver without colors

Another variation showing different flocked materials and some color accents

Above are two examples of robot toy designs, created by Dall-E (also known as ChatGPT-4).

I find Dall-E to be the best at "listening" to exactly what I'm asking for in my toy design prompts. Many other image generators tend to overlook key details mentioned in the prompt. Now, let's look at what Google's Bard created. Note: This was done using the exact same prompts I used with Dall-E.

Google . . . I'm speechless. (Was this thing shot in the chest?)

This one's even better. And by "better" I mean worse.
Why the colors? Why does it look old, grimy and used.

These things would give nightmares to the Island of Misfit Toys. They do look as if a child created them. The colors are so muted and conflicting. If AI was involved with these designs, I'd like to know how it came to this point. Personally, I feel Google shouldn't have released Bard to the public. They should have kept it closed until their technology was at least on par with the AI world a year ago.

UPDATE: My Top 10 AI Images of 2023 - - - - - - - - - - - #1 Moldy Bread Monster

- - - - - UPDATE - - - - -
Needed to update with the latest version of the Moldy Bread Monster. All AI generators have been updated several times since the creation of the first image -- so this is now the latest version circa January 2024. I mean, this newer guy is actually made of BREAD! The former dude was more wool-like.

Love, love, love this updated version. Much better mold.
The goofy expression has a very Adventure Time/Gravity Falls vibe.
The red veins in the eyes are simply priceless. Bravo, AI -- Bravo!

- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - 

Originally posted on 12-22-2023

NOTE: These images were not arranged in any level of greatness or priority. I’m gathering the 10 images I feel represent my best AI results in 2023. I hope you find them amusing, interesting, and inspiring too.

Behold, the slice welding Moldy Bread Monster

Meet the Moldy Bread Monster, born on the day there was a half loaf of moldy bread in our kitchen, and I REALLY wanted a sandwich! (Hate when that happens.)

The MidJourney prompt was simple, though I added the word “dancing.” The dancing wasn't quite captured, which often happened with version 4 back then. But the character's uniqueness, especially with it seeming upset about a slice of “good bread” in its left hand gave me a good laugh. (Especially since I was in a similar situation mere hours ago, complaining and holding a “bad slice’ of moldy bread.)

Runner-Up AI images of 2023

Here are a few of the AI images I generated during 2023 that I felt were interesting enough but didn't make the top 10. 

Manic Santa Claus was a popular trend during December, this one was my weirdest

Using someone else's prompt, I edited it to portray Billy Elish as a warped version of the Poison Ivy character, (mostly picked her because she used to sport green hair.) This came out much better than expected.

Red Sonja fighting a gigantic viper creature

This one came out of nowhere! I believe I simply asked for a "Vivid Colorful Santa Hat" for photos of pet costumes. I have NO IDEA why this was the image I got instead. Disturbing.

Another disturbing image, this was "a photo of Humpty Dumpty circa 1910."  

A beautiful fantasy city scene. I especially like the warm lighting in the foreground and the insanely doom-inspiring details of the towers in the distance.

My Top 10 AI Images of 2023 - - - - - - - - - - - #10 Darth Vader Worn Revisioned

There used to be a great website called Illustration Friday. Every Friday, a new word was given, and artists would submit images based on that week's word. For the week when the word was "WORN," I chose to interpret it as "damaged and shabby as a result of much use." Using this definition, I depicted Darth Vader as worn out from a long day. Many years later, I uploaded that image into Leonardo.ai, activated the 3D settings, and received this result:

In less than 20 seconds, my 2D drawing was transformed into a 3D version. The only disappointment was that the smoke trails from the lightsaber looked more like gray wood with stones. However, the Darth Vader character turned out very well, with excellent lighting, reflections, and textures.

The original sketch was digitally drawn within an hour and posted on the Illustration Friday website.

Sometime later, I hand drew the design using brush pens, Copic markers, and a white Gelly Roll pen.
This was the image uploaded into Leonardo.ai to be revised by AI technology.

Baby New Years 2024 - The Year of AI

HAPPY NEW YEAR!!!

Every year around this time, I recreate versions of my original Baby New Year character kicking in the new year, which I created 10 years ago in 2013. It has always been very popular among visitors to this site, and it definitely gets stolen by MANY people who download other's artwork to call it their own and sell it online.

This year, something a little different: I'm using AI to create Baby New Year based on my original drawing.  

Behold, the AI-generated version of the 2024 Baby New Year

To be honest, this little guy was a combination of 4 AI images, which had to be altered, assembled, and touched up -- so it probably took me much longer to do than simply draw a new illustration. 

The first Baby New Year image from 2013, using brush pens, a Gelly Roll pen, and Copic Markers

The original concept pencil sketch in 2013

My Top 10 AI Images of 2023 - - - - - - - - - - - #9 Cthulhu According to AI

In his famed novel, Call of Cthulhu, H. P. Lovecraft describes a statue of Cthulhu as:

"A monster of vaguely anthropoid outline, but with an octopus-like head whose face was a mass of feelers, a scaly, rubbery-looking body, prodigious claws on hind and fore feet, and long, narrow wings behind."

I prompted his exact wording into Leonardo.ai and generated several dozen images following that description. All of the images generally had similar aspects, but the level of difference was totally unexpected. Next, I entered more prompts, but this time provided much more details about character placement, size/proportions, and the background/environment setting. Which produced this:

This image above instantly became my favorite image of the group. Cthulhu's size in the far distance, the foggy air, and the mysterious shielded warrior, such fantastic visuals generated in one image. 

More prompted images of Cthulhu are seen below, seen in ranges from traditional to insectoid.

My Top 10 AI Images of 2023 - - - - - - - - - - - #8 Joker Circa 1910

A nice feature of AI image generation is that you can suggest creating an image set in a specific period. For example, imagine if the Joker fought Batman around 1910. AI tends to do a better job with individual characters than with multiple people in one image. This is often because when describing several people in text, AI can lose focus on which person should receive specific details, leading to a tendency to morph two or more characters together.

[Sidenote: If you’re interested in what inspired the character Joker, check out the silent film “The Man Who Laughed”; a truly eerie depiction of a character that would influence comic book/movie culture for decades to come.]

Fabulous sepia toned image of the Joker back in 1910; the expression alone is priceless

Here are some samples of attempts to depict the two characters in a 1910 fighting scene. Note how Batman often appears more manic than his classic stoic expression. It's as if the Joker is wearing the Batman costume. 

Additionally, Batman's armor is too technically advanced to represent his costume accurately in that timeline. Apparently, it seems AI is pulling reference to the Batman’s costume mostly based on the more modern Batman films instead of the cloth based costumes in earlier times.

My Top 10 AI Images of 2023 - - - - - - - - - - - #7 Celebrity Not-Quite-Right-Look-A-Like

To me, nothing is more impressive than when an AI program successfully captures the likeness of a famous person. However, AI companies are now striving to limit this aspect of image generation due to the growing threat of Deep Fakes in multimedia. If left unchecked, this could lead to worsening political scenarios, increased legal actions from celebrities regarding the misuse of their likenesses, and a general rise in public deception, fueling falsehoods and conspiracies. This could potentially lead to complete anarchy in a world already susceptible to misinformation.

This image doesn’t quite look like Joey Heatherton, but it captures the style of her time.
The green smoky background, black dress, bobbed hair, and mystic pose makes this a fantastic image.

Most American-based AI sites, tend to not do great exact likenesses. They do permit likeness usage generally as a reference in prompts to generate similar-looking characters. But if referring to anyone famous in your prompt, the censorship restrictions are usually maximized. (This is not true from many foreign AI sites, especially those based in Japan.)

In this particular case, I referred to a lesser-known celebrity from the early '70s who I saw on TV all of the time on variety, talk, and game shows as a kid: Joey Heatherton. (Not only was she cute, sexy, and bubbly — she was the inspiration to Catherine O’Hara’s character, Lola Heatherton, one of my all-time favorite comedy shows, SCTV.)

Catherine O’Hara as the over-the-top-manic-laughing, Lola Heatherton, on SCTV in the late ’70s
Yes, that's Catherine O'Hara, known for Home Alone, Beetlejuice, Best of Show, and Schitt's Creek.