The prospect of creating art using AI

With the great development of artificial intelligence (AI) technology, now artworks created from digital technology have become a potential development trend.

According to Argentine artist Sofia Crespo, this is a new trend when people write down rules for computers to use algorithms to generate new ideas and art models.

This field is beginning to attract the interest of many art collectors, with many of the works fetching high prices at auction. American artist and programmer Robbie Barrat sold a work called “Nude Portrait#7Frame#64” at Sotheby’s for £630,000 ($821,000). Four years ago, “Edmond de Belamy” was auctioned off at Christie’s for $432,500.

However, this burgeoning trend has faced a new challenge, as tech companies begin to launch AI technologies that can create photo-realistic images in seconds.

In the 1960s, artists in Germany and the United States paved the way for the trend of creating works of art from computers. Currently, the V&A Museum in London (UK) still keeps a collection that is more than half a century old, including the work “Plastik 1” created by artist Georg Nees in 1968. He used the Random numbers from the computer to create geometric designs for your sculptures.

Today, digital artists all work with supercomputers and systems called adversarial creative networks (GANs) to create images that are much more complex than artists in the past. GANs are a collection of AI technologies, capable of creating images from initial instructions, assessing whether the output is exactly as required. If the product is defective, the machine will send the image back to the AI ​​system to correct.

Artists such as Crespo and Barrat both assert that artists still play an important role in the art creation process, even though their working methods are not in the traditional way. Barrat said he did not create images, but created a system that can produce images, while Crespo claims she had to use many lines of code to bring about the desired results.

Tech companies are hoping to bring this technology to mainstream consumers. Google and Open AI are touting new technologies that can be creative and create realistic images for users without programming skills. They have replaced GANs with more friendly AI models that can convert common commands into images.

Google’s Imagen page has loads of such creative on-demand images, such as a cactus in a hat and neon glasses in the Sahara. Meanwhile, Open AI emphasizes that the Dalle-2 tool can bring about any art style.

Although the emergence of AI raises concerns that humans will be replaced by machines in many areas from customer care to journalism, artists see this as a potential rather than a threat. Artist Crespo tried out Dalle-2 and found it to be a new level of visual creativity. Meanwhile, Camille Lenglois of Center Pompidou – Europe’s largest cultural center and contemporary art museum – also supports this view, arguing that machines are not capable of creation and production. Image does not help make an artist.


Leave a Reply

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