Since last week, social networks have been inundated very special images. Generated by artificial intelligence, some are very faithful to the title. Others appear deformed. The truth is Give her has become a trend: it is a site that generates a representation of what the user writes. Even what we ask for does not exist in the real world.

Dall-E is a technology created by a San Francisco company, OpenAI. The company aims to create “safe and useful general artificial intelligence” and is recognized in the computing environment to work on GPT-3 , a tool that generates images from very basic indications such as, for example, a text. They also use Co-pilota tool that helps automate coding for software engineers.

What does all this mean? That Dall-E can generate images with words. From an algorithm that considers 12 billion parametersthe system creates images with letters only.

It is therefore not a bank of images: images are created from texts. So, for example, if we want to see David Bowie cooking, we can ask him to design it for us. And the result can be a little creepy at times, but no less faithful to the representation of the input input.

Artificial intelligence recognizes the basic parameters: “Diego Maradona having tea with Darth Vader” is a situation that has never existed and will never exist. Yet opening this article recreated the image.

That’s what it is, how it works and why this tool is all the rage.

First: why Dall-E?

The name is a play on words between the surrealist painter Salvador Dalí and Wall-E, the Pixar character. In January 2021, the company unveiled the first version of the tool on which it worked a limited resolution of 256×256 pixels.

The second version began testing in April, but in a closed beta.

Already with images of 1024×1024 and extra tools, began to circulate.

And last week they started flooding social networks: user threads on Twitter, Instagram images and all platforms fell prey to Dall-E

Dall-E vs. Dall-E mini: why are there two Dall-Es?

Basically Dall-E mini follows in the footsteps of another more developed software known as Dall-E 2, but although the two platforms they use the same softwaretheir approach and final results are very different.

Dall-E 2 is primarily a private service that offers access based on a long waiting list. Dall-E mini, on the other hand, is free for everyone. This means that Dall-E mini may often not be available with too much traffic. and slower loading times.

Even the mini equivalent struggles to match the same kind of quality as the original. Indications can often appear blurry or of poor quality. Dall-E 2, on the other hand, is for the most part accurate and of high quality.

I mean, at the end of the day, Dall-E mini would seem like a gateway to the paid service.

The curious thing is that the free service generates more bizarre images and this has made them go viral much more than those of Dall-E 2, what would the payment be (and more powerful as a technological tool).

How to use Dall-E mini

It’s quite simple: go to the Dall-E mini website. Here the user will be prompted to enter a command and execute it. The command must be in English, but in natural language: for example, if we want to see a corgi dog wearing a suitYou have to write it in English. “Cogi in tuxedo”.

Two questions: the first is that the process takes about two minutes in general to result.

The second is that it usually gives an error because the servers are crashing. If this happens, you need to try repeatedly times until you take the order.

You can also request different styles of images with your command by adding phrases like, “Like a baroque painting”, “With panoramic lens” and so on.

The limit is the imagination. And social networks are the first to know.