It has been claimed that an art computer may soon draw your ideas for you.
Text-to-image AI illustrators are rapidly improving in sophistication and are appealing to many as a way to produce high-quality images instantly from text prompts.
Using a computer to draw images instead of a human illustrator can, however, make artists think about the safety of their work.
It takes a lifetime to perfect and hone the motor coordination and creativity needed to be a professional artist, and the prospect of being replaced by a neural network will leave a bad taste in the mouth of many. However, this problem may still be many years away, with the technology still in its infancy.
Illustrative programs, such as Google’s Imagen and Open AI’s DALL-E 2, use neural networks to organize images into categories. These images coincide with specific text messages and visually amalgamate ideas into the final product, which hopefully reflects what the human user imagines as accurately as possible. This is a learning process, with schedules supposedly getting more accurate by the day.
Huge amounts of computing power are required to train AI illustrators on huge datasets. This acts as a limiting factor, allowing only the largest tech conglomerates with deep pockets in the running to develop the technology.
Unfortunately, these programs are not available to the public, and thus skeptics claim that the technology is simply not yet there to navigate the nuances of art.
Tech companies responded by saying that the risk of their software being misused to create malicious images and content is too great to allow for open access.
However, steps are being taken to address this problem. Google is developing a list of malicious images and concepts that need to be blacklisted, thus enabling the technology to be deployed. Whether this is just a deadlock tactic to allow programs to have more time to develop is up for debate.
Human artists, on the other hand, can welcome AI in its current form, especially those who work with digital tools. Companies like Adobe have slowly been adding AI capabilities to their products. Speeding up workflows by using programs to automatically crop shape outlines or find specific video frames allows artists to free up time to truly experiment and showcase their creative flair.
By eliminating repetitive tasks, human ingenuity can take firm control, something that AI has no way of replacing in its current form.
No industry will benefit from this development more than animation. In Asia, where anime is a revered international cultural export, artificial intelligence tools that can reduce the need to draw frame by frame and automatically color scenes are a big deal.
The automatic coloring programs are now used by the OLM production studio, which is responsible for bringing the Pokemon franchise to life. The developers of the Chinese University of Hong Kong have further developed sophisticated coloring programs, such as Style2Paints, which are openly accessible on the Internet.
The AI revolution is far from automating an artist’s work. If anything, it contributes to their productivity, freeing up time to explore original ideas. The creative industry, however, will be affected by these developments. The amount of work and deadlines set by clients will change as the art process becomes more efficient.
But expecting a machine, rather than a human being, to create a bespoke masterpiece is still a long way off.