News

DALL-E 2 and Midjourney, two allies for industrial design

Times evolve and the conception of certain disciplines also as a result of the application of new technologies that seek to provide greater precision to the results. This is the case of DALL-E 2 and ChatGPT either midjourneyinnovative tools that put artificial intelligence at the service of companies and that seek to replace conventional methods.

Although there is debate about the ethical use of AI imaging platforms, big tech companies like Google and Microsoft are competing to take control of the market, analyzing the repercussion that they can have in sectors as specific as industrial design. In this sense, the assistant professor of design at the Rochester Institute of Technology, Juan Noguerahas ensured that they will not replace the critical thinking acquired, but that they will improve the techniques.

DALL-E, the origin of everything

Talking about DALL-E is talking about an image generator, created by Open AI, and that, like ChatGPT, is in charge of using the AI ​​to generate images from text, in such a way that it will draw everything we want from nothing, since it is capable of understanding natural language. This is possible thanks to the fact that it is based on GPT-3a language model trained with millions of parameters.

DALL-E’s wake has been followed by other tools such as MidJourney and Stable diffusion, although its degree of perfection and precision is unlimited. And it is that he has been trained with a large library of photographs and artworkbeing able to combine concepts, styles and attributes for a single image.

The first version was presented in 2021, although in 2022 OpenAI presented DALL-E 2 and now it is expected that it will emerge in the medium term DALL-E 3 to generate better images from texts.

The creation process that DALL-E follows is similar to that of other similar AI tools, as it follows a diffusion model. First encode and interpret text (both in English and Spanish), identifies the features and styles to draw, and then creates image information from the initial request. finally worth it a decoder that paints the image starting from said text. Every time we ask him to draw something he will do it in a different way, since his interpretation will vary.

To use DALL-E it is as simple as access the web enabled by OpenAI and ask him to make a drawing based on our request, and in a matter of a few seconds he will provide four different images from which we can choose the one that best suits our needs. You may even be asked to generate new images from it. promtp.

Its application to industrial design

DALL-E or Midjourney are capable of revolutionizing industrial design, understood as the creation of everyday objects from a deep investigation after which they will create their own inspiration boards. For example, jonathan ive (designer of many Apple products), was inspired by luxury watches to create the Manzana watches. Now, with these new AI tools, it will be possible to obtain for the same object thousands of designs, shapes and textureschoosing the one that interests us the most.

Noguera himself assures that from those 1,000 images that DALLE-2 or Midjoruney can generate on the same thing, new results can be obtained, not intended by the designer, which allow the product’s characteristics to be improved. He even devised himself a nostalgic souvenir project to connect people with the place they had visited.

With this experiment, Nogueras obtained such unusual products as erasers or boxes of frowning tissues. He inserted the word ‘Ancient’ into his directions, and little by little he was getting iconic symbols of the place he was referring to. He finally managed to create a mockup in Photoshop of a tissue box in the shape of the Agua Volcano located south of the city of Antigua (Guatemala).

He then created a 3D computer drawing, a CAD model, using actual terrain data and produced a functional product, even incorporating matches into it, i.e. creating a volcano that could contain matches.

A door to the future

AI technology is here to stay and to perfect the realities that exist, and even creating new horizons. Thus, Noguera considers that these tools will form part of the short term industrial design training. He himself has had his graduate students at the Rochester Institute of Technology generate his own products from these efficient tools.

Perhaps this is the sign that we should not be alarmists, but that AI can be used as a source of inspiration inside the designer’s toolboxbut never as a substitute for your source of knowledge and work.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *