All you need to know about Dall-E Mini, the AI-powered art maker

There is no shortage of Nightmare Fuel on the Internet. The latest source may just be an artificial intelligence tool called the Dall-E Mini, which has garnered attention on social media thanks to the bizarre, funny, and sometimes disturbing images it creates from text prompts.

Images generated by AI showing surfing Batman.

Surfing Batman.

Erin Carson / Dal E mini

Dall-E Mini It allows you to type in a short phrase describing an image, one that theoretically exists only in the depths of your soul, and in a few seconds, the algorithm will show it to your screen.

Odds are, you’ve seen some appear on your social media feeds while people are thinking of the most wild motivators they can do – maybe it’s Jon Hamm eating pork, or Yoda robbing a convenience store.

This isn’t the first time that art and artificial intelligence have captured the internet’s attention. There is a certain attraction to seeing how an algorithm deals with something subjective like art. In 2016, for example, actor Thomas Middleditch made a short film based on a script written by an algorithm. Google has produced more than a few tools that link art and artificial intelligence together. In 2018, the Arts and Culture app allowed users Find doppelgangers in famous paintings. or Google automatic drawing It will figure out what you are trying to draw, and fix it for you.

There are other text-to-image systems, such as OpenAI’s Dall-E 2 and Google Photos-This is not available to the masses yet.

Here’s what you need to know about the Dall-E Mini and its AI-generated art.

What is Dall-E Mini?

Dall-E Mini is an AI model that creates images based on the prompts you make. In an interview with post iBoris Daima, programmer, said he initially created the software in July 2021 as part of a competition set up by Google and the AI ​​community called Hugging Face. Daima did not immediately respond to a request for comment.

How it works?

Anyone can type a prompt and press the Run button (although it is possible that you will receive an error message about traffic to the tool and have to try again). The Dall-E Mini will publish its results in the form of a 3×3 grid containing 9 images. A note about the tool on its website states that it was trained on “unfiltered data from the Internet”.

How good is artificial intelligence?

Unsurprisingly, the Dall-E Mini has little success or error. In an interview with I News, always said AI is better with abstract painting, not with faces. The landscape of the desert is very beautiful. It looks like Dolly Parton’s pencil drawing could steal your soul. Paul McCartney eating turnip will take years of your life.

A cat made mostly of pink laser light.

Here’s a cat made with a laser.

Erin Carson / Dal E mini

However, Dima said the model is training (this ability to learn is one of the things people love — and fear — about AI), which means it can improve over time. And with the widespread use of the Dall-E Mini, the goal is for you to stumble upon the most unusual images you can think of, not necessarily for a perfect impressionistic view of the Waffle House. It’s more about dreaming about the most weird non-existent images – which probably shouldn’t exist – and bringing them into damned existence.

Dall-E also has a note that creating images can have a less fun aspect and can be used to “reinforce or exacerbate societal biases”.

Is the Dall-E Mini related to the Dall-E 2?

No, they are not related. Dall-E 2 It is also an AI image generation tool that was launched as a research project this year. It was created and published by the artificial intelligence research firm OpenAI and is not widely available.

What types of photos do people create?

On social media, you find no shortage of quirky Dall-E Mini creations, from Thanos at Walmart looking for his mother, to Jar Jar Binks who won a Great British Bake Off. Here are some other highlights.