When artificial intelligence designs objects
The constant debate … Can an Artificial Intelligence have imagination? To be creative? For many "never", for others "we have been working together for a long time and I have some studies", and for a few, of course! In a few years.
Artificial intelligence (AI) relies on vast amounts of data and algorithms to create its 'creativity' which is subject to randomness, for now (as some Silicon Valley master's engineer would say). We do not know what the future will be, but the Artificial Intelligence in design it is evolving by leaps and bounds.
Of course, human creativity is driven by feelings, hormones, neurotransmitters, etc. And many other factors that are hardly - if not impossible - replicable in an algorithm, but this does not prevent AI will accelerate the product development process and we have to start training (By the way, Google offers training from HERE).
Moreover, and as a subsection, the designer Phillip Starck created the first chair designed with AI a few years ago. You may like the final design more or less, but there it is …
One of the leading companies in artificial intelligence is Open AI. Almost every month, he launches a new experiment showing us the potential of AI.
This time, he has surprised us with the neural network called DALL-E (This is supposed to be a WALL-E / Dalí pun). Create new and unique images from a natural language written context that describes what we want to see.
Researchers have discovered that it can create new anthropomorphized versions (adaptations - transformations) of animals, objects, scenes, etc., that is, AI can show us new object designs, which did not exist before according to our guidelines.
The DALL-E has the ability to analyze text and combine unrelated concepts to create new objects
For example, based on two words; armchair and avocado. Contextualized in… «an armchair shaped like an avocado. an armchair that imitates an avocado »we obtain the following designs.
Some designs created are not bad at all. Or if we combine the words; clock, green, pentagon. Contextualized in… «a pentagonal green clock. a green clock in the shape of a pentagon »we have the following variants.
What's more, DALL-E provides new capabilities in 3D rendering through natural language. You can independently control the attributes of a small number of objects, and to some extent, how many there are and how they are arranged relative to each other.
Too you can control the location and angle from which a scene is rendered, and can generate objects according to precise angle specifications and lighting conditions.
And it could not only become a new tool that could be used in industrial design in the near future. They are also provided examples in the field of home interior design.
Create various interior scenarios based on the written paragraph… «a living room with two armchairs and a painting of the coliseum. the painting is mounted on a modern fireplace. '
Imagine being able to combine this tool with the parametric design of buildings or with others used that are already being practiced with artificial intelligence in the architecture sector, crazy!
And another example of how it is able to reproduce scenes in the field in the field of interior decoration.
As AI expert Alberto de Torres put it… “The risk is that design will become very homogeneous and lose its human dimension, as designers become guardians of AI-driven ideas rather than users of technology as a collaborative system and as a tool ”.
We will see how AI evolves in the field of design, but professionals in the field will have to start training in artificial intelligence so as not to miss opportunities if they want to survive in a world with so much competition.
The reference article is from the Openai page that in addition to having other application examples of the new DALL-E neural network, we can also play a bit with the text to recreate new designs.
If you liked this article, share it!
