Artificial Intelligence (or AI ), or that branch of technology that studies how to make computer systems “intelligent”, that is, capable of simulating the capabilities and behavior of human beings, is making truly amazing steps forward, allowing you to perform operations that were once unthinkable.
If on the one hand this technological progress raises some important (and sometimes worrying) issues concerning ethics (see for example the phenomenon of deep fakes ), on the other hand AI is useful for carrying out various operations, some of which are useful. to do various kinds of jobs, others rather frivolous (but definitely fun!).
In this article I want to show you everything by explaining how DALL · E works , an online portal that, using Artificial Intelligence, allows you to create images using simple textual descriptions. You understood well! Images that do not exist are created by an “intelligent” algorithm that creates them by reading and interpreting the text command given by the user. If I have intrigued you, stop the chatter and have fun!
Index
- What is DALL · E
- How to use DALL · E
What is DALL · E
Before showing you how DALL · E works , it is my duty to explain to you what it is . As I mentioned in the introductory lines of the article, DALL · E is an Artificial Intelligence algorithm, developed by OpenAI , which is capable of generating images starting from textual descriptions, making itself capable of learning concepts of space and time.
Among the various things it allows you to do, there is the possibility to make realistic changes to the context of the images that are generated with DALLE or to the images that are uploaded to the platform, which also allows you to create different variations inspired by the image “original”.
We now come to the question of costs . At the time of writing, DALL · E can be tested for free in its beta version, but beware, it works with a credit system : to be precise, you get 50 credits upon registration and then 15 free credits every following month .
Each credit offers the possibility to make an original request to DALL · E ( prompt , in jargon), and each request returns four images . Otherwise, always with a single credit, it is possible to make a request for modification / variation , which returns three images.
If you need more credits, you can purchase them with prices starting at 115 credits / $ 15 (which is equivalent to 460 generated images).
How to use DALL · E
Now we can finally move on to the funniest part of the article by seeing how to use DALL · E. Basically, just go to the home page of the service, complete the registration and start generating images through text descriptions.
To proceed, therefore, first go to the main page of DALL E and register via email ( Email address ), via your Google account ( Continue with Google ) or via your Microsoft account (Continue with Microsoft Account ), taking care to use the text field or the appropriate button and following the instructions on the screen to complete the procedure.
After registration, indicate your name and surname in the appropriate text fields , click on the Continue button , specify your telephone number in the appropriate text field (making sure that the correct international prefix is selected from the menu showing the flag of your country of residence) and click on the Send code button .
Once this is done, enter the 6-digit code from the SMS that was sent to your number to validate it. Then read the DALL · E welcome message and click first on the Continue button and then on the Start creating with DALL · E button .
At this point, use the text field located in the center of the page to write the word or phrase that will allow the DALL · E algorithm to generate a corresponding image, then click on the Generate button .
From the tests I have carried out, it seems that the DALL · E algorithm is also able to understand the Italian language, so you don’t necessarily have to write in English. In the example below, in fact, DALL · E has managed to create four versions of a dog eating a pizza and the result is really interesting. However, for greater accuracy it is recommended to use English.
Once the image of your interest has been generated, click on the symbol (…) that appears when you hover the mouse (you can see it at the top right of each generated image) and, if you wish, generate other versions of the image by selecting the wording Generate variations .
By selecting the Edit item , on the other hand, you can access some tools to retouch the photo, thanks to which it is possible, for example, to delete the background (the rubber icon ) and superimpose other images (the postcard icon with the + ). Once satisfied with the result obtained, click on the button with the arrow pointing downwards (top right) or select the Download item (after clicking on the button (…) at the top right of the preview of the image itself).
Do you want to upload a photo to transform with DALL · E? This can also be done (as long as you don’t upload photos of people without their consent). To do this, click on the Upload button (top right), upload the image to edit and click on the Continue button . Subsequently, apply a crop to the image using the DALL E editor, click on the Crop button and then decide whether to edit the image ( Edit image ) as I explained earlier or whether to create variations automatically ( Generate variations ).
Don’t have the faintest idea which image to create with DALL · E ? In this case, know that you can click on the Surprise me button (top right) and then on the Generate button to create images suggested by the algorithm itself. Interesting, right?